BlueDustStudio / Chris Porro

Months back I bought a pair of Avantone MixCubes. Here is a summery along with a fuller explanation.

Summery: At $240 (passive pair) I consider my MixCubes a good investment towards better mixes. They don’t sound good. That’s not the point. I use them as a second perspective for checking levels, making sure things like bass and kick translate to the mid range, and comparing my mixes to reference mixes.

Reasons for purchase: For a while I mixed almost exclusively on my Mackie Hr824s. These are midfield monitors, with an 8″ bass driver, and what Mackie calls a bass radiator. The bass radiator is supposed to have the advantages of a ported design without the disadvantages. In other words extended low end without ringing problems. But if you check out the waterfall plot on the HR824 it sure looks like it rings a lot between 20-100hz. In fact after I purchased these I learned they have a reputation for  exaggerated bass. Not ideal.

I was not happy with my mix results. Then a friend and accomplished mixing engineer recommended I get a single MixCube. I also read a compelling case for MixCube type speakers in Mike Seniors “Mixing Secrets for the small studio“. BTW, that book is the best one I’ve ever read on the subject.

Here is my distillation of Mike’s points and a few of my own I hope.

• Mid-range first. Since a MixCube will not reproduce the very highs and very lows it forces you to get the mids right. Midrange is what all listening systems have in common. It is where the ear is most sensitive. And it will make or break your mix. You still need to check your highs/lows on a full range speaker.
• Avoid comb filtering. Whenever the same sound arrives at your ear at different times you get this bad thing called comb filtering. Comb filtering isn’t something to be taken lightly. It can sound pretty extreme at times. It can occur both acoustically in the air (direct signal is speaker, delayed signal is off wall) and digitally (a delayed signal played along with the original). If you want to hear the digital version put a track through one speaker. Copy that track and push it back less then 20ms. Play them together. Ok, but how does this relate to a speaker? Most speakers have 2 drivers (sound sources) and they have an frequency overlap. If your ear isn’t perfectly equal distance from them you will get some comb filtering. With one driver, and especially when using just one MixCube, comb filtering is  minimized.
• Single driver. You may wonder how a speaker with 2 or 3 drivers splits up the frequency range and gets everything just perfect. It doesn’t. And often the crossover is located in the all important mid-range. With the MixCube your precious midrange is minimally messed with.
• Un-ported design. Ported designs get extended lows. but they also get extended ring. Big picture this is a distortion of the signal. IE, what you hear is not what you have. A MixCube does not have much bass but what it does have is accurate The roll-off is natural without spikes.
• Lack of bass = Good thing. What excited room modes? Bass. So mixing with a low volume low bass speaker can help you avoid whatever coloration your room might be adding.

Two Camps: If you start nosing around and asking questions you might notice opinions on MixCubes are divided. Some people believe they are used for a “real world” or worst case scenario check. In other words how will my mix sound on some really bad consumer speaker. Others, like myself, see them as real mixing tools. Some very good albums (Michael Jackson’s Thriller) have been mixed on MixCube style (Auratones AKA”Horrortones”). I have also seen some modern mastering rooms with a set.

How they have effected my mixing? This is pretty subjective. I like to do quasi-scientific experiments whenever I can but I can’t think of a way to do this one. Sure I could do 2 mixes. One on the Mackies and one on the MixCubes. But all that might prove is I was in the zone one day and not the other. Or that I did more reality checks on mix A then mix B. In practice I wouldn’t mix 100% on either speaker anyway. Not anymore.

But I do find it easier to judge a lot of things on the MixCubes. Levels, eq adjustments, reverb levels. I believe, but can’t prove, my mixes sound better using the MixCubes.

One thing I’m considering right now is the way mixing in mono effects a mix. Something on my too try list is using a single Hr824 with my master buss summed to mono. That might reduce some comb filtering and make mixing adjustments easier.  This is exactly how I use the MixCubes. I have heard quite a few people endorse the single speaker with mono summing approach. Part of that is checking for phase problems should a mix go to mono (phone, mono radio, PA system) but part I suspect has to do with limited comb filtering.

Amp Pairing:  I’m a fan of active speakers. But no active MixCubes were available when I purchased. So I started looking into amps and cables. If it’s worth doing it’s worth overdoing to a fault right?

The first realization? I didn’t need much to power these. In fact, the Byrton 2B (50 watts per channel at 8 ohms) I ended up buying is overkill.

Like most specs a speakers numbers don’t tell the full story. Avantone lists a recommended power from 10-200 watts. But the real number to look at is the sensitivity. It looks like this: 93dB @ 1w / 1 meter. This says 1 watt will drive the speakers to produce 93dB spl (of some test signal I assume) measured at 1 meter. That’s a whole lot of loud for just 1 watt.

The way the math works out every 3dB increase requires double the wattage. A 10dB increase requires 10 times the wattage (decibels are Logarithmic).  According to this a 200 watt amp will produce just 3 dB more then a 100 watt amp. Wha? Yeah, it’s true.

Here’s how it works out for my 50 watt Bryston and the MixCubes.
93dB @ 1w / 1 meter so then…
103dB @ 10w / 1 meter
113dB @ 100w / 1 meter
Now take away 3 dB to arrive at 50 watts
110dB @ 50w / 1 meter

I don’t monitor at 1 meter, more like 2 or 3, but still 110 dB is loud. The signal I send the Bryston from my software mixer is usually -40 dB.

Bottom line: You don’t need much to power these for mixing at reasonable levels around 70 dB spl. I think 10 watts should do it. I definitely can’t use 50 watts.

One other thing of interest. You may see these listed as “full range speakers”. This does not mean they produce the full frequency range (far from it). It means the full frequency range is sent to one driver. Most monitors have a crossover and each driver handles a partial range.

Set up: There isn’t much to the set up. But there sure can be if you subscribe to “magic speaker cable”.  At one point I spent far too much time on the cable thing. I’d make recordings with several different cables and listen carefully. To be honest there are differences. I don’t think it makes sense to buy super cheap instrument or XLR cable. For one they do start to effect your sound. Second they go bad faster due to cheap construction. On the other hand some cables make ridiculous claims with matching price tags. Cable is one area where Socrates was right. Take the middle.

But speaker cable, cable that carries a much higher signal, is different.  So again I looked into it. Whats wrong with me? The best single source of information I found is here.

Here is my personal condensation of the above link. Super expensive cables are a racket. Speaker cable resistance as a percentage of total system resistance is what matters. You want to keep cable resistance < 5% of system resistance. In a system with a decent amp and decent contact points the system resistance is your speaker. You know, 8 ohms. So your run of cable should have less then 5% of 8 ohms. Umm … Just check out that link. They have a handy chart listing gauge of wire and length of cable.

I just happened to have a spool of 12 gauge low voltage underground wire handy from a backyard lighting project. And, consulting chart, I can use up to 50 feet before I might hear something.

I’m liking this underground low voltage wire. It’s cheap, and designed to be underground. Makes me think it’s corrosion resistant.

12 gauge  is pretty thick. The MixCube specs state “accepts up to 12-gauge wire, spade lugs, dual banana plugs.” From examining the posts I think they could fit a thicker gauge but I’m not going to try. My solution is already overkill.

Burn in: You are supposed to “break in” you MixCubes. When I first purchased the MixCubes I complained to a friend about their harshness. He asked if I had broken them in. Speakers, being mechanical, need to stretch a bit and stabilize. The result I’m told is a smoother sound. Whatever that means. I followed the instructions that came with the speakers for break in but must admit I didn’t notice much. That doesn’t mean they didn’t change. It just means I didn’t take measurements.

If you want to do the break in (which consists of hours of play at around 80dB) without going mad they do have a tip. Face the speakers close to each other and invert the phase on one. Should be much quieter. I decided not to use this trick. Maybe I’m paranoid but it seems the stretching on the inverted speaker would be a little different.

update 8/26/11
After a few weeks the Fractal is still very quiet. If I did it again I would mount all 4 standoffs with JB Weld. I estimate I could position the fan better and keep the mount more square this way.

In this post I’m going to detail a fan mod on my Fractal FX Standard. The mod makes the Fractal A LOT quieter! I’ll post before/after noise measurements, before/after cooling temperatures, detail the mod, list parts you’ll need, and give my rational for wasting your life and money.

The nuts and bolts of things are at the end of the post. Go there if you could care less about why.

Rational: As I’m going to say in another post I love my Fractal Standard. But one thing I don’t love is it’s loud fan. For a $1,100 unit (bought used) I wouldn’t expect fan noise to be an issue. And for many people it isn’t an issue. For example, you would never hear this fan over any realistic practice level. And since no mics are needed to get a great sound out of the unit fan noise is moot. But my space is a very quiet mixing room. I can hear my hard drives, my computer case fans, the faint buzz/hum of my Bryston B2 power amp into my Mixcubes. Guess what? the Fractal fan is louder then all those combined (no science was used to access the truthfulness of this statement). But seriously, the Fractal fan is loud. It’s a ball bearing 60mm fan that rotates at 3600 rpm and pushed  15.5 cfm of air. Tiny jet engine.

At first I was a little confused why the Fractal would have a fan at all. Just like my computer it has a processor with a heatsink. But unlike my computer the unit is sealed so no exchange with outside air can occur. The Fractal fan just blows air down on the heatsink and averages out the internal air temp. How much difference could this make?

I took some temperatures with an infrared thermometer. First I programmed a preset that got the Fractal running at full (96%) cpu usage. Then I let the unit warm up for about 20 minutes and started taking readings. The difference between the stock fan and no fan at all was about 26° F. Highest temp with the fan was 96° F. Highest temp with no fan was 122° F. If you decide to nerd out and take some temperatures make sure you give the unit at least 20 minutes to warm up. I think an hour is even better since it takes a while for all the components to warm up and reach a stable internal air temp.

So the stock fan does cool things quite a bit. But to me 122° F isn’t all that hot. Mind you the heatsink on my graphics card with no fan was 137° F and my hard drive was 103. My guess is the Fractal would do fine with no fan except in extreme conditions. But isn’t that what engineering is about? A factor of 10 safety margin. I’m not advocating removing your fan. For one it would void your warranty. And consequences of a melted Fractal don’t seem to outweigh the benefits.

