I like to listen to Classical Music, but I don’t have much time to do it in the quiet environment of my den. Most of my listening is done in the relatively noisy environment of an automobile. Because of the wide Dynamic Range of Classical music, this presents somewhat of a problem. I therefore have decided to burn 2 CDs of my LPs – one with no adjustments to the Dynamic Range for my home environment, the other to be listened to in my auto. Thus I am exploring Dynamic Compression.
After much searching on the Forum for info on this topic, I have decided to experiment with Chris’s Dynamic Compressor (CDC). I am currently conducting A-B listening tests of captures in Audacity of Dvorak’s “New World Symphony”, an excellent example of wide Dynamic Range. I am using the default values in CDC, adjusting only the “Rate of Compression”, up to but not exceeding .05. I was surprised to hear how much the Loud passages were increased, along with the quiet passages. I noticed that the Playback Meter registered in the Red, indicating Clipping. I didn’t think that this was supposed to happen?
So I went back to my original captures to find out what was happening before compression. The following paragraph occurs before use of Chris’ DC, and has nothing to do with it.
I discovered that when I “Amplify” at -1.0 dB, I get no clipping on playback, with the Playback meter registering -2dB during the loudest passages. But when I “Amplify” at -0.5dB, clipping does occur during playback with the Playback meter going into the Red. This surprised me. Is this normal? I think the proper conclusion is that I should not “Amplify” beyond the -1.0dB level to avoid clipping. Question: Have I reached the correct assumption?
Charlie, I was pondering this recently following some of you recent potings on the forum - the two CD solution os good way of dealing with this - but my thoughts ran along the lines of “why on earth can’t the makers of car audio players fit them with onboard Dynamic Compression, switchable on and off with a button?”
Great question. As much as I enjoy classical music in the quietness of my home sanctuary, I have much more time to listen to it when I am on long road trips – frequently lasting 6-8 hrs in length. I find it very annoying to have to keep fiddling with the volume control in the auto, constantly adjusting it up and down. Right now, I am trying to learn how to use the proper adjustments for Chris’ Dynamic Compressor to find a happy compromise for the loud/quiet issue.
I have already decided that the often suggested Compression Ratio setting of .5 is probably too much for my taste. I suspect that I will be in the range of .2 - .5 setting on the Compression Ratio and something less than .99 on the Maximum Amplitude setting.
You stated in an earlier post that you normally “Amplify” in Audacity at .6. Do you find that this setting avoids clipping?
I usually prefer to play on the safe side… and I rarely amplify anything beyond -3dB or -2dB. When recoding live audio I usually try to set the input gain so that the peaks will occurr at about -6dB. Maybe this is being too safe but I always prefer to have lower volume than clipping…
my thought is – why do they insist on so much dynamic range in the first place. 30 dB is the practical limit to listen to playback – unless you are in an anechoic chamber where no outside noise gets in.
One reason I wont go to a live concert is the low parts are under the people noise. and the highs are amplified so much that i have to wear ear plugs and headphones.
Great minds Charlie, that’s exactly what I do
The “Home” version I do not compress at all. The car (auto) version I compress with CDC - usually on default settings (unless the defaults sounds wrong).
CDC will amplify everything above the threshold, but amplifies very loud sounds less than it amplifies sounds that are just above the threshold and in that way makes all of the sound a bit more equally loud so that you can hear it over traffic noise. It should not push the meters in the red, but the exact amount of amplification depends on the content of the music - with some music the processed sound will have higher absolute peaks than other music. Not surprisingly many of the settings can have an influence on the maximum peak in the processed track, and that is what the last slider (Maximum amplitude) is for. If the meters are pushed into the red you need to “undo” (Ctrl+Z), adjust the “Maximum amplitude” slider a bit to the left and try again.
If the processed sound looks to be a bit too quiet, you can use the Amplify effect to bring the level up a bit.
Brilliant - that could be how you make your next $million
-3dB is possibly a tiny bit over cautious if you are then burning directly to CD, but the audible difference between -3 dB an 0 dB is very small. I usually go for a peak level of about -0.3 dB before the final export, which after thorough testing I’m satisfied is sufficient to avoid any clipping, even when converting to lossy formats. When recording live I aim for the same target as you (-6 dB) unless I’m recording an “unknown quantity” such as a musician that I’ve not seen or heard play - in which case I’ll allow substantially more so as to avoid the need for retakes. The bass player in my band always plays 4dB quieter during sound checks than during “takes” and he is remarkably consistent at that
After some trial and error tests today, I have discovered that I can avoid clipping after compression by simply setting the Maximum Amplitude in CDC at no more than .95. This seems to work fine. I’m still playing around with the Compression Ratio setting. The default as you know is set at .5, but that almost over does it for me. I think the answer for me is probably going to be somewhere between .3 -.5, maybe varying somewhat depending upon the musical selection.
Thanks for the help.
And the next $million after that will come from getting turntable manufacturers to implement Brian Davies’ ClickRepair in hardware and then integrate it into the TT. I have actually discussed this with Brian - but he doesn’t have the time, the inclination or the contacts to pursue it.
Ok Bruno - youv’e convinced me - I’ve changed my personal workflow to amplify to only -2.0dB to give a little more headroom
No this is the level that I normally aim for too when I am transcribing LPs or recording FM off-air. With Audacity set to work in 32-bit float there is plenty of headroom for later amplification without distortion. Digital is a big change from the old analog tape days when (with the right tape) you could safely push the recording into the red for brief excursions to get a goo saturated signal on the tape.