Multitrack

There are plenty of USB 1.0 audio interfaces out on the market, and some of them tout multitrack recording functionality. Be wary however of any USB 1.0 sound interfaces as a digital multitrack recorder as the interface does not have the juice to cater for anything more than 2-in/2-out.

Your numbers certainly get tight if you’re wanting to experiment with 24-bit/96khz audio. Most interfaces will not do 2-in/2-out at this quality level in full duplex. Full duplex refers to the cards ability to record audio and playback audio at the same time. In general, you would want this for multitrack audio, even if it was just for monitoring the performance while recording.

A better option would be to go for either PCI, PCMCIA/Cardbus, FireWire or USB 2.0. All of these above interfaces are best suited to multitrack audio conditions.

If you’re just starting out recording your own music at home, or for the band, getting set up is fairly painless these days. Your best bet is a computer, either Mac or PC, in addition to some other basic recording equipment. The other additions I’ll cover in a later article, but today, I want to talk a little about the computer.

Now, there are many truths and myths around the Internet regarding what kind of computer you need. Here are some simple pointers:

You Don’t need the Latest and Greatest

Although some people would like to tell you otherwise, you do not need the latest and greates PC or Mac to record some simple multitrack audio. In PC terms, something as old as an Intel Celeron 1.8Ghz CPU will get you a lot of mileage when recording.

Although not the Fastest, get the Best

What does that mean? Well, although there is equipment around that is fast and fantastic, it’s better to get a PC where the components are well designed and well made. For example, having a cheap quality power supply and motherboard in particular can cause siginificant problems down the line. So, don’t worry about the fastest, but get something good!

If you need further help with choosing parts and specs, post any questions in the comments feature below.

Go with what You are Comfortable with

Music is an art. That means that you want something that will help you create music, not get in the way. From start to finish, aim to choose your components, from the hardware to the software according to what you are comfortable with. Download some demo software and have a play around, find which is more suitable to your style and preferences.

These tips should get you on your way in finding the right computer for your home recording experiments. Please ask any questions below if anything is unclear.

This post can originally be found at another one of my sites: Binary Muse. But, since the information would be very helpful to most readers of this site, I’ve posted it here also. Enjoy!
A question came up on a forum that I was loitering at:

How would the quality differ from a Maya (USB 4-channel audio card) to an M-Audio (PCI Delta 44 audio card)?

I thought, good question, and I pecked out this reply. I worked on it so hard I thought it’d be good to put this here also in case anyone else finds it interesting.

The difference is mainly in the quality and design of the board, the quality of the preamps (if they have any), and the quality of the converters (digital to analogue and vice versa).

As far as simple technical specs go, you’ll also find differences in bit-rates (16-bit vs 24-bit) and resolution (44khz vs 48khz vs 96khz), and also noise floor ratings (noise introduced from the components only) and “headroom” (the amount of signal level you can pump in before distortion occurs) which I think is also referred to as dynamic range.

I’ve got a pretty good ear and can definitely tell the difference between a cheapy multichannel card and a good one. You can get it down to technical specs, and those mentioned above are the first step.

Digital is fairly unforgiving on audio sources, so you can really tell the difference with a cheapy card with cheap converters as compared to a well designed card with decent quality converters.

For a subjective comparison, you will notice that a better quality card will give you a “sweeter” sound, sounds will be reproduced more accurately, and you will be able to discern more detail in a better quality conversion.

Some ways to compare (if you’re an audio freak, or interested in becoming one), although not too technical but more subjective, (if you have the ability to have them side by side) is this:

• Get a hold of a good quality piece of audio software (I like Ardour in Linux)
• Don’t connect any source to your sound card, but enable monitoring on the inputs. Watch the meter, does it record a signal? How much of a signal? This is the noise floor.
• Another way to check noise is to record nothing for about 30 seconds and then once recorded, zoom in on the waveform. See any signal? If you have a CRO, hook it up to the inputs and see how well a sine wave can be recorded.
• Record an instrument with complex sounds (acoustic guitar is a good one) and listen to the original instrument while recording. Then play it back, how accurate is the recording as compared to the real thing? Can you hear the dynamics of the instrument? ** NOTE: this is a difficult test as it also depends on mic quality, preamps outside the sound card and the speakers you use to playback. BUT, if you use the same setup in each test and just change the sound card you’ll hear a difference.

I’m sure you could come up with a similar experiment to test headroom, although my brain hurts after typing all this, so I’ll leave that for someone else.

If you really want to compare how crappy sound cards sound, get out a laptop and run some of these simple tests. Most laptops have terrible quality audio cards, it should almost be a crime, although I understand the design constraints and interference from other componenets in a laptop chassis. The worst I’ve ever heard was a Sony VAIO FX770, if anyone has one of those, give it a whirl.

I just remembered, there are other things to consider also:

• Some cards don’t work with other OSs like Linux and Mac OS X.
• Some cards perform poorly speed wise, like the ability to run buffers at 8ms, etc.
• Some cards have crappy drivers (slow, buggy, consume too much CPU, poorly supported)
• Bandwidth capabilities on the interface chosen (USB 1 cannot transfer many channels of audio due to bandwidth constraints)
• USB cards tend to use a lot of CPU (design of the chipset)

As far as the comparison of the M-Audio and the Maya on paper:

Sample Rate: M-Audio: 96khz Maya: 48khz

Bit Rate: M-Audio: 24-bit Maya: 16-bit

Dynamic Range (A/D): M-Audio: 99dB Maya: 85dB

Dynamic Range (D/A): M-Audio: 103dB Maya: 87dB

Just on paper it looks like a BIG difference (also note that dB are logarithmic, meaning that a 3dB increase is the equivalent of 2x - I think maybe an audio geek can help out here, but I believe that’s right)

I hope someone has some fun with this. If anyone can comment or add anything feel free!

The other day I realised something important, I enjoy digital multitrack recording and everything about it.  The realisation came when I stumbled across a forum where some people were asking questions about multitrack sound cards.  I was able to pass on some advice to the people in the forum and realised that maybe I should place some of my writings on the web somewhere more accessible.

So, here we are.  Hopefully this site will become an informative resource for digital multitrack audio recording covering the recorders themselves, sound cards, software and all the other hints and tools that go along with it.

Enjoy.