A larger fan: My very first thought was to replace the 60mm fan with a larger 80mm fan. 60mm fans have to work very hard to push air. A larger fan could spin less to push the same amount of air. There are far more quiet 80mm fans then 60mm fans. A 80mm fan would likely be deeper putting it closer to the heatsink (you have to know the layout of the Fractal to appreciate this). I assume this would cool better.

All this sounds good but as usual things aren’t that simple. The stock fan directs a pretty strong stream of air down on an equally small heatsink. A larger fan might move the same amount of air (cfm) but not direct it as well on the heatsink.

There are lots of other things you can consider. Static pressure and the larger dead zone in the middle of a larger fan. But who cares? My infrared thermometer will tell me the reality of things.

Before/After Noise: Ever notice how fan noise stats are pretty much meaningless? There seems to be no standardized way to measure this across the industry. Manufacturers are “generous” with their ratings. And 30db of whine sure is more annoying then 30db of pink noise.

So I decided to just set up this mic and record all fans inside the unit. Everything here was held constant except the fans. Take a listen. Here is the stock fan, the Coolink SWiF2 800 80mm with and without rubber washers, and the Nexus SP802512L-03 80mm. As you can hear the rubber washers don’t make much difference. For a fan that has a rep for being quiet this Nexus sure wasn’t!

Fractal fan mod01 stock fan by cporro

Fractal fan mod02 coolink by cporro

Fractal fan mod03 coolink rubber washers by cporro

Fractal fan mod04 nexus by cporro

The big take away for me is that fan speed has a direct relationship to noise. Here the slowest fan (Coolink) is by far the quietest. It also pushed the least amount of air but I’m ok with that so long as my infrared thermometer readings are decent.

The other thing that occurred to me: This particular setup is more prone to noise then normal. You have a fan mounted to a large metal panel. Turning the panel into essentially a speaker. This type of set up would not occur in a computer. The fan would either mount to something that does not vibrate well (CPU) or mount over an opening.

Before/After cooling: After shooting the heatsink in various places I determined where it was hottest. The total variation was only about 4° F anyway. I took a lot of readings always logging the ambient temp (usually about 70° F). Here is how it boiled down.

The stock fan cooled to about 98° F. The Nexus (Nexus SP802512L-03 80mm Case Fan) to 98° F. The Coolink to 102° F. As I said earlier with no fan the unit got to 122° F.

Given my belief that the fan is just a precaution to begin with I am comfortable using the very quiet Coolink.

You will need:

• Quiet 80mm fan, 3 pin. (I used a Coolink SWiF2 800 80mm case fan) This fan is 25mm deep which is not an issue unless you use a lot of washers.
• JB Weld (they make a few different types. The original seemed best suited)
• Sand paper to prep metal for JB Weld.
• 4 standoffs (Hex Standoff, Female-Female, Aluminum, Screw Size 6-32, Hex Width 1/4 In, Length 1/2 In, Thread Length 1/2 In) grainger.com You could use your stock standoffs but the new mounts are permanent. If you go back to a 60mm fan you will need to buy some more.
• machine screws (6-32, 1 1/4″) from a hardware store. Better to get the ones where the head mounts flat not beveled. Remember the bottom of the new standoffs will have a bit of JB Weld on them so the screws should not go all the way to the bottom of the standoffs.
• Washers. You may be able to get away with not using these. I like them because they keep the screws from jacking up the fan chassis and distribute pressure evenly.
• Rubber washers. Optional. I thought these might decouple the fan better from the Fractal chassis making it quiet. In practice is seemed to shift the noise form one area of the spectrum to another. Kinda like noise cancelling headphones.
• Infrared thermometer. Only needed it you like to check your work. Also good for rating the temperature of hot chicks. Wha? Only 94°? Must be wrong.

Mod:

1. Remove the top of the Fractal. Caution: the fan is mounted to the top panel of the Fractal but it plugs into the board on the bottom of the Fractal. Don’t just pull the top panel off or you will rip the fans power wires right out of the board! After you remove the 10 tiny screws slowly lift the top panel. Then peak inside and see how things are connected.
2. Remove the stock fan.
3. Prep the top panel for JB Weld. I roughed up the areas where the new mounts will be. I also rubbed the bottom of the new standoffs. This will allow JB Weld to get a good bond.I decided to use one of the exiting mounts (The rear right one when the top panel is on the unit and you are in front of it). This places the new fan offset over the heatsink. That way the dead spot in the fans center is not centered over the heatsink.
4. Assemble the fan with new mounts. I ended up using a washer under the screw head and another between the fan and standoff.No need to crank things down here. In fact that may bend things out of square.
5. Apply JB Weld. If you are using one original standoff you only need to apply JB Weld to the other 3. I carefully dabbed it on with a chopstick. I didn’t want to use too much since it will “weld” up the bottom threads.
6. Once it’s in place make sure it’s level and doesn’t rock. Now allow to cure overnight.
7. When you put the top panel back on you may want to insure the fan power wire is out of the way and not resting on any potentially hot things.

Here are a few notes on things to watch out for.

If you use a lot of washers you might not clear the heatsink. Try to insure the new mounts are square as possible. To make up for any variance you can try rubber washers.

Congratulations. You just spent 80% of you time on .02% of the problem.

This is the 3rd truss rod adjustment post I’ve made. It’s going to be very short and sweet. After much sleuthing I think I found the perfect washer. It looks very similar to the spacers used on the Stewart McDonald Truss Rod Rescue Kit. As you may recall I had a very difficult time finding a properly sized washer. The inside and outside diameters have to be exact.

The best solution I could find at the time was a metric lock washer bent flat. You have to stack several washers to get the right height. The footprint against the wood is a bit small in surface area. The stacking isn’t perfectly aligned… You get the idea.

Because I was also looking for some computer case stand-offs I happened upon a whole other category of standoff and spacers. And from an online industrial supplier I found what I believe are perfectly sized spacers. Here are the specs and where I got them.

grainger.com

Round Spacer, Clear Hole, 18-8 Stainless Steel, Screw Size #10, Inside Dia 0.192 In, Outside Dia 3/8 In, Length 3/16 In

Whatever you use as a washer here are two notes of caution:

• You may never be able to get it back out. Make sure whatever you try out either has lots of space or you are willing to commit to it. I inserted two of these perfectly sized spacers and had a super hard time getting the second one out.
• Make sure the material does not compress. I put a very tough nylon spacer in (as seen above). It compressed changing my relief and was a total nightmare to remove.

Over at recordingreview.com I made the comment I could hear the difference between 24 bit dithered and non-dithered to 16 bit. I got called on it, “I’d like to see you pick it out double blind”. I felt the same way so I set up a semi-scientific test. Here she is with my findings.

Summery: You can hear the difference. It’s super subtle and not worth much concern. But there is no reason not to dither down to 16 bit. So why not do it?

Here are the samples. Listen for yourself. My understanding is SoundCloud plays back as MP3s but if you download you get the original file. The original is a wav at 16bit/48k.

No dither by cporro

Dither by cporro

I used a Shure KSM44 mic to record various drum hits on my acoustic kit. Recording was done at 24bit/48k. The sample I chose for the test is just a few snare hits. I exported the audio raw from Samplitude. One export was dithered using Samplitude’s triangular dither. The other was not dithered. Both exports were at 16bit/48k.

I then bought them back in the original recording session (actually the exports were set up to go directly into the Samplitude project). I toggled back and forth between dithered and undithered. The dithered sample had fuller sounding snare hits. It also sounded less harsh.

But doing a blind test inside samplitude is a pain. You need to close your eyes, toggle the samples around enough so you have no clue which one you are hearing, make your decision, open your eyes, and then log the results.

So I used Foobar with the ABX Comparator component. This allows you to do easy blind testing. And how did this go? Not so well. I wish there was more documentation on this component because I had a hard time  interpreting it’s results. For example it calculates the chance you are guessing but that stat never added up for me. I did 40 trials where I just guessed and that stat never got close to 100%.

In my trials for picking out dither with ABX Comparator I did skew towards not guessing. But it was not the concrete case I expected. I really had to listen hard and sometimes I wasn’t right.

Back to Samplitude where things get strange.  Again I am able to pick out the dither with 100% accuracy. I must have done 30 trials this way. The chances of guessing correctly 30 times is extremely low. Someone who remembers probability chime in.

Inside Samp I decided to do a null test to see what the actual difference was between the dithered and undithered versions. The noise was below -80db. I certainly could not hear that.

So what is going on? All I can figure is Samplitude is playing back the audio differently then Foobar.  If this it true it’s kind of a shame because I was really looking forward to having something like ABX Comparator for blind tests. Wah!

In the previous post I removed the truss rod plug, added washers, and adjusted the truss rod. But after tightening strings back to tuning pitch it still needed a bit of adjustment. Since the truss rod was already turning hard I decided to try another trick: Apply back-bowing pressure as I turned the truss rod.

There are some pretty cool specialized tools for this job. You can check out Stewart- Mcdonald for those. Because this is a one time deal for me on a $400 guitar I decided to go full on ghetto. This is where I look at my heap of tools and devise a good enough solution.

What I used: One custom cut length of 2″x4″, two 4″ C-clamps, a shim, 2 other small blocks of wood, and some rags. 4″ clamps barely fit the neck, 2″x4″, and block of wood. If you are buying clamps new get something wider then 4″.

Here are few notes to keep you from destroying your guitar.

• Take it easy with the clamps. You can very easily bend your neck with this setup. I’ll talk more about this later.
• Don’t clamp on top of your strings. Take them off or loosen them so they can be pushed off the fretboard. Clamping down the the strings will ruin your strings and possible your frets.
• Do not clamp directly on the guitar. This will damage the wood. See the block of wood and rag in place? That’s so the clamp doesn’t demolish my fretboard.
• I’m using a shim here for two reasons.  Shims are soft and about 1.5″ wide. I don’t want to put anything hard with a small surface area against my guitar or it might damage it. Second, shims have varying thickness which makes them easy to adjust for this particular set up.

Just wanted to throw those points out for the people long on doing and short on reading.

The basic idea is to assist the truss rod as you adjust it. You should be able to get a bit more adjustment out of it and not have that uncomfortable feeling when it turns hard.

It takes a bit of fine tuning to get this set up. For me two drum thrones worked out well. I put on the clamp near the 14th fret first. Not snug but with a little play so I could insert the shim aka fulcrum.

