Wednesday, June 28, 2006

It's Not My Mac

Last night I picked up a Logitech USB headset on the grounds that even if the audio quality isn't very good, I could use it to (1) test my computer and verify whether this device also produces glitches during recording, (2) use it as a reference to determine whether the low-level issue is specific to my Snowball, and maybe (3) use it to record, at least temporarily, while the Snowball is getting fixed.

What I determined is that recording works fine using the USB headset: it does not drop buffers and "glitch" the way that the Snowball does. If I record with a number of plugins going, I can wind up hearing glitched audio if I monitor on the headphones, but it doesn't appear to affect the recording. Not really a problem since I actually intend to record fully dry, using minimal CPU, and master later.

So, this means I have a pretty strong case to claim that the Snowball has a problem with the buffer loss, not my computer. I'm going to send it to BLUE and let them try to figure out what is wrong.

The audio level of a dry recording is actually comparable to what I'm getting from the Snowball, at an RMS level of about -35 dB, even with the gain on the Logitech headset turned all the way up. This suggests to me that I just have to live with it and try to compensate in post, since I can't add more gain on either device prior to the A/D conversion.

I also spent a little time experimenting with the "Fish Fillets" trio of Floorfish (downward expander/gate), Blockfish (compressor), and Spitfish (de-esser). The results with the audio from the Logitech headset have been mixed. Despite the "deluxe" label, the Logitech headset is just not really a high-quality audio device. The cord from the headset is strangely thin, and it seems to pick up quite a bit of electrical noise from other items in the room, including perhaps my Netgear wireless network hub, depending on how I turn my head (maybe I need to turn it off when I record).

The mic, although mounted away from direct breath blasts, still manages to be sensitive to pops and has some strange resonances, which manifest as certain consonants suddenly "essing" or "clicking" a lot. If I move my head, I get cable-handling noise. It does do a reasonably good job of rejecting background noise, though, which is good. I think I'm just going to go ahead and use it, and do what I can with the plug-ins.

One problem seems to be that the levels I get using the plug-ins as _insert_ effects while recording, monitoring the results, does _not_ match the levels I get when I use them as _master_ effects and post-process the dry file. Not at all. I have no idea why this should be the case -- the insert effects are set to fully "wet" -- but it means that I can't effectively _preview_ just how the inserts are going to change the dry recording; I have to apply it to at least a section of the recorded file, listen to the result, tweak the settings, apply again, etc. I don't know whether this problem is specific to DSP Quattro, or not. I should give it a shot in another plug-in host, like maybe Audio Hijack Pro.

Next time I hope to have some examples illustrating the results I'm getting when applying the plug-ins, and maybe I can recruit some feedback about how to improve the settings as much as possible.

No comments: