Volume and bit rate reduction, software "audio engine"

M

mjhigg

New member
I have a couple of digital audio questions that have been nagging at me.....

1. I seem to remember reading somewhere that when converting analog to digital, it is preferable to go as hot as possible because lower volumes are actually recorded at a lower bit rate. Is this true, or has Geary's Hampshire Special Ale (HSA to us Mainers) seeped into my sulci?

I ask this, because I am trying (for the sake of ease of use) to get rid of all of my outboard gear, including my dBx166, which I have always used before converting to digital. If this is true, though, it would seem preferable to use an analog compressor or limiter before conversion to maximize the signal, rather than compressing with a plugin after going digital.


2. Occasionally, I hear reference made to the "audio engine" of a software multitrack program being revamped, or sounding better than others. This seems a little strange to me. I guess I always thought that the quality of your digital audio depended on your converters, and that, after going digital, the multitrack just routed it from there. Anyone have different experience with this?
 
mjhigg said:
1. I seem to remember reading somewhere that when converting analog to digital, it is preferable to go as hot as possible because lower volumes are actually recorded at a lower bit rate. Is this true, or has Geary's Hampshire Special Ale (HSA to us Mainers) seeped into my sulci?
Click to expand...

What you've read is actually not that far off. Consider 16-bit audio. At digital 0 all 16 bits in the word will be turned on, meaning they will all be 1. Anything less than that will have 0s in the word. If the signal is really low, then several of the last bits will be 0, effectively lowering the resolution. In itself this is not bad, but if later on you're going to increase the volume of the signal, apply EQ and such, then there simply isn't enough information and you start dealing with rounding errors. So the hotter the signal, the more "on" bits, more information and more accuracy for further DSP.

I ask this, because I am trying (for the sake of ease of use) to get rid of all of my outboard gear, including my dBx166, which I have always used before converting to digital. If this is true, though, it would seem preferable to use an analog compressor or limiter before conversion to maximize the signal, rather than compressing with a plugin after going digital.
Click to expand...

Your thinking is rational, but I wouldn't use a limiter. I would use compressor only if there are a lot of transients in the audio that you're recording, something like vocals probably would be a good idea. I wouldn't compress things like synths though as their volume is usually pretty consistent. While doing this you don't want to overcompress, just use enough to tame the signal w/o sucking the life out of it, or "hearing" the compressor. For example, when recording myself, I have realized that my "k"s "t"s and "p"s are quite sharply enunciated. I actually do this on purpose as then I distort the hell out of it, but I still like to keep the inteligibility of the words. I use a compressor before the AD conversion to tame those transients, but only those w/o affecting the rest. That way I can record a hotter signal w/o worrying about those transients clipping the converter.

2. Occasionally, I hear reference made to the "audio engine" of a software multitrack program being revamped, or sounding better than others. This seems a little strange to me. I guess I always thought that the quality of your digital audio depended on your converters, and that, after going digital, the multitrack just routed it from there. Anyone have different experience with this?
Click to expand...

This has to do with several aspects. One is digital summing, i.e. how signal gets affected when you mix it down. In most DAWs and audio sequencers you'll notice that the more tracks you add the more likely it is that the mix outputs will clip. It is a good idea to keep a tab on this, and lower the faders on the individual channels to compensate. Because of this you'll hear a lot of debates about digital vs. analog mixdowns. Another aspect has to do with downsampling and dithering. This comes into play when your origianl tracks are recorded at say 24/96 and your final product is going to be delivered on Audio CD in which case you have to downsample and dither your audio at 44.1/16. Obviously the better the dithering alogrithm is, the better the quality of the downsampled/dithered audio will be.

Hope this somewhat clarifies things for you.
 
You must log in or register to reply here.
Back
Top