Then I put on the clamp near the first fret. I adjusted the clamp at the 14th fret so that it grabbed on. Not to tight. Just enough that it would stay in place. Over at the 1st fret there was about 1/4″ between the the back of the neck and the 2″x4″. That’s plenty.  Then I tightened the clamp at the 1st fret just until it started to apply pressure.

My approach was to sync up with the truss rod and not get ahead of it.  I reasoned that there should be a small clamp adjustment then a small truss rod adjustment and so on. Back bowing the neck too much before the truss rod was adjusted for the bow could strain the truss rod.

Without the strings on this is a guessing game. You can’t know for sure how much concavity the neck has until you tune back up to playing pitch. So after what I guessed to be enough I removed the clamps and tuned up. I got lucky as the adjustment was just right. But I would caution against going too far too fast. Yeah it’s a pain to remove the clamps and tune but it’s a bigger pain to ruin your neck.

FYI your truss rod only effects the first 12 frets so don’t expect it to solve problems above that. It also won’t solve problems with uneven frets. That’s another story and a costly repair at the fret mill.

After the adjustment I took a few measurements to make sure everything was good. I like to use Fender’s specs as a ballpark.

First I put a capo on the first fret and used my finger to press down at my last fret (22nd). Now check out the gap between the E-string and 7th fret. Here mine is with a .011″ feeler gauge. Fender’s spec is .010″ for this particular Strat.

Feeler gauges come in sets like this. The idea is the gauge should fit between the string and fret without pushing the string up at all.  Note: My feeler gauges have seen better days.

Ok, the gap looks good. Now I’m going to adjust the string height to near Fender specs.

It’s ok to adjust the saddle height down with strings in tune. It is not a good idea to adjust the saddle up with strings in tune. This will put stress on your saddles and strings. So I always detune the string until it’s pretty loose then adjust the saddle up.

I like to adjust my strings as low as possible without major buzzing. It’s an iterative process. Lower string, tune, pick it at every fret, bend it at every fret. Sound good? then lower it some more.

I adjusted my string saddle so that the bottom of my low E-string was about 3/64″ from the top of the fret. Fender’s specs are 4/64″. I don’t shred or have a light touch so super low action is not my goal. My other string heights were in the same ballpark.

At the end my guitar was within Fender’s specs and playable again! Ok, it probably could use a bit of fret work to fine tune it. But it’s a big leap forward from where it was the day I foolishly bought it.

Depending on how this went I was prepared with an alternate title: How to wreck your Strat in 5 easy minutes. Instead this method worked really really well.

I’m some kind of magnet for bad truss rods. I’ve had 3 Strats all with the same issue. Truss rod out of thread. When this happens you run out of adjustment and can’t correct a neck with too much concavity.

I have one Strat I’ve been playing for about 15 years. It’s one of the more deluxe models so I didn’t want to jack that one up. I took it to a respected guitar repair man and he fixed it to the tune of $700. Then I did some math. $1500 guitar I bought used for $580, put $700 into, and can probably only sell for $900. hmm.

I decided the approach to take on my “more affordable” strat would be different. I bought it used for $380. American Highway One Strat. I don’t consider myself an optimistic person but I was that day when I took a look down the neck and thought “Oh, I’ll just adjust the truss rod”. My new iron clad rule is: Never buy a guitar if you can’t adjust it there on the spot.

I did a bit of sleuthing about the process. You need to soften the glue that holds the truss rod plug in place. Then back the truss rod nut up. This forces the plug out. Remove the nut. Add washers. That way the nut does not run out of thread. Re-insert nut and see if you can adjust the neck to where it should be. Since strings will cause the neck to bow your final truss rod adjustment should take place with strings on and tuned.

I found varying ideas of how to get the truss rod plug out. I heard streamer, hair dryer, exacto knife, and a few others. The hair dryer method is supposed to be from Fender’s repair papers. My repair guy mentioned an exacto knife.

The exacto knife method seemed far fetched to me. Seemed you would butcher it trying to cut around that small radius hole. I scraped that idea right away.

I also had a hard time believing the hair dryer method would work. Here you shield the headstock with foil just exposing the truss rod plug like so.

Then heat the plug with a hair dryer until you see it start to sweat. Back the nut out pushing out the plug. I tried this and it did not work for me. Plug sweat fail. Softened glue fail.

The other method I thought about was inserting a soldering iron into the plug hole. This made a lot of sense since it would evenly heat the plug from the inside out. The heat gets applied uniformly unlike the hair dryer method. And if anything burns it will be the plug not the headstock.

Pressing on. Tape fail!

I taped the strings out the way. Don’t do this. I used painter’s tape. It’s not supposed to tear paint off or leave a residue. Not 100% true. Apparently guitar work is more exacting then painting. It left a very thin film on my fingerboard. I’m not going to sweat it that much. Will probably come off with some playing.

From left to right: Cheapo soldering iron, utility knife on top of tin foil, needle nose pliers, 1/8″ Allen wrench with ball end, wife’s hair dryer.

You can throw out the hair dryer, foil, and utility knife since they are for the hair dryer method. The rest is very useful.

Some notes:

Using a typical Allen wrench can be a real pain imo. This specialized Allen wrench can be used at a slight angle, has a very strong shank, and can be turned continuously.

This soldering iron is a low wattage model. But it loosened up the plug quite fast. I think a high wattage model may be overkill and start your plug smoking. My model came with an assortment of tips. I chose the one that best fit the opening for my plug. (Man, it’s hard to write this stuff with a straight face.)

Needle nose pliers were good for pulling the plug free. The nut will only back it out so far. They were also great for bending the lock washers and bending a paper clip into a useful tool. More below.

Here’s how things looked before.

Once the glue gets soft you want to back out the nut fast. So First I backed the nut out as much as I could. Then I inserted the soldering iron making sure it did not touch the wood of the headstock.

I didn’t want to heat things any more then I had to. I gave the soldering iron about 30 seconds then tried to back the plug out. Rinse repeat. Don’t force it. Once the glue warms up it comes out easy.

It took about 2 cycles of 30 second heatings and then … sweet success! Wow, that was refreshingly easy. And the plug hole is in relatively good shape.

Getting the nut out took a bit since it is almost the same diameter as the hole. This is where an unwound paperclip with a slight hook at the end may be helpful. Use the needle nose pliers to bend the hook.

Here is the nut.

After I got the nut out I cleaned the Allen wrench side of it with a needle and vacuum. You don’t want that gummed up when it’s time to adjust your truss rod. Then I added a very small dab of Chap Stick on the threads. I’ve heard of using oil, Vaseline, dry lubes. To me it made sense to use something that’s a solid at room temperature. That way it doesn’t wick into the wood easily.

Now comes the most time consuming part of the process. Finding washers. I walked into two large hardware stores. What an amazing assortment of washers that weren’t the right size. I guess it would kill Fender to make their guitars fit a readily available washer. Lots of people run out of truss rod adjustment. This fix is not rare.

I got washers that needed to be drilled out. I got washers that were slightly too wide. I got a nylon spacer that I wasn’t sure I could trust. Would nylon deform under stress? I didn’t want to find out.

If the washer is too wide it won’t fit into the plug hole. And if it’s center hole is too narrow it won’t fit over the trust rod’s threads. You don’t want those threads getting striped by a snug washer.

Useful fact: The truss rod is a #10-32 thread. When you go to the hardware store have your truss rod nut and a #10-32 machine screw in hand. It’s an exacting fit.

What was just the right size? A metric lock washer. I believe it was the lock washer for a M4 sized screw. I wish I wrote it down but that’s pretty close, maybe right. I didn’t want to stack 4 lock washers on top of each other. So I planned on twisting the “lock” out of them with two needle nose pliers.

If you can take some exact measurements it may be worth ordering the washers online from an automobile supply place like this.

It was pretty easy getting them flat. See.

I put 4 of these flattened lock washers down the hole for my truss rod. They fit perfect. Then I inserted the nut and screwed it back down. I don’t think I would ever put back a wooden plug. But I don’t want the hole to get filled with crude. So I inserted a bit of backer rod. It will maintain a crude-free zone and is easily removed. Not much for looks though.

I wish this was where the perfect ending occurred. I was able to adjust my truss rod better. Here is a shot of the gap at the 7th fret. There is a capo on the 1st fret and I’m pressing down on the last fret. That’s a good way to check the concavity of the neck.

This is a 1/16″ gap and way too much. After adjusting the truss rod with the washers in place I got it back to factory specs. That’s checking with a .011″ feeler gauge.

But after the strings went back on to tuning tension I lost some of that. I’m in a much better place, but still not Fender specs. Unfortunately the truss rod nut is turning very hard at this point. The next option is to apply some back bowing pressure to the neck and see if I can turn the truss rod a bit more. I only need a bit. But that’s another post (maybe)

I’m looking forward to this post. It’s less technical and more about a philosophy. I won’t talk about motherboards or logarithmic ratios. Instead I’ll record some mics and listen to the differences. I’ll give you my take. Then I’ll post some of the recordings so you can form your own opinion.

Back story. I know this very generous guy named Pete. He shall remain last-nameless so this post will not bring upon him a storm of lending requests.

Pete lent me out a set of Shure KSM32s, a set of Shure KSM44s, a Shure Beta 52, and a Rode N2. Knowing his senses may return to him any moment and thus his mics I did some test recordings. How does a $800 mic might stack up against a $200 one? Would it have less noise? Sound more open? More natural? Use less gain? Handle dynamics differently?

There is a lot to know about mics. I could spend a whole post on one mic’s features. But more and more these days I realize it comes down to how they sound. There was a time when I read ravenously on pickup patterns, dynamics, condensers and proximity effect. Now I just place and listen. Imo this tells a fuller story. Not to mention specs can be deceptive.

This is an odd bunch of mics so let me give you the once over.

The AKG D112 and Shure Beta 52 are both kick drum mics (unless you find them useful for something else). I’ll record my kick with both and compare. Here you can see what I consider the sweet spot for kick drum placement. No secret. Lots of people like this placement. Gives a solid thud and with the diaphragm pointed at the batter a good amount of click. How meaningless are those term?

Take a listen.

