Friday 13 August 2010

Recording, plugged in

Sorry for so many posts in so few days, but I've got all this stuff that is bursting out and just can't wait!  The problem is that when I first started this blog I knew I was writing to no-one, whereas now I know there are more than three people listening - I feel as I have to approach blogging with a bit more restraint and respect!

Recorded sound from a guitar needs to be good.  Note that I didn't say great.  Great is great, but great can begin to mess with your life - visiting a recording studio or dedicating a room in your house as a studio, filling it with sound absorbing material, buying fancy microphones...great sound starts to take over from the actual playing.

Recording via a webcam onto youtube isn't good.  So many guitarists, posting videos, wasted, because it sounds, well; crap.

So in my continuing efforts to simplify recording good sound I recently tried out USB guitar cables.  I tried two, both are cheap Chinese units, one is a direct rip-off of a Behringer product, the other was essentially a modified microphone headset product.  Both were approx $AUD20 delivered slow boat from China to my door.

The intent of both is that you have one end that plugs into your guitar, one end that plugs into a USB port on your computer.  From a very simplistic viewpoint the cable consists of the right impedence at the guitar input, an A/D (Analogue to Digital) conversion, and the USB data stream to your PC.

Now I always talk about recording direct to PC. And when I say PC I also mean Mac.  The bottom line is that PCs are everywhere, they are extremely flexible, they can be a bit of a pain, but you don't need to buy anything very special to be able to record good sound.  You probably already have a PC, you are probably looking at it right now.  So let's not talk about the other recording medium options, this blog is going to be long enough already...

In my article I wrote several years ago about recording direct to PC, I talked about using the built in soundcard.  One problem with that is the input impedence - if you are going into a "line-in" socket on the soundcard (avoiding any extra pre-amp), then you'll probably have to max out the gain on the guitar and in software, if you are going into the microphone socket you probably have to turn the gain down to one click above zero.  It's a hit and miss affair.  It's how I've recorded for years, reasonably successfully, until I changed to a laptop that caused me no end of problems using the built in soundcard.  Hence the experiment with USB guitar cables.

The immediate advantage of the USB guitar cable is it's built with the right impedence for a guitar and has the right jack.  With any luck, the A/D conversion, done by a tiny little standard chip that lives in the USB connector (basically a little soundcard) will be even better than the soundcard in your PC.

If you've looked into this sort of thing before you've probably heard "latency, latency, latency."  As in the delay between your signal being converted and recorded.  There are (for Windows, Mac is not an issue) low latency drivers - the problem is that latest versions of Windows abstract hardware (a soundcard has to meander its way through layers of internal software to do anything) - so special drivers are required to bypass this and cut out the delay.  Which means the drivers are dedicated to a specific chip, and worse, the sound editing software needs to be able to handle the driver.  The main low latency driver I have come across is the ASIO driver originally developed by Steinberg, and if anything was close to a standard that would be it.  Unfortunately, it is proprietary and something like the free sound editing program Audacity is not allowed to include it.

One of the USB cables I used was ASIO driver compliant.  I found the ASIO driver tedious, and my favourite sound editing software did not support it.  Whereas the other USB cable was not ASIO, you just plug it in and Windows automagically loads generic drivers and it just works with every piece of sound editing software I have - because really, it's just another soundcard.

The more I thought about the latency issue, the less important it seemed.  The ASIO USB cable has an extra port on it for headphones - because an electric guitar player wants not only to hear what he is playing, but wants to hear it through digital effects as well. Yeah, head bangin' out some Metallica with a Kirk Hammet software effects plugin.  Ah ha! - latency is about playing a note, gets converted to digital, goes into the PC software to get distorted, gets converted back to analogue and finally into headphones.  Yeah, you wouldn't want that trip to take more than say 50 milliseconds, it would be very disconcerting.

But for us acoustic players, unless we are playing against some effects, it's not an issue.  So what if your recording is 50ms behind realtime...doesn't make a difference, it still got recorded!

I did a test to gauge what the latency of the standard driver USB cable was - I recorded the normal driver USB cable and the laptop soundcard at the same time, then zoomed into the signal and you know what - there was less latency on the USB cable relative to the soundcard!  Since I've never had a problem listening through headphones on the soundcard, the USB cable is not going to be a problem either.

Look I'm giving USB guitar cables a big thumbs up for just recording your acoustic playing plugged in.  Going further, but being well aware your patience for this blog is at it's end, have some bullet points:

* Capturing Video and Audio - when you use your webcam, it is using the integrated microphone.  Bleugh.  What you need is some software (for example virtualdub) that allows you so say "no, don't use that crappy webcam microphone audio feed, use this nice USB guitar cable feed instead".  I was very pleasantly suprised that virtualdub easily synchronised the audio and video feed, which had plauged me with problems on soundcards - nothing worse than audio and video out of sync!

* Signal to noise ration - yes, there is noise.  I found it to be less than the soundcard, but it's not zero.  I think the specifications talked about having a -80db noise floor, I was seeing more like somewhere between -60~50db, but I didn't do a proper analysis, that's just a thumb suck.  With the guitar unplugged but the USB cable online, zero noise.  That may be as a result of a noise gate.  With the guitar plugged in, but not being played, an "insignificant" (good versus great) amount of noise, undetectable to my ear, but apparent on the recorded waveform, was present.  Is this from the guitar or from the cable?  I don't know.

