Audacity Forum

There's not a lot of difference, but the USB sounds "cleaner" and a shade brighter to me.

Do you know what bit depth the ART runs at internally? There's a bit of high up noise on the xonar recording that sounds remarkably like "dither" noise. I wonder if the ART is running at 24 bit internally and down-sampling to 16 bit before it sticks the signal back out of its Line-out. One would expect that the ART input to line-out would be all analogue, but perhaps not. On the other hand, perhaps the xonar is struggling with the low signal level. Either way, the USB seems to have the edge to my ear.

It would be interesting to hear how they compare when handling a bigger signal - the difference would probably be a lot closer.

You did miss one combination. The MacBook Pro has an excellent stereo digitizer. Analog Line-Out from any of the preamplifiers straight to the 1/8" stereo Line-In of the MBP. It can be a mono plug. You'll just get "Left" in the show.

That's how I do all my recordings.

Yes, it does bother me that the "analog" pathway had that haystack hump of trash around 20 KHz. That's pretty classic digitizer gone wild. Otherwise, there's not a lot of difference.

Also a little odd that they seemed to have similar "dark" noise. I wonder if we're looking at the microphone noise floor.

fffffffffff, but way in the background.

Koz

kozikowski wrote:You did miss one combination. The MacBook Pro has an excellent stereo digitizer. Analog Line-Out from any of the preamplifiers straight to the 1/8" stereo Line-In of the MBP. It can be a mono plug. You'll just get "Left" in the show.

Actually I didn't forget about that... I tried it before, but with no luck, probably because I was using a stereo cable (the only I had available then).
I shall try it again soon with the mono cable.

kozikowski wrote:Also a little odd that they seemed to have similar "dark" noise. I wonder if we're looking at the microphone noise floor.

fffffffffff, but way in the background.

That's what I thought too. Either it's that or the noise coming from the preamp itself...

stevethefiddle wrote:Do you know what bit depth the ART runs at internally? There's a bit of high up noise on the xonar recording that sounds remarkably like "dither" noise. I wonder if the ART is running at 24 bit internally and down-sampling to 16 bit before it sticks the signal back out of its Line-out. One would expect that the ART input to line-out would be all analogue, but perhaps not. On the other hand, perhaps the xonar is struggling with the low signal level. Either way, the USB seems to have the edge to my ear.

It would be interesting to hear how they compare when handling a bigger signal - the difference would probably be a lot closer.

I searched but couldn't find any info about the internal bit depth of the art. At first I thought it would be strange if they did analog-digital-analog, but then I realized another thing... on the analog outputs of the art (line and headphones) you can listen to computer playback sound (coming from the usb connection) like an external soundcard and there's a mixer balance button: all to the left just the sound of the mics, all to the right just the sound from the computer (and in middle all shades between one extreme to the other). So I wonder how the mixing is done here...

bgravato wrote:

kozikowski wrote:Also a little odd that they seemed to have similar "dark" noise. I wonder if we're looking at the microphone noise floor.
fffffffffff, but way in the background.

That's what I thought too. Either it's that or the noise coming from the preamp itself...

Probably a bit of both, but considering you're recording a relatively low level sound at a distance of about a meter? that underlying fffffff is pretty low level - it's less obtrusive than what we are suggesting may be dither noise. I suspect that you would need to spend a lot more money to get the SNR noticeably better. For the money that you've spent the SNR is better than I would have expected - I could live with it

bgravato wrote:So I wonder how the mixing is done here...

Without the technical information we're guessing, but it seems likely that it is done digitally. I don't think we can hold this against the ART - against a full-scale signal we're looking at better than -60dB with the gain cranked up to maximum. To get this into perspective, it's about a 16dB improvement on the MicPlug - 16dB is a lot.