Akg d112 kick by cporro

Shure beta52 kick by cporro

The Shure KSM32s, KSM44s, and Rode K2 are large diaphragm condenser mics. Generally this group is sensitive and captures detail well. So I will test them as drum overheads and later with vocals with acoustic guitar.

The rest of the condensers are some of my mics. The Behringer ECM8000 is a $50 test mic I purchased for room acoustic measurements. It has a small diaphragm and a pure omni pickup pattern. Shure sm81 – small diaphragm condenser, cardiod. Rodes NT1-A – large diaphragm condenser, cardiod. AKG C2000B – medium diaphragm, cardiod.

Test Methods: If you’ve read a bit about acoustics and mics you know that many things  effect sound. Every spot in a room sounds different. Some sound extremely different (corners). Pointing a mic directly toward the source or off axis makes a difference. Proximatey to the source makes a difference.

How can you eliminate many of these variables for a better test?

There are 3 methods I considered. All of them good enough for government work.

1. Constant placement. You can mark exactly where the mic is. Record. Then swap in the next mic in almost exactly the same position.
   
   I tried to place all the diaphragms at the same position and axis. I marked the position with this hanging tape which was easy to knot up out of the way.
   
   The shortcoming of this method is your performance needs to be very consistent otherwise you could perceive mic differences from performance differences.
2. Constant performance. Here we flip flop it. Same performance but mics have to be in slightly difference positions. I tried to have all the mics pointing directly at the source. Here you can see they form an arc. I would resist the urge to put them too close to each other as this will color the sound especially with the omnis.
   This method makes it less jarring to A/B mics since you can solo back and forth between mics and the performance is smooth. But the mics are not all in the same acoustic space. On top of that some sources are directional. For example you can point all the mics at a vocalist but the vocalist’s mouth will only point at one mic.  The mic being sung to will sound brighter.
3. Constant performance and placement. Here you place the mics as in method 1. You then play a constant source such as a CD. Both are fixed. Perfect right? Not really. Your speakers are not going to reproduce a live performance. Chances are they will not produce the highs, lows, or dynamics or a real drum kit. I was interested in how the mics handled a live performanc and never tried this method.

There are also a few other things to try and hold constant. Cables, preamps, and converters can all effect the sound. Ideally they should be the same. My tests were less than idea in that respect.

I tried out both method 1 and 2. After much listening I determined method 1 was more helpful. It’s a strange thing to describe but even with different performances you can pick out the character of a mic.

Findings: For kick drum I strongly prefer the Shure Beta 52 over the AKG D112. It has more attack and thud. It just sounds more alive. I had to eq (boost near 5k, cut near 200hz) the AKG D112 to approach the same sound.

Here are the cardiod drum overheads.

Akg c2000b drums by cporro

Rode nt1a drums by cporro

Rode k2 drums by cporro

Shure ksm32 drums by cporro

Shure ksm44 drums by cporro

Shure sm81 drums by cporro

The Rodes K2 and the Shure KSM44 both have variable pickup patterns. Here they are as omnis in the same placement as above. The Behringer ECM8000 is a full time omni.

Behringer ecm8000 drums by cporro

Rode k2 omni drums by cporro

Shure ksm44 omni drums by cporro

Here are the recordings of vocals with acoustic guitar. These are all cardiod.

Akg c2000b voice by cporro

Rode nt1a voice by cporro

Rode k2 voice by cporro

Shure ksm32 voice by cporro

Shure ksm44 voice by cporro

Shure sm81 voice by cporro

And the omnis…

Behringer ecm8000 omni voice by cporro

Rode k2 onmi voice by cporro

Shure ksm44 omni voice by cporro

Omni vs cardiod. One thing to remember is my room is on the dry side. I expect this minimizes the differences of an omni since there are less room reflections. I hear the omni pattern on these mics as more open, less forward, more balanced, less claustrophobic, with a hint of the room. Listen in particular to the way the omnis sound when the vocals get loud. I think they do a more graceful job. It’s like built in compression or lack of proximately effect.

Conclusion: There are only a few things I can say with certainty. I prefer the Shure Beta 52 over the AKG D112 on kick. I marginally prefer the sound of omnis in this room. The Behringer ECM800 has way more noise then the other mics.

After that it’s tricky. I hate to repeat what I’ve read again and again. That mics are subjective. That matching the source to the mic is more important then simply having a great mic. That placement can be more important then mic choice.  But the more I listen and record the more I realize this is true.

Makes me wish I spent less time over the years reading about gear and more time listening to it! If you like that last statement check out recordingreview.com. Brandon has some pretty interesting things to say about the diminishing returns of expensive gear.

For example he has a shootout for guitar cab mics. Included is the sm57 and the highly regarded Royer R121. The differences aren’t huge. I’d be surprised if a person new to recording could pick them out at all. Plenty of people did not prefer the Royer. To my ears the Royer did sound the best. But did it sound 13 ($1,300) times better then the sm57 ($100)? Not to me.

More confused then ever? Let me throw one more at you for consideration. Think about mics in terms of mixing. What mic and placement will make the source fit best in the mix? People bag on the SM57. It can’t capture highs like a condenser. It’s not very sensitive. It doesn’t have anything close to a flat frequency response. So? As many have discovered it makes thing like snare and guitar cabs sound great in the mix.

Or how I learned to start worrying and hate the tape machine.

I’ll go through the process of adding another 8 tracks (ADAT) to an audio interface so we can record more tracks simultaneously. On the way I’ll cover some other points of interest like jitter, word clocks, and reference levels.

One interesting thing I’d like to point out upfront: Different converters/preamps can introduce different latencies even when in “sync”. More about this below.

Here is the setup. The album I’m mixing was recorded on 16 tracks of tape. The recording engineer is a real analog guy and can’t do the transfer to digital.

At the time I only had 10 analog inputs on my interface (RME Fireface 800). My first thought was to A/D convert in batches. The first batch would have tracks 1-10 the second 11-16. Can you tell I’ve been working in digital? Then I remembered something about tape from back when I owned a Tascam 424. Tape does not play back at a consistent speed. You can’t hear it pitch up and down like vibrato. It doesn’t end the song a quarter note flat. But if you try and do 2 transfers, as I suggest, good luck lining them up. This could be an even bigger problem if you have mics that are supposed to be in phase but are now in two different transfers. I confirmed this with the recording engineer who said he’d tried it. Then he said don’t try it.

Plan B. Sync my Fireface up with another analog to digital converter adding tracks. Sounds expensive for a one time use. My Fireface was around $1800. Another quality box would be in the same price range I thought. Maybe. After reading the review in Sound on Sound as well as some positive remarks on Gearslutz I bought a Behringer ADA8000. I know Behringer is not perceived as high fidelity. That doesn’t guarantee this unit sounds bad. Best of all it was around $300 which is super cheap for 8 channels of preamp and converter.

Can I hear a difference? Yes. The Fireface sounds warmer to me. Or if you want to spin it in favor or Behringer… the ADA8000 sounds cleaner. Is it better or worse then my Fireface? I can’t tell. Imo different isn’t necessarily better or worse. Take a listen and tell me what you think.

You can listen on this player or download by clicking the downward arrow on the right end of the player. If you really want to nerd out load them in your DAW so you can loop a small section and listen to each carefully.

Fireface 800 by cporro

Behringer ADA8000 by cporro

Here is how I organize the rest of this process. Somewhere at the end of this post I will go through the specifics of this particular gear.

1. Choose your master device.
2. Make data connections.
3. Make sync/clock connections.
4. Designate devices as master or slave via software or hardware.
   

Choose Your Master Device: Digital audio systems are kinda like Highlander. There can be only one! (master device) Which one should you pick? Briefly, the one with the better clock. RME has an excellent reputation for building quality products. So I’m just going to set my Fireface as the master. But what’s a clock and why does it matter?

When a device takes samples (recording) or plays them back it needs to do it at very regular time intervals. Hopefully everyone has seen that comparison between an analog waveform and a digital waveform. If you haven’t let me show you why I don’t have a career as an illustrator.

Figure 1: This is a simple analog waveform. It is continuous. Every point in time and amplitude is defined.

Figure 2: Digital audio samples this waveform at regular time intervals. Blue vertical lines show where samples are taken.

Figure 3: Now we take those sample values and reconstruct the waveform. The original wave is shown in black. The blue “steps” are the digital reconstruction. Note that the reconstruction has the same basic shape as the original analog wave.

Figure 4: But what if we don’t sample at regular time intervals?

Figure 5: Then we reconstruct the waveform with these samples taken at irregular time intervals. Jitter! Some intervals are too close together. Amplitude does not change much over that time. This produces the flat areas of distortion. When time intervals are too far apart you get big vertical drops in the waveform because a large amplitude change occurs over that time. I didn’t spend tons of time making figure 4 map exactly over to figure 5 but you get the idea.

Your clock determines the time intervals. If your clock is inconsistent you get the jitters. Er, I mean, jitter is what happens when your clock is inconsistent. Can you hear jitter? You sure can.

Which brings us back to picking a master device. It’s called the master because we will use it’s clock as a master clock. All other devices will be “slaved” to that clock. Again, all things the same, use the device with a better clock as your master.

Make Data Connections: The only way to get 8 or more tracks of digital audio into my Fireface is through ADAT.  (It also has one input for S/PDIF or AES/EBU formats. But these are 2 channel formats and I need 8 channels.) Technically it’s called ADAT optical since the connections are optical Toslink cables.

My rule of thumb is not to buy super cheap cables. Some people will argue that digital is digital and super cheap cables will work fine. But there are other considerations. I’ve read of Toslink cables with ends that won’t fit or break off. I’m not convinced super cheap fiber optics can’t effect data transmission. I also would not buy cables any longer then I needed. All signals degrade over distance.

Make sure you take the protective covers off the ends of the Toslink cable and connect the output of the slave device to the input of the master device. Since these are optical connections be careful not to muss up the ends of the cable.

Make Sync/Clock Connections: This is the connection that passes the clock signal from master to slave device. There are 2 ways for me to do this for ADAT. I can run it via an ADAT connection (Toslink optical cable) or a BNC word clock connection. BNC is a common coaxial cable connection.

I sniffed around to see if either was preferable. As I could not find a strong case for either I choose Toslink.

So just connect the ADAT output of your master device to the ADAT input of the slave device.

