Topic: Evening out large level differences across a set

[aberg]

I recorded a show the other night and was a little too jumpy with my level corrections. There are some segments during the 1st set that are near clipping, and others at -3db peaks or -6db peaks. I have Audacity, Audition, and WaveLab right now and can use all 3 if need be. What is the best approach using any of that software to nicely even the whole set out to good levels... I know I can just select chunks and amplify, but I was wondering if there's a better approach.

[Patrick]

normalize the file.  any program can do it. 

[mmmatt]

compression would even things out a bit but many here do not like to use compression.  If you do, don't go over a 2:1 ratio and 1.6:1 or so is better (less compression)  make your threshhold about 18-20db.  All my opinion of course.  Going through the line and trying to increase at spots can be more noticable than comression if you aren't very careful but if you don't like compression that end result will be better.  I think the compressor in wavelab is pretty good.  I use that ne a lot.

matt

edit
normalizing will not change the levels of insividual passages.  Normalizing is like just turning up the volume knob on the whole file.

[aberg]

compression would even things out a bit but many here do not like to use compression.  If you do, don't go over a 2:1 ratio and 1.6:1 or so is better (less compression)  make your threshhold about 18-20db.  All my opinion of course.  Going through the line and trying to increase at spots can be more noticable than comression if you aren't very careful but if you don't like compression that end result will be better.  I think the compressor in wavelab is pretty good.  I use that ne a lot.

matt

edit
normalizing will not change the levels of insividual passages.  Normalizing is like just turning up the volume knob on the whole file.

Yeah, I knew normalizing wouldn't do the trick. I've never used compression before, just hard limiting, which might be a form of compression? I don't really like the result of hard limiting... what exactly does compressing it do?

[mmmatt]

compression will reduce the volume of the parts of the set over the threshhold by the amount specified

[aberg]

compression will reduce the volume of the parts of the set over the threshhold by the amount specified

So is that the same as hard limiting? What I really want to do is bring the -3db and -6db section up to near clipping without actually clipping, and without it all sounded fuzzy like the hard limiting feature in Audition... any ideas?

[BayTaynt3d]

This really sounds like a job of fader volume automation to me. Compression will alter the sound and normalizing won't help here, so it seems like ideally, you'd just reverse the volume changes with volume automation. Not sure about the app you use, cause I use others, but you should be able to get a volume line across the entire track, and then you can move it up and down in the spots you need to, but like mmmatt says, it's a fine art to do it and not make it noticeable -- but from the sounds of it, it's ALREADY noticeable. I'm sure this is obvious now, but this is why you need to leave some headroom and set it and forget it -- because mopping this up in post sucks. On another point, if you are taping open (not stealth), it's nice to jot down on paper the timecode where you make these adjustments during the recording so you can find the exact spots in post easier. Good luck...

[aberg]

This really sounds like a job of fader volume automation to me. Compression will alter the sound and normalizing won't help here, so it seems like ideally, you'd just reverse the volume changes with volume automation. Not sure about the app you use, cause I use others, but you should be able to get a volume line across the entire track, and then you can move it up and down in the spots you need to, but like mmmatt says, it's a fine art to do it and not make it noticeable -- but from the sounds of it, it's ALREADY noticeable. I'm sure this is obvious now, but this is why you need to leave some headroom and set it and forget it -- because mopping this up in post sucks. On another point, if you are taping open (not stealth), it's nice to jot down on paper the timecode where you make these adjustments during the recording so you can find the exact spots in post easier. Good luck...

Good point. I generally like to get as close to 0db as possible, but maybe it's best to set for peaks around -3db or even lower and just boost it all at once later if need be.... a lesson learned today. Thanks! I'll try and look into that volume automation tool... sounds like it would be the same as just selecting chunks and amplifying tho.

Can anyone also recommend the best way to minimize slight bass distortion? For some parts that I had the levels too high, the bass clips a bit and there's some slight audible distortion.. nothing too serious, but with cheap speakers or headphones, it would be bad...

[mmmatt]

compression will reduce the volume of the parts of the set over the threshhold by the amount specified

So is that the same as hard limiting? What I really want to do is bring the -3db and -6db section up to near clipping without actually clipping, and without it all sounded fuzzy like the hard limiting feature in Audition... any ideas?

hard limiting is compression but it is drastic compression.  Essentially you are cutting the dynamic peaks completely out.  with a compression scenario like I stated above it won't be nearly as noticable.  When you are manualy adjusting the levels in post, you have to do it gracefully.  It sounds easy but it isn't!  Some programs have "envelope functions and adjustable fader options.  By just jaklcing up points you are going to just make the problem more apparant unless you were jerking your levels up and down and you find the eact points.  A good way to do it manually is to find a point where the level goes from low to high.  find the point where you raised it up.  Lets say you raised the level up 6 db over a period of 6 seconds.  In this scenario you would then take 1 second chunks and do the first at -1db, then next second -2db, then -3, -4 until you hit the -6 point and then drop that whole thing down until you reach the next time you dropped levels.  Then after you get it all evened out go through and normalize to 96% or whatever you like and you should be goldn.  To just jack it up and down 6db at a time will be a mess.
    Don't discount the compression though... it isn't nearly as harsh as hard limiting.  you should try it before you discount the idea.  it may help and it may not, but it takes only a few minutes to know.  Either way, when you are done dealing with this bullshit you will really appreciate Taint's comment about leaving headroom and only adjusting for an over when you are in the field.

Matt

[aberg]

I think it's just taking some time to get used to my new rig.... but thanks for your advice. I will try it when I'm home from work. Compression sounds like much less of a pain, so I'll try that first and see what I think... I can also post a screen shot of the waveform if you might have other advice upon seeing it.

[momule]

On your bass prob's you may try cutting the freq range between 60-90Hz (adjust to taste) I find that’s where most of the crunchy stuff comes into play.  You can do this either by a good old fashioned graphic EQ or better yet a Paragraphic EQ.