Something we have not looked at much, but since we're talking SNR, the volume level rises rapidly as the microphone approaches the sound source (inverse cube law isn't it?). With the microphone position we are looking for the "sweet spot" - the best balance between all relevant factors. Too far away and the direct signal is too low, too close up and you capture the sound from only part of the guitar (there is a huge difference in sound from the air projecting from the sound hole, the resonance of the wood, and the direct sound from the vibrating strings). Also there is the "proximity effect" which boosts low frequencies as the distance from source to microphone decrease. This last effect is one that we have not really played with. The proximity effect can be countered (if not too great) quite easily by rolling off a little bass - either in the microphone with the bass roll-off switch (the SC1100 has a bass roll-off switch?) or in Audacity after recording. If there is a sweet spot closer to the guitar than you are currently using, that has a nice balance of all of the guitar, but suffers only from the proximity effect, then by using a little gentle bass roll-off you will be able to take advantage of the greater sound level and achieve improved SNR. It will however reduce the relative amount of ambiance that you pick up.

The other thing we have not really talked about is dynamic compression. I have serious reservations about dynamic compression on acoustic recordings in that I have so often heard it horribly overdone. I like dynamics - it brings light and shade into the picture. Having said that, if used sensitively it can offer very pleasing results.

When listening to real-life live music, your ears will automatically compensate for variations in sound level - a kind of natural compression. This is triggered by all sorts of cues, including visual cues. When you see a guitarist is about to take a mighty swipe across the strings, your ears know that it is going to be loud and brace themselves for the impact. Listening to a recording on the other hand, you miss many of the cues - the only stimulus is the sound. "Hearing" is a truly remarkable ability - Notice how you can listen to someone in a crowded room and hear them distinctly, but listen to a recording and you hear just a wash of noise? Ears are able to "focus" on sound.

Before I ramble off too much - where am I going? Dynamic compression, when used sensitively, can assist the ears. It can to some extent compensate for some of the cues that are missing in a recording. A side effect of compression is that it will increase the relative amount of ambiance in a recording - another reason to keep the reverb, natural or otherwise, reasonably low in the original take. As said before, reverb can be added, but is can not easily be taken away.

<<<reverb can be added, but is can not easily be taken away.>>>

Unless you use the PAF.

I would expect processing the latest efforts through the same, but more gracefully applied noise gate would take out the very last of those noise effects. Then apply the compression as needed, but I don't see any great need.

That's for a stand-alone performance. If you produce anything commercial, you may find that everyone else's performances are a lot louder than you are and since the digital channel overloads in the same place for everybody, that's when you need Chris's Compressor and bend the expression variations a bit.

Koz

stevethefiddle wrote:Something we have not looked at much, but since we're talking SNR, the volume level rises rapidly as the microphone approaches the sound source (inverse cube law isn't it?). With the microphone position we are looking for the "sweet spot" - the best balance between all relevant factors. Too far away and the direct signal is too low, too close up and you capture the sound from only part of the guitar (there is a huge difference in sound from the air projecting from the sound hole, the resonance of the wood, and the direct sound from the vibrating strings). Also there is the "proximity effect" which boosts low frequencies as the distance from source to microphone decrease. This last effect is one that we have not really played with. The proximity effect can be countered (if not too great) quite easily by rolling off a little bass - either in the microphone with the bass roll-off switch (the SC1100 has a bass roll-off switch?) or in Audacity after recording. If there is a sweet spot closer to the guitar than you are currently using, that has a nice balance of all of the guitar, but suffers only from the proximity effect, then by using a little gentle bass roll-off you will be able to take advantage of the greater sound level and achieve improved SNR. It will however reduce the relative amount of ambiance that you pick up.

Yes the mic has a low-cut switch, I've tried that, I think I even posted a sample of that... IIRC Koz didn't like it much then hehe (here's the link for that sample)

I bought that article on soundonsound that Koz mentioned and I've been reading it. They talk about some alternative mic positionings which I might try. (BTW if anyone interested in the article I can email the pdf to you... you know... for evaluation purposes... if you really like it you should buy it )

stevethefiddle wrote:The other thing we have not really talked about is dynamic compression. I have serious reservations about dynamic compression on acoustic recordings in that I have so often heard it horribly overdone. I like dynamics - it brings light and shade into the picture. Having said that, if used sensitively it can offer very pleasing results.

I don't fully understand yet how compression works (I know the basic idea but...) I'll have to read deeper about it, but I think I'll save it for a bit later, if I start approaching too many variables and different techniques at the same time I will start to straggle too much.