Designate Devices as Master or Slave: Remember you can only have on master device. Everything else must to set to slave. The ADA8000 has a switch on the pack panel. The Fireface makes the switch via software. Check the manual for your device if you are unsure.

Here are the specifics for the Fireface and ADA8000.

ADAT optical is always 24 bit and mostly 48k. You can use 96k if you really want but your bandwidth stays the same. So you only have 4 tracks at 96k, not 8.

Fireface 800:

1. Make sure it is in fact set to master. This is found in the Fireface settings software. Clock source should be set to internal. That’s RME’s term for Master.
2. The Fireface constantly looks for input clock signals. If it finds one it will automatically sync with it. If you use the Fireface as master make sure not to send any word clock signals to it.
3. Make sure ADAT is set up as an input. The Fireface allows you to limit the bandwidth on the Firewire bus if you want to. For example you can limit input to just 8 analog inputs. I set bandwidth to all channels to be sure adat is received. I’ve never had the need to limit traffic on my Firewire bus. Again this is in the Fireface settings software.
4. My DAW seems to drive my interface’s sample rate. Changing my DAW’s sample rate will change my Fireface’s control panel setting to the new sample rate. This may be a foolproof design but I would check to make sure you are at 24bit/48k.

The Fireface has a nice set of LEDs that indicate what is connected and if it is in sync.

A steady green light as seen here says ADAT 1 is connected and in sync. A blinking light would indicate signal with no sync.

Behringer ADA8000:

1. The ADA8000 does not use a software control panel. Instead you use a switch on the back of the unit. There are 4 postions: 2 master and 2 slave. In the slave postion you need to set it to ADAT as seen here.
2. There are 2 indicator LEDs on the front of this unit. You want it to look like this. Only the green locked light should be on.

We are sunked!

It would be a great idea to do a test now. This truly great idea was never followed by yours truely and causes a delightful work flow stoppage during the tape transfer. I suggest you make sure all your inputs properly route to your DAW tracks and you can record to them.

ADAT 1 did not get routed where I expected it to. 10 analog inputs and 2 S/PDIF inputs put the first ADAT at input 13.  I hadn’t considered the S/PDIF inputs. My DAW’s input weren’t labeled so it appearted I was missing 2 tracks.

I’m writing this post for people who are similar to myself except they have yet to build a DAW. If you have been building machines for years you still might learn something about building a DAW since some considerations are different from building a general use machine. There are lots of guides on DIY computer builds out there. My focus will be on what I learned from the actual build not what I learned from reading .

In my last post I talked about the process of selecting a DAW. The choices were:

1. Mac Pro
2. Large PC manufactures like Dell and HP. I’m not taking about consumer machines. I’m talking about “workstations” like the HP Z600 or the Dell T5500. Unlike consumer models these run Xeon processors.
3. Boutique DAW builders like http://www.reyniersaudio.com
4. DIY builds like the one I just finished.

My conclusion was: Buy a workstation model from a big manufacture or a Mac Pro. Before I go into the details of the DIY process I’d like to recap my reasons.

I am technology agnostic. I have no preference for Macs, PCs, or anything else. I only want my technology to perform and be reliable. Over the years I have used 2 Macs (G4s), a big manufacturer PC (Dell precision 380), a boutique DAW builder machine (ADK Intel E6600 Dual Core Processor) and now the DIY build I finished a few months ago (Intel i7 950). The Macs gave me the least amount of trouble. The boutique DAW and my current DIY machine gave me the most trouble. I’d put the Dell in the middle.

As things stand right now my next machine will be a Mac Pro running Bootcamp. In fact, a studio owner in LA told me he only builds and uses that very combo due to performance and reliability.

Here are my arguments against a DIY DAW build. (This is no way to start a long DIY DAW post.)

1. DIY will likely not save you money. You have to see the big picture: The price of the parts, your time doing research, assembling, troubleshooting, and necessary software such as Windows 7.
2. DIY will take longer to get up and running unless you know exactly what parts you need, can purchase them locally, assemble them fast, have no troubleshooting, and can set up the bios and OS fast.
3. It’s easy to build your own PC! I’ve read that many times. Perhaps it’s true. But I believe it’s not so easy to build a high performance and reliable one.
4. Dealing with several manufactures instead of one. I have at least 5 different parts manufacturers in my DIY DAW. If there are hardware problems I will need to figure out which one it is and then deal with that manufacturer. With a Mac I just deal with Apple and they do the troubleshooting. One exception here is if a part fails quickly. In this case you may be able to return to the retailer such as Newegg.
5. Boxes and clutter. My DIY DAW build left an array of boxes, packing materials, extras, and manuals in it’s wake. I don’t want to store these or organize them. With an assembled machine you get one large box.
6. Resale. In my experience name brand machines have better resale and a larger pool of buyers.

Still, for better or worse, I just had to build my own DAW. Perhaps because I consider experience to be the best form of information.

So why would you want to DIY? Some people enjoy building things. Some people like to know exactly what components are in their DAW (super quiet fans and hard drives). If you built it yourself chances are you can repair it yourself and not rely on questionable repair shops. Some people have been building computers so long they have streamlined the process.

I want to stress that I did not go into this build without some research. I read a respected book on the topic (Scott Mueller’s Upgrading and Repairing PCs 19th Edition). I read many posts online. I asked questions in technical communities. And I received the advice of seasoned DAW builders. There were still holes in my knowledge that I will cover here.

Expect to spend most of your time selecting parts. All the specs you need to know should be available online. If the retailer does not list it you can go to the manufacturers site.

Some tips:  Find out what other DIYers and boutique DAW builders are using. Also check online retailers feedback on the item. This feedback is undoubtedly bias since people with problems tend to leave feedback more then happy customers. I still find it useful. I look for well written and specific feedback rather then emotional and general feedback like “this sucks”.

Here are the parts I used and why. My cost was around $1,200 including Windows 7, two 1T Drives, and 6Gigs of RAM. I bought everything at newegg.com.

• CPU/processor: Intel i7 950.
  
  This is a low/mid priced Intel processor. I think it’s a good bang for the buck. If you plan on overclocking the lower priced processors stand a better change of large gains. The 950 also uses the higher performance LGA 1366 socket. This would allow me to upgrade my CPU to a faster i7 chip such as the Core i7-9xxX Extreme Edition. Checking the Wiki page for the i7 I see they are using a new socket for the latest i7s. Dang. The 2 big things to get right here are the performance of the processor and the socket. For performance you can see what others are using in DAWs. For example many DIYers use the i7 chip as well as some boutique DAW builders. There are benchmarks available but I wouldn’t put complete faith in them since they are in part dependant on hardware and software you might not use. Socket! The socket is the connection between the CPU and motherboard. It is actually part of the motherboard. Just like a light bulb a CPU and motherboard MUST have the same socket.
• Motherboard: GIGABYTE GA-X58A-UD3R.
  
  If the CPU is the brain then the motherboard is the central nervous system. It connects the CPU to everything else: Memory, hard drives, audio card, video card, etc. I picked this particular board because it had just what I needed with little extra. My last board was kinda pricey and included all kinds of extras I never used. The big points here are to match the socket with CPU and make sure the board will have the connections/slots you need. For example, I have an old UAD-1 PCI card. PCI is being replaced with newer PCI-e. So I had to make sure this board had a PCI card slot. I also wanted lots of SATA connections for many hard drives, some firewire ports, usb 3… you get the idea.
• Case: COOLER MASTER Centurion 590
  
  A case may be more important then you think. Motherboards come in different sizes (form factors) as do cases. So all motherboards will not fit all cases. This case is an ATX form factor. This is the most popular one and my recommendation. There is an advantage with sticking with industry standards as upgrades and future compatibility should be better. A case will play a roll in how your computer is organized. For example it will determine where fans mount, where drives mount, where power supplies mount, how much room you have to work inside the machine, the airflow for cooling, and where front panel connections (usb, firewire) are located. This case is a no frills and inexpensive model. It has plenty of venting and lots of bays for multiple hard drives. Make sure your case form factor and mother board form factor are the same!
• Ram: CORSAIR XMS3 6GB CMX6GX3M3C1600C7
  
  Do not buy cheap off-brand RAM. Figuring out the problem is RAM will take some time. Testing RAM is not fast. Then you will need to remove and replace it. So imo the “savings” of cheap ram are not worth it. Your motherboard must be compatible with your ram. There is a type of memory (ddr3, ddr2), the channel mode (dual-channel, tri-channel), and the speed (1600MTps, 1333MTps). All things the same DDR3 memory is the fastest. Channel mode has to do with how many sticks are accessed at the same time and how fast the ram is. The higher the channel mode the faster the memory. Tri-channel is faster then dual channel. Tri-channel memory is designed for 3 sticks to get accessed at the same time and therefore sold in packs of 3 sticks. If you were to buy 2 sticks of tri-channel memory they might work but not at the faster tri-channel speed. Just don’t do it. The speed is what you’d expect. Higher speeds are faster all things the same. Before you buy a ton of RAM make sure you can use it. Windows XP Home Edition can only use about 4 gigs of ram. Even if you can support 64 gigs of ram such as Windows 7 Home Premium 64 bit your applications may still be running in 32 bit mode and only use 4 gigs ram. The ram I bought is name brand ddr3 tri-channel at 1600MTps speed. It was mid-price. Unfortunately my motherboard does not officially support 1600 speed ram so that extra cost and speed may be going to waste. One way to test with would be to change my memory settings in the BIOS and do some testing. (I’ll put that on my list.) One last thing. When I built my machine it was possible to buy RAM faster then the stock motherboard speeds. So my advice is don’t go crazy with expensive high speed ram.
• Hard Dives: Seagate Barracuda 7200.12 ST31000528AS 1TB 7200 RPM SATA 3.0Gb/s
  
  For hard drives I want something quiet with a good DAW track record. Hard drives make a lot of noise. Most of the noise in my system is from hard drives. How do i know? I unplugged them all and restarted into BIOS. Very quiet. Drives are getting super cheap these day and I’ve heard on good authority they are also getting less reliable. As with ram do not buy cheap off brand drives and always back music up to a second drive. I have one drive for OS, one for music, one for samples, and one for backup. Solid state drives have been around for a while but they are still expensive and don’t offer any huge advantages for DAWs. They are good for quick random access times such as booting. But DAWs are mostly reading large continuous files so the price for performance does not make sense to me yet. They are very quiet with no moving parts. That is nice if you record near your DAW.
• Mobile Racks: iStarUSA T-5-SA. These are basically enclosures for your hard drives that allow you to remove the bare drive easily. They are damn handy if you need to swap in drives or run with data in a hurry. I like this model because it does not have a fan. Tiny enclosure fans fail often and can be noisy. This model is open and allows decent air cooling. I have another model that makes some vibration noise. So far this one is quiet.
• PSU or Power Supply Unit: COOLER MASTER Silent Pro M700
  