* Multi-track recording - now this is where it gets interesting.  The cable is suitable for any instrument, it worked fine with the electronic drums, and I'm sure a keyboard would be the same.  Four USB ports - four cables - four multitracked inputs - instant band recording, with separate track editing and mixdown!  This is something I'm hoping for.  My concept: I'm playing through my guitar amp, my drummer mate is playing through the drum amp, we are hearing the sound and playing off each other.  A feed is taken from each amp, and fed via two USB cables into multi-tracking software.  The feed comes from headphone sockets that all amps generally have.  It works just fine...but you guessed it...when you are plugged into the headphones socket the main speaker is turned off!  Not so good for jamming.  I'm going to have to think a bit harder about that one.

There is plenty of stuff still left to talk about, let me know if there is a direction you are thinking about to discuss!



  1. Just an idea... you could use a Y splitter cable from the instrument... one end goes to the amp, one to the computer.

  2. I'm sure there exists such a thing as a splitter, but unfortunately that link isn't it - that one is taking two mono signals and turning them into a single stereo plug.

    I don't think there is an issue with splitting, because the input impedance of the cable is 1M Ohm (quite high) and amplifiers are about the same. A guitar output impedance is a couple of K Ohms. What this means is there would be insignificant reflections from the guitar to the two devices (not impedance matched like an antenna), the two devices in parallel wouldn't look much different to just having one (eg 1M in parallel with 1M is 0.5M, not a _huge_ difference) so the signal would not be degraded much at all to either - guitar to device is just about transfer of voltage.

    If that didn't make any sense, it's because I'm not totally sure ;) I think I should give it a try and report back!


  3. All the impedance stuff is way over my head... but it may be worth a try to solder a cable up as a test. *You* wouldn't actually *buy* a pre-made cable anyway... would you? :)

  4. Ha, to be honest, these days I would almost always buy a pre-made unit - saves me the time to build it and almost always a machine or someone in a sweat shop is going to do a better job than I am :)

    But, in this case, I didn't find any pre-made cable so I had to build one. Okay, so it cost me about $3 and about 30 minutes to make it (15 of which was finding my soldering iron)...

    Results? Pretty much exactly as expected, a dip in volume when you are plugged into two loads. To my ear there was no change in frequency response, it sounded exactly the same.

    That's the thing, it's pretty much all _real_ impedence. As far as signal transfer goes, the only difference is that there is a bigger resistor on the other end. If the amp/USB cable input had more capacitance and/or more inductance then that would change the frequency response and yes, you might start to hear a difference. However I expect the input stage of any amplifier or A/D converter would have very low capacitance and negligable inductance...otherwise it is messing with the incoming signal!

    Conclusion: Splitting a guitar signal (active pickup in my case, but the results should be the same for passive) across two loads (an amp and an A/D computer cable in this case) drops the signal level a bit on both devices, maybe introduces a bit more noise, but the signal has no apparent change just turn it up a bit more! :)

    Perhaps I might do a loss analysis...pluck a note over and over again, exactly the same every time, and read off the waveform the difference between 1 load and 2 loads. However it's going to be 1/2 - if the two devices impedences are the same.


  5. I didn't know that those cables (guitar at one end, USB at the other end) were not common yet. In a way, hearing you h

  6. Sorry Jaw, I'm pressed "Send", so let me continue...

    Hearing you were trying those cables, and knowing that you know about technique (I don't), it made me feel proud about the choice I made around two years ago (when I started my blog), to buy an audio interface device with USB cable. The only reason I did that was that whenever I had tried to record direct to PC with standard port (small one into "line-in"), I had loose connection problems, so I thought USB was a good option (without knowing anything about technique, impedence and so on...). Yes it is a more expensive option, but still reasonnable (the one I have is called Lexicon Lambda, around $120 I believe). What do you think, for others I mean (or even for you)?

  7. I seem to recall we discussed USB cables a while back, I didn't give them a chance at the time because the $5 direct cable I made was working out fine. When it was no longer working out for me, and I needed the ability to record multiple channels (left/right channel on a direct stereo plug is fine for two channels, I was thinking more than that) I looked harder at USB cables.

    The key for me is that you can get a generic USB guitar cable delivered via slow boat from China to your door for around $20. It has a sound quality better than my computers soundcard and it is a lot less mucking around with gains and ports in the hit and miss world of direct-to-soundcard cable.

    I'm pretty happy to say that for anyone who wants to record plugged and get a really good (not necessarily great) sound, a USB guitar cable is the way to go. I wish I could hand them out to all my favourite guitarists who record onto youtube without a decent audio feed! :)

  8. I've been waiting for this for a while now as you know. It all sounds fairly simple, I can't wait to try it. Thanks for the tips JAW

    Ryan G

  9. I will also update my webpage with the new info on USB guitar cables. The webpage has always been too wordy, I need to simplify it down into manageable chunks. The more guitarists that record plugged in audio on their video for the internet the better! :)