One of my many hobbies is archery and one of the most painful neverending processes in archery is the so called "bow tuning". In order for the arrows to fly perfectly straight you have to ajust a lot of variables to achieve that, and they're all connected, so when you change one variable it will affect the others too... At some point you just want to forget about tuning and start focusing on actually shooting and just give a damn about how straight (or not) the arrows are flying...

stevethefiddle wrote:When listening to real-life live music, your ears will automatically compensate for variations in sound level - a kind of natural compression. This is triggered by all sorts of cues, including visual cues. When you see a guitarist is about to take a mighty swipe across the strings, your ears know that it is going to be loud and brace themselves for the impact. Listening to a recording on the other hand, you miss many of the cues - the only stimulus is the sound. "Hearing" is a truly remarkable ability - Notice how you can listen to someone in a crowded room and hear them distinctly, but listen to a recording and you hear just a wash of noise? Ears are able to "focus" on sound.

So true... When I'm at live concerts I often close my eyes or look away from the stage in order to focus more on the sound and be less distracted by the visuals... It's amazing how the brain can naturally create such complex filters...

stevethefiddle wrote:Before I ramble off too much - where am I going? Dynamic compression, when used sensitively, can assist the ears. It can to some extent compensate for some of the cues that are missing in a recording. A side effect of compression is that it will increase the relative amount of ambiance in a recording - another reason to keep the reverb, natural or otherwise, reasonably low in the original take. As said before, reverb can be added, but is can not easily be taken away.

At this rambling pace I believe this thread might reach the 20 pages record soon

bgravato wrote: Yes the mic has a low-cut switch, I've tried that, I think I even posted a sample of that... IIRC Koz didn't like it much then hehe (here's the link for that sample)

Too extreme isn't it. The cut-off frequency (100Hz) is too high and too steep. Most of that recording is markedly bass-light, except for the notes at about 11.5 seconds where it sounds like it has caught the sound hole more.

Something I've noticed on several recordings so far is that there's been a bit over-emphasised at about 100Hz, which could be giving the impression that some of the recordings are more bassy than they really are. I can't tell why it's happening - mic position, microphone, the room, or whatever, - but the next time you get a recording that sounds a bit bass heavy, try using the equalizer effect in Audacity and just dropping the 100Hz a bit.

bgravato wrote:I don't fully understand yet how compression works

It's a black art

bgravato wrote:I think I'll save it for a bit later, if I start approaching too many variables and different techniques at the same time I will start to straggle too much.

Sounds like a wise decision. There's no rush - you are already getting very nice recordings.

bgravato wrote:At some point you just want to forget about tuning and start focusing on actually shooting and just give a damn about how straight (or not) the arrows are flying...

Remind me of that if ever you are shooting arrows anywhere near me

I've been trying some different configurations, namely putting the mic higher above the guitar but closer to it and facing more down. At first try it sounded better, but after a few tries not sure anymore... Everytime I go through the positions I tried before I have different feelings about it... So I'm going to stick to the "vanilla" setup (+/- 30cm in front of the 12th fret and stick to it for a while and focus on recording something (after my neighbour stops singing in the shower...)

BTW, I forgot to comment earlier that the position of the mic you see in the pics is not 1m from the guitar, it's more like 50cm. The last recordings I did I had it a bit closer like 40cm or so. Today I tried getting it as close as 20cm but still not sure about which I like most... So sticking to the 30cm flavour and focusing on other things...

stevethefiddle wrote:

bgravato wrote:At some point you just want to forget about tuning and start focusing on actually shooting and just give a damn about how straight (or not) the arrows are flying...

Remind me of that if ever you are shooting arrows anywhere near me

Beside practicing I also teach archery, the first thing I say to the newcomers is that "we have a golden rule here: if anyone is in front of the shooting line no one shoots". Safety above all (and if the arrows don't fly straight they still hit the target... just not so consistently...)

Close doesn't work. That picks up too much accidental trash and doesn't sound right. Who listens with their ear closer than a foot/300mm away from a guitar opening? There's a sweet spot between too far away/slightly dull and too close and mechanical or off tonal balance.

Koz

I think I might have found a sweet spot but who knows... maybe tomorrow when I try it again it will probably sound different

Steve, any tips on the calf reverb parameters? I've tried playing around with it but it always kind of sounds the same to me.... and it always feels like a bit too much reverb...