  Again, don’t go cheap-o off brand. When PSUs fail it might not be obvious. Failure can manifest itself in many ways. As with motherboards you need to get the right form factor, most likely ATX. PSUs are rated by how much power they provide (Watts) . There are a few online PSU calculators out there that will put you in the right ballpark. Scott Mueller’s book also has some basics wattage numbers for various devices. But wattage is not all-telling. Some quality PSUs rated at 350 watts will outperform a 600 Watt rated cheapo model. So take these numbers with a grains of salt. I bought far more then the minimum I required. Too much power is always better then just enough. Besides, I will undoubtedly add devices and increase power consumption up the road. I prefer modular PSUs. They only use the connections you need making your computer less cluttered and easier to work on.
• Video Card: SAPPHIRE 100292DDR3L Radeon HD 5450
  
  There is no need to spend your money on a high performance graphics card for a DAW. But you don’t want one with a crappy and loud fan that burns out in 9 months. The fan on my old graphics card had a protracted and noisy death. Soon after that the card failed probably due to poor cooling. Now I only buy cards with heat sinks. I have two 1280×1024 resolution monitors so I bought a single slot card that will support them. I also wanted to use digital connections so this card has one dvi and one hdmi output. I had to run a hdvi to dvi cable to my second monitor since it does not have an hdvi input.

There are a few other things to consider like an audio card and optical drive. Audio cards or interfaces are their own big topic. Just make sure you motherboard has the proper slot or connection to support it. Optical drives are so cheap all I can say is get something reliable. Mine cost about $20.

Before you buy locally or online make sure you have double checked to make sure everything will work together. My list was revised several times.

If you record quite sources near your computer consider noise. I touch on this in a few other places. Your main culprits are hard drives and fans. Remember there are fans on the heat sink and power supply unit. Here is a cool site that specializes in all things quiet.  They give detailed specs which many online retailers to not.

The next phase is assembly. This is supposed to be the easy part and I hope it is for you. I had some snags.

Things to have on hand:

• Manuals. The most handy of which is the motherboard manual. Some things are printed on the board but these can be hard to read or obstructed. I found my motherboard manual indispensable.
• Screwdriver with magnetic bits: Otherwise threading will be difficult and you won’t be able to recover dropped screws easily. Don’t worry about using this magnet by your machine. It’s safe and used by many techs.
• There will be lot of screwing. In this case it won’t be enjoyable. A small battery operated screwdriver will save you time and aggravation. One thing to consider if you go this route is the bulk of the screwdriver. Sometimes this will make fitting in tight spots difficult. But you can get extra long bits to remedy this. Be careful with an electric screwdriver. You don’t want the bit to jump off the screw and go tracking across the motherboard.
• A good light source. Some of this work is very fine (such as inserting the CPU). You want to be careful and see what you are doing. I read quite a few stories of people bending pins on the CPU socket. LED headlamps are super cheap and very useful so nerd it up.
• I found it impossible to make some front panel connections without needle nose plyers. You may or may not need some.
• Tweezers. You may need them to pull jumpers.

Stay organized. It’s pretty temping to just rip open all the new parts and start assembling. In fact, that’s what I did. Bad idea. If you ever need to return something in original packaging good luck. Aside from that all the little extras will get mixed together and you’ll wonder where that one screw went. I suggest you open one part at a time keeping all extras and packaging organized.

Order of assembly. There is definitely a wrong way to put things together. For example if you install the hard drives first you will probably need to pull them back out to get the motherboard in. The assembly steps below should keep your work and aggravation to a minimum.

One assembly tip is to mount the motherboard on a motherboard tray and install RAM, CPU, and heatsink outside of the case. Having it mounted on a a tray is very important because installing RAM and the heatsink require some force. If you don’t mount to a  tray you could damage your board. I recommend the tray be on a firm flat surface and not your couch. Installing outside the case gives your hands more room to work. You will have more control while applying force and have more delicacy while installing the CPU.

Unfortunately, I did not have a motherboard tray so my motherboard was installed in the case. Wah.

static discharge: In 2 years I’ve noticed static discharges only one time in my studio. So I am simply touching my hand to the case to discharge any static and handling parts by their chassis not components. If you have a high static environment (or you aren’t brave like me) you may want to consider static discharging devices: The anti-static bracelet with tether and an anti-static mat for laying out parts.

Don’t power up anything or even connect a power cord until the machine is fully assembled.

Motherboard Assembly: before the motherboard goes in we need to prep the case. The idea is to make the case as accessible as possible. Remove the side and front panels. If your case shipped with a power supply installed remove that. Btw, most power supplies that come with cases are low quality. You probably want to replace that anyway. If you plan on replacing the fans that came with the case remove them now. Fans that ship with cases also tend to be of low quality. I’ll talk about what to look for in fans latter.

Now we’ll mount the motherboard on a tray outside the case or mount it in the case.

Motherboards do not mount directly. They mount on stand offs. Here is a picture of some stand offs inside my case. Standoffs create about a 1/4″ space between the board and tray. Without this space your board could short out. Some standoffs screw in like these. Others are like nubs built into the case.

My board has 9 holes for mounting it on the stand offs. It is possible that your case can take more then 9 stand offs. These are for extra support. Just make sure your board is designed to take the extra stand offs.

If you are mounting the motherboard on a tray outside the case you can skip the next 3 paragraphs. After you install the CPU, RAM, and heat sink you can return to them.

It should be pretty obvious how the motherboard fits in the case. The mounting holes in the mother board line up with the stand offs. The back panel connections on the mother board will line up with the i/o shield on the back of your case. This is where most of your connections are like USB, lan, PSU, Firewire, etc.

Before you place the board into the case make sure the back connections of your board will fit the i/o shield on your case. If it doesn’t you’ll need to remove the shield and replace it with the one that came with your motherboard.

I think it’s best to get the i/o shield fitting first and then line up the mounting holes. Make sure none of the little metal grounding tabs on the i/o shield go into the ports on the motherboard.

With everything lined up insert the mounting screws. Now the board is well supported and ready to install RAM, CPU, and heatsink.

If your motheboard is in the case you might want to make a few difficult connections right now. Later, when things are cluttered, it will be harder.

front panel connections:
These are the connections at the front of your case for USB, Firewire, reset, power LED and a few others. On my case they are attached to the chassis and can’t be removed with the front panel. Some of these connections are very easy to make. My USB, for example, is just a standard USB header on the motehrboard. But the connections for power LED, reset, speaker, chassis intrusion, and system status are very tiny and hard to make. In this picture you can see that many of them are single wires going into a header on the board. I had to use a needle nose plyers and some patience to get them in. Imagine how much more difficult that would be later when the case has more cables and drives.

(Note: If your front panel connections are attached to your case’s front panel you may not be able to make these connections since the front panel may need to be removed to install drives.)

With my particular case and motherboard I needed to do a bit of sleuthing for these connections. When I found the correct motherbaord header in the manual the labeling did not match up exactly with the labeling on the case’s connection pins. This would be so much easier if the connections were bundled together in a foolproof inset like so many other connections!

I have heard you can order a header insert to make the job easier. This allows you to plug the small connection tips into the insert and then slip it into the header. but I was unable to find one that would work for my board. My old board came with one. Here is a company worth looking at. You can also check with your motherboard manufacture.

CPU Installation:
This is a somewhat delicate job. Let’s do it while the case is uncluttered. The CPU connects to the motherboard via an array of pins that press into a grid. This particular socket (LGA 1366) has 1366 pins. You can bend them easily.  The only thing that should touch them is the bottom of your CPU during installation. Bent pins can destroy your board and void your warantee. Be careful.

The CPU should be held by the edges. You don’t want to touch any contact points where it connects to the socket.

Both the CPU and the socket have a protective plastic cover which will need to be removed. The plastic cover on the motherboard socket is held down by a load plate. The plate is held down by a load lever.

Push down on the lever and pull it out a bit so that it clears the clip. This will release the the tension on the plate. I’ll try to be clear but you may want to check out some videos on You Tube for the nuances. Here’s one.

1. Push down on the load lever and pull it slightly away from the socket to release it from the clip. Swing the lever up and out of the way.
2. If the load plate does not come up with the load lever you can push down on an edge and it will come up. Here you can see the load lever is released and the load plate is up.
   
3. Remove the protective plastic cover on the socket. You should see an array of pins (some sockets have an array of contact points).
   
4. Remove the plastic cover from the CPU. Remember to hold the CPU by the edges.
5. Get the CPU oriented correctly for the socket. There are notches on the CPU that will match the socket. Once you have it oriented the right way drop it in the socket. You may need to push it a bit until it falls and seats.
   
6. Put the load plate back on
7. Press the load lever down and lock it back under the clip.
   

Heatsink Installation:
Processors generate a lot of heat and unless they are cooled they will overheat. Next we will install a heat sink/fan. This operates much the same as a radiator on a car. Heat is transferred to a thermally conductive metal with lot of surface area. Then a fan is used to dissipate the hot air. Essentially the heat flows out of the CPU, across a thermal grease or phase change material, into the heat sink, and then is blown into the air.

What is thermal grease or phase change material?  Even though the heat sink will be clamped down on top of the processor we still need a thin layer of material between them to insure the best heat flow. This is not optional.

If you are installing the stock heatsink that came with your CPU you will probably already have the thermal conducting material (phase change material) pre-applied to the heatsink.  It will look like a few gray strips on the bottom of the heatsink. It will have a protective cover or film on it which needs to come off.  If you purchased a heatsink separate from you CPU you will need to apply some thermal grease.

It’s  worth checking out some You Tube videos on apply thermal greese to get the nuances. Here’s one.

Basically, you take a small amount and spread it out evenly over the heat spreader (top of the CPU). Some people use a card to do the spreading. I think it’s easier using a finger. You don’t want thermal grease on your finger or finger in your thermal grease. Wear a glove or put a plastic bag over your finger. Don’t spend lots of time deciding on what brand of thermal grease to buy. They all work about the same.

1. Remove the protective film/cover from the heatsink’s thermal material.
2. Orient the heat sink. The fan on the heat sink needs to plug into the CPU fan header on the motherboard. Make sure your current orientation of the heatsink allows you to reach the CPU fan header. Make sure the wires from the heatsink’s fan do not obstruct the fan. When I powered up my machine for the first time a red LED on the board lit up indicating my CPU was too hot. The power wire from the fan was ever so slightly in the fan’s path. It stopped the fan and the CPU did not get proper cooling. (Modern CPUs do not burn up when they overheat. They throttle themselves down. I didn’t want to test this out.)
   
3. My stock heatsink as 4 retainer pins which are used to hold the heatsink firmly to the CPU. The arrows on the pins are a bit confusing. To install you need to turn the pins opposite of the arrows. The arrows indicate the correct direction for removal not install! So turn them opposite of the arrows and set the pins into the holes on the motherboard. Now you need to a apply a decent amount of pressure downward to lock the pins into place. I like to put a thumb on opposite corners to apply even pressure. You will hear a click when they are locked. Now do the other corners.
4. Connect the fans power wire to the CPU fan header on the motherboard. All fan headers are not the same. Make sure you use the fan header for the CPU fan.

Be aware if you are installing a non-stock heat sink the mounting may be quite different since the heatsink may have to fit many different sized sockets. Check out the heatsink’s manual for instructions. Before you buy it make sure it fits your socket.

Ram Installation:
Make sure you buy the right RAM. Check your motherboard manual for specs. RAM manufacturer’s websites may also have a tool for matching RAM with motherboard. Your motherboard must support the type of RAM (mostly ddr3 these days), the channel mode (dual, tri-channel), and the speed (1333MTps, 1600MTps, etc).

Installing RAM is not as simple as buying RAM and pushing it down into a slot.

• My motherboard manual as well as many other people recommend installing all the same memory. Same brand, speed, chip, model.
• The memory slots on your motherboard are designed to be filled in a certain order. Consult your manual. For my motherboard and tri-channel memory it’s pretty straight forward. The first 3 sticks go into slots 1, 3, and 5. These slots are not next to each other but they are all white. So that’s pretty simple. The next 3 sticks go into slots 2, 4, and 6 which are all blue. But things can get tricky if, for example, you decide to install dual channel memory.
• Don’t mix and match sets. Even if you get the same brand and model you still should not mix them up. For example if you are buying tri-channel memory the first pack (3 sticks) should go in the first three ram slots as specified by your manual. The next pack of ram (3 sticks) should go into the next 3 slots as specified by your manual. In other words don’t shuffle 2 packs of ram even if they are exactly the same. If you do mix them up you can sort them back out by looking at the serial numbers or bar codes printed on them

To install:

1. At the ends of the ram slots are clips. These need to be open. So push them away from the slot.
   
2. Ram is a bit like the CPU. It has notches and will only fit one way. Align the notches in the ram with the notches on the slot.
3. Applying even pressure across the stick press the ram down into the slot. This takes more force then you’d expect. When it is properly seated it should click into place and the clips will close.
4. After all the RAM is seated double check that all the clips are back in locked position and the all the RAM appears seated the same.
   

Now you should have the RAM, CPU, and heatsink installed on your motherboard. If you’re installing on the motherboard tray outside of the case you need to put the board in the case now (covered above).

Let’s install drives, expansion cards, and the power supply. I am going to install power supply first, then drives, then card. This keeps the center of the case uncluttered for as long as possible. That way we have more room to work.

Power Supply Installation: Depending on your case your power supply can mount in a few different places. Mine mounts on the bottom of my case. One thing that troubled me is the manual did not specify if the fan should be mounted to draw air into the case or push it out. I could mount either way. The manual was about as detailed as Ikea instructions and I couldn’t find anything online. Perhaps it does not matter, in which case I’d still like to hear that from the manufacturer. In the end I mounted it with the fan pushing air out of the case via a grill on the bottom of the case. I reasoned that puling air in from essentially the floor of my studio would introduce more dust.

My PSU had 4 screws for mounting to the case and a rubber gasket. I assume the rubber gasket is for a better mount. It should also make for less noise due to vibration. I fit the gasket between the PSU and the case. Then lightly snugged the mounting screws.

Most motherboards will have a 24 pin connection that powers the board (and devices connected to it) and a 4 or 8 pin connection that powers your CPU. Higher performance CPUs will take an extra 4 pins of power. Check your manual for where these are located and plug them in. These connections have a fool proof design and can only be plugin in one way. So look at the headers and the ends of the cables and match them up. You will notice not all the connection pins have the same shape.

You will also need to connect power to your hard drives, optical drive, and perhaps some other devices. This is one reason I like modular power supplies. They allow you to just use the cables you need. This power supply came with an assortment of cables. Pick out the ones you need, connect them to the PSU, and then to the device it powers. My old PSU had a heap of wires hanging off it. Some I needed. Many were just in the way.

Hard Drive Installation: While you have the drive out you should check and see if it has a limiting jumper on the back of it. One of my drives had one. It was the tiniest jumper I’ve ever seen and I couldn’t imagine removing it once the drive was installed. Jumpers make physical connections and change the behavior of components. This particular jumper limited the drive to a slower transfer speed (1.5 GB/s instead of the drives rated 3GB/s). You may need tweezers to pull a jumper out.

Now we need to mount the drive in the case. Drives generate heat and may need cooling but I don’t sweat that much. At one point I was reading white papers on drive failure. My rule of thumb is your drive needs a fan if it’s too hot to keep your finger on it. I have 3 drives in a 4-in-3 device module with a fan and 2 more drives in mobile racks with no fans. All of them are very touchable.

I dig mobile racks. You can remove the bare drive super fast, swap in a new one, or run with your data.

Mounting my drives inside the 4-in-3 device module was a slow and tedious exercise in threading screws and aligning. You can’t remove drives easily. You have to pull the whole unit which means unplugging the fan and all the other drives. Should be called a 3-to-4-times-more-work module.

My case came with tool free mounts. I don’t like them. They are fast but they don’t hold the drives tightly. I’m suspicious this could cause vibrational noise.

Mounting the drives, mobile racks, 4-in-3 Device Module, etc is pretty straight forward. Some cases may have spaces where a 3.5″ drive mounts directly. In other cases you mount the drive inside a device and then mount the device inside the case. Once mounted we need to make some connections.

Chances are you will be installing SATA drives. They need 2 connection: Power from the PSU and a SATA connection to the motherboard. The power supply connection has 5 pins. The SATA connections to the board are a bit trickier. For example, my board has ports for several different SATA standards. Usually a faster standard will support a slower one. For example I could plug a slower (1.5GB/s) device into a higher speed port (6GB/s).  But the opposite will not work. make sure your port supports the speed of your drive. Also, if you are planning on setting up a RAID you need to plug into the ports that support that.

I have noticed that some of my SATA cables have a release clip on them and some don’t. A few times I inserted a SATA cable without a release clip on it and found it very difficult to remove later. Perhaps the header was designed for a cable with a release clip. I don’t know. Now I only use the cables with a release.

Here is a picture of how I worked my SATA cables out of the way. These run into the space between the rack for hard drives and the side panel. There is about 1″ of space there. It keeps things much cleaner.

As I said earlier I still don’t think solid state drives make sense at their current prices. Yes they are dead quiet. Yes they are fast at random access tasks such as boot up. But for audio I never have any issues running 50 48k/24bit tracks off my Barracuda drives. For me the cost of SSDs outwieght the benefits.

Expansion Card Installation (Video and Audio Cards): Make sure your motherboard supports the type of card you are inserting. Cards need to be seated in a slot. They also need to have their bracket mounted on the rear of the case where the port is. Here is the rear of my case. I opened the bracket space by removing the bracket that came installed with my case. You can see the blue plastic tool free mount or clip is also released.

I like to insert the card into the bracket a bit then seat it in the slot. As with RAM try to apply even pressure as you seat it. The slot may have a clip that clicks when the card it seated. Once seated we can secure the bracket. My case has tool free brackets. I don’t mind using them for cards since they don’t vibrate like hard drives. Otherwise you will have to use a screw to secure the bracket. Done.

Fans: Most people are going to want a quiet DAW. That means quiet fans and few of them. A large fan (120mm) can rotate less and push the same amount of air as a smaller fan (80mm). All things equal a large fan will be quieter pushing the same amount of air.

Note: Fans tend to make a white-ish noise which I find much more listenable then the tonal noise of hard drives. When I was recording very quiet things near my DAW it was tonal noise from the drives and a few other components that was a problem.

Another spec to consider is the bearing type. Cheap fans tend to have cheap bearings that fail sooner. As they approach failing they will get louder. Sleeve bearings don’t last as long, and are louder at high rpms. Dual ball bearings last longer and are quieter generally.

I see no reason to run more then 2 case fans. One drawing air into the case. One pushing it out. Remember you also have a heatsink fan and a power supply fan bringing the total to 4 fans. That noise can add up. If you record quiet sources near your machine you should consider that.

I run my case with both side panels and the front panel off. I’m not sure I even need to run my 2 case fans. Some people will argue the case will take on more dust if it is open. Maybe. Maybe not. Your intake fan is going to draw in some dust anyway. Your computer will definitely be a bit louder with the panels off. Even more so if the panels have an acoustic treatment. For me it makes more sense to be able to access my computer easily, see what LEDs on the board light up, and touch components to see if they are hot. Just my personal preference. I also stuck with the stock and cheap fans that came with my case. Replacements may follow.

If you really want to know if your case fans are necessary or if running the machine with panels off makes a difference you can run a CPU stress test for both options. This will tell you which scenario provides better cooling. More on diagnostics at the end of this build.

Fan installation is pretty easy. Make sure your case can mount your fan size. Fans are sized by diameter in millimeters. Common sizes are 80, 92 and 120 mm. Try and plug them into a fan header on the motherboard so that the wires are out of the way. There should be an arrow on the fan indicating the direction of air flow. A common way to set up fans is an intake on the front of the case and an output on the rear.

Powering Up: It would be nice if everything ran perfect the first time. In case it doesn’t lets leave the case as open as possible. My side panels and front panel have never been on my machine. With them off I can see all the LEDs on my board. These LEDs can tell you some useful stuff like CPU temperature. Check your manual for what LEDs you have and what colors indicate trouble. With the case open you can feel if something is too hot and check that fans are running.

Plug in a mouse, a keyboard, and a monitor.

Connect the power cord to the power supply. Flip the power switch on the power supply. Finally hit the computer’s power button.

Installing drivers and OS: There are a few more things you need to do before you are ready to install your DAW software.

Drivers that came with your motherboard: Back when I ran windows XP you needed to install these from the motherboard DVD or find them online. Finding them online was a terrific pain because the motherboard manufacture did not have the latest drivers posted on their site. So you had to download them from the individual vender’s sites.  But Windows 7 installs with many drivers so you may not need to install any of the ones that came with your motherboard. Not to mention your motherboard’s drivers will likely be old. The approach I take is to install Windows 7 and then see what devices are unrecognized in the device manager. If you need to install others find them online at the manufacturer’s website. Now, if Windows 7 does not install a working LAN driver you will not be able to get online. In this case you may need to install the LAN driver from the motherboard.

Utilities that came with your motherboard: I don’t install any of these. They aren’t essential and have caused me problems. IMO the less installed the better.

So, knowing that, you may want to skip any software or drivers that came with your motherboard and go straight to installing Windows 7.

Troubles: Here I will summarize the issues I had with this specific build and how I solved them.

Bad RAM: It took me a few days to figure this one out. When RAM fails you don’t exactly get a message “bad RAM, please replace”. This goes back to my original idea about buying a pre-build machine from a big manufacturer. When those machines are assembled they are tested. Anyway, I got several cryptic BSOD (blue screen of death) errors. Looking them up on Microsoft’s site was not much help either. But I did find that one of the errors could be from bad RAM. I tested the RAM using memtest86. This test takes a few hours. It passed. After a day or so I realized I was testing the memory at my system’s default speed (1099Mhz) and not it’s rated speed (1600Mhz). At the rated speed one set (3 sticks) quickly failed and went back for exchange.

BIOS freeze: To change the RAM speed I needed to get into a screen of the BIOS called intelligent tweaker. It’s the overclocking page of this BIOS. Unfortunately this board will freeze up on that screen unless your keyboard is connected via an old PS2 connection. So several times I had to power down and restart. I had no idea what was going on until I found a few others with the same issue. I hoped the newest BIOS would solve this. It didn’t. So I had to run out and buy a $3 USB to PS2 adapter.

Testing and Diagnostic: You don’t need to use any of these programs. But after building your DAW you might want to test it a bit and see if it melts.

• Memtest86: This free program allows you to test your RAM. I ran it from the Ultimate boot CD.
• Ultimate Boot CD: This is a free compilation of helpful software on a bootable CD. They allow you to download an ISO image which you burn to CD. Being able to boot from the CD is important since your machine might not be booting from it’s OS drive.
• Sandra lite: The free version of this program allows you to get lots of information on your system and run some benchmarks. I used it to benchmark my drives. I wanted to see their actual transfer rate and compare some of the older drives to the newer ones. I also bench marked my CPU.
• Real Temp: Allows you to check the temperature of Intel CPU cores.
• Intel Burn Test: Stress test for CPU. Good for checking cooling. I found Intel Burn Test more stressful on my system then Prime 95. Anyone know a direct link for this? I hate putting up links from software aggregators.
• prime 95: Another stress test for CPU. This is on the Ultimate Boot CD.
• CPU-Z: A good program for getting information on your hardware such as motherboard model, chipset, socket, specific CPU.

Alright. Done! Now, was all that worth it?

I’ve waited some time to post. That’s because I didn’t want to write anything until my research was finished and my opinion fixed.

I had a 4 year run with my old DAW (Digital Audio Workstation) but it’s time to move on. Sure there are workarounds. I could freeze tracks, maybe upgrade the CPU, but basically it doesn’t have enough juice to keep up with my ideas.

I’m the type of person who likes to know all my options and spend some time considering them. At this stage of recording it makes sense to operate this way. Ten years ago I knew little and was happy with simply buying a popular machine and using popular DAW software. That approach was fast but had it’s pitfalls. Like when I found a out a month later some software I really liked wasn’t developed for my OS, or that I didn’t have any PCI slots for my old DSP card, or that I payed a lot of money for a graphics card I would never put to use. Ooops.

Here is the disclaimer I’m going to use for the next few posts that detail my daw selection and building process: I’m not an expert when it comes to building or selecting machines. Even so, I think I can add some useful information because I’m approaching it from a newbie-ish perspective. People that have been building machines for years seem to have forgotten all the knowledge they have acquired along the way.

Here is an example. I’m my research for a DIY DAW I received a recommendation to use 3 sticks of mid-grade ram. I believe what the poster was trying to say was “Don’t pay for expensive fast ram. Ram speed is not the issue for a DAW. Also don’t get super cheap ram. The problems you can have with cheap ram aren’t worth the savings”. Here is what he assumed I already knew: My motherboard is designed with DDR3 ram in mind. DDR3 ram can only run at top speed if you install sets of three. You should already have checked with the motherboard or ram manufacture for compatibility or downloaded their qualified vendor list (QVL). If you plan on overclocking you will need ram that is stable above your default memory bus speed.

Not to fault this guy. He was trying to help. He just forgot all the things he actually did know or learned years ago building his first machine. What I hope to do is fill this knowledge gap.

All the information I’m going to pass on is from researching online, reading Scott Mueller’s 19th edition Upgrading and Repairing PCs, reading the documentation that came with my parts, and building my DAW.

Options:

The way I see it you can either buy a pre-built machine, have a daw builder build one for you, or build it yourself. Over the last 10 years I have tried all three approaches. That’s not scientific, but it’s good enough for an informed opinion.

A few years ago I would say consider your OS. Now that Macs have Intel chips and can run Boot Camp I’m not so sure. I talked to one studio owner and DAW builder who only uses Mac Pros and runs Boot Camp. He says these machines are fast, stable, and have no virus issues. So it would seem any software you can run on a PC you can now run on a mac.

If you run audio software developed only for Macs I think you have to stick with a Mac. To my knowledge there is no Boot Camp equivalent for the PC. There may be ways of simulating a Mac on a PC but I’d be careful. If I went this route I’d like to know what support is like, how much processing power you loose, if it’s stable, and if anyone doing serious audio is actually running it.

I am not going to consider laptops. IMO a desktop will always beat a laptop for a DAW. Laptops have slower drives, no room for your DSP cards, cost more for the desktop equivalent, etc. The only advantage I see in a laptop is mobility. I don’t need that.

Here we go.

Pre-built Machine: If you get a pre-built machine from one of the big manufactures make sure it is a business or enterprise machine (HP and Dell list these as workstations). These machines are designed to do heavy workloads and be dependable. My rule of thumb is most consumer grade goods stink.

I looked at 3 manufactures: Apple (Mac Pro), HP (Z series) and Dell (Precision series). All of them use Xeon processors. The Xeon is basically the enterprise/server version of Intel’s i7 chip. I doubt there is much difference in DAW performance using a Xeon vs an i7. However, Xeons do cost more then i7s. Take note potential DIYers.

The Mac starts at $2,500. The Dell and HP start at less. But if you configure them close to the Mac specs they are over $2,000. Don’t assume Macs cost more. At some price points they beat HP and Dell with similar specs. Go to the respective sites and configure the machines you want and compare prices.

When I had my Dell Precision series machine it performed well for the price, was very quiet, and the only thing that ever failed was the optical drive. They sent a “tech” who just swapped in a new drive. I could have had the part mailed to me and done it myself. I wish I had because the goon with a screwdriver they sent scared the hell out of me. It was apparent he wasn’t that skilled and all I could think about was him jabbing my hard drive (not backed up of course) with that screw driver and loosing it.

The Dell was both quieter and had less hardware failure then my next machine which was a custom DAW build. huh?

Daw Builders:

The idea behind daw builders is they select components with audio in mind and tweak your system for audio. For hardware: quiet fans and hard drives, chipsets and processors that test well with audio applications, inexpensive video cards (we don’t need the added heat or cost of gaming video cards). For software: optimized bios and OS settings. Some builders will preload your applications too. When the machine arrives you just need to enter your license information. Cool.

About 4 years ago I bought one of these machines. Although I found the owner of the company to be a good guy who was generous with his time I just can’t give them a very good grade. It’s just one machine so maybe it was a fluke, but that machine had more problems then any other machine I’ve owned. I’m not going to go into details here. But if you do buy a boutique DAW ask a few questions. If something goes wrong who will work on the machine? Do you ship it back and loose it for a week? Are You responsible for finding a qualified person to do the work locally? How long do you warranty the hardware? How long is your tech support free?

DIY DAW build:

DIY DAW building is not necessarily simple or fast. There are people who have been building machines for years. It may be simple for them. There are also people who happen to get lucky. They pick the right parts, don’t get any lemons, and things move right along. I am in neither category.

These days I’ve read you can’t really save money by building your own machine. When you take into account your time, the cost of parts, buying critical software ($200 for Windows 7) you may be paying more.

The advantage is you can select exactly the parts you want and know your system very well. If something goes wrong you will probably be more qualified to fix it then the guy at the local computer shop. After my build I certainly feel more qualified then the tech-thugs at the local repair place. They managed to bang up my case replacing a hard drive. Then the tech working on my machine almost dropped a hard drive 5 feet onto concrete. In the last 4 inches he made a miraculous save. I made a decision to not return.

Another advantage is you will or should be using industry standard parts. When you buy from a big company you may end up with proprietary components. This could make future upgrades or replacements a pain or impossible.

If you are inclined to go this route but don’t know where to start please hold on. In my next post or two I will be detailing my DIY DAW build in numbing detail. I’ll include the thinking that went into component selection, the missteps I made, reviews of specific parts, and (I hope) some easy to grasp concepts.

My conclusion? Option 1. Buy a business or enterprise model machine from a big manufacturer. A big manufacturer has the resources to thoroughly test these systems for reliability. If something goes wrong they will most likely send a tech to swap in a new part. You are up and running fast and can get back to making music instead of installing ram and tweaking the bios.