Topic: Help with post editing - quiet show with loud claps

[Belexes]

I taped Peter Gabriel last night, who is doing a symphonic performance sans the traditional band.  The music is very quiet, but the applause/claps very loud.  I was going to hard-limit the claps (to -1 dB?) and then boost the music portions.  Is that the right order to do this or does anyone have any other suggestion?  I edit primarily in Cool Edit Pro and Audacity. Thanks!

[drewloo]

I had a similar situation like that not too long ago and ended up adjusting the gain down on the clapping just after the applause started until just before it ended.  Had to reduce the clapping from between -3dB to -7dB doing each channel by itself and then raised the entire file until the loudest part peaked at -.1dB.  Took a while, though, but I couldn't hear any abrupt gain changes during the applause unless I cranked the stereo up all the way and even then I really had to listen for it. 

I tried to do it with in wavelab with the Sony Graphic Dynamics direct X plug-in but could never really get the hang of it and always ended up compressing the music, too.  Later on I realized that maybe I should have just focused on highlighting and compressing the sections of applause instead of trying to do the whole file at once. 

However you do it I'd bring the applause down first and then raise the gain on the whole file to just below clipping.  It looks like the applause sections went over 0dB pretty bad in some spots?

[page]

I taped Peter Gabriel last night, who is doing a symphonic performance sans the traditional band.  The music is very quiet, but the applause/claps very loud.  I was going to hard-limit the claps (to -1 dB?) and then boost the music portions.  Is that the right order to do this or does anyone have any other suggestion?  I edit primarily in Cool Edit Pro and Audacity. Thanks!

I'd hard limit the applause and move on. I suspect that you may have other noise issues once you amplify the music to near 0 that might need some work.

in audacity figure out the loudest musical peak, and then hard limit to just above that. Then run the limiter again with a residue value of somewhere around 0.5 for about 2-3db under that peak level, it's sort of like a poor man's apogee soft limiter. That should bring up the RMS levels of the music enough that it's not bad.

Got a sample?

[Gutbucket]

I had a similar situation like that not too long ago and ended up adjusting the gain down on the clapping just after the applause started until just before it ended.  Had to reduce the clapping from between -3dB to -7dB doing each channel by itself and then raised the entire file until the loudest part peaked at -.1dB.  Took a while, though, but I couldn't hear any abrupt gain changes during the applause unless I cranked the stereo up all the way and even then I really had to listen for it.

Works best for me.

[yltfan]

I use the envelope tool in Audacity for this.

[morst]

I would try the way Page is suggesting. What is your peak value now? Did the clapping go "over?"

I taped Peter Gabriel last night, who is doing a symphonic performance sans the traditional band.  The music is very quiet, but the applause/claps very loud.  I was going to hard-limit the claps (to -1 dB?) and then boost the music portions.  Is that the right order to do this or does anyone have any other suggestion?  I edit primarily in Cool Edit Pro and Audacity. Thanks!

I'd hard limit the applause and move on. I suspect that you may have other noise issues once you amplify the music to near 0 that might need some work.

in audacity figure out the loudest musical peak, and then hard limit to just above that. Then run the limiter again with a residue value of somewhere around 0.5 for about 2-3db under that peak level, it's sort of like a poor man's apogee soft limiter. That should bring up the RMS levels of the music enough that it's not bad.

Got a sample?

[Belexes]

Clapping went way over 0 dB.  I didn't want to set my levels based on the claps and wanted to set the level based on the performance.  The music mostly peaked at -6 dB and I taped 24/44.


Some good suggestions here. I may have time to try what Page suggested as well as what Drew said.

[page]

Clapping went way over 0 dB.  I didn't want to set my levels based on the claps and wanted to set the level based on the performance.  The music mostly peaked at -6 dB and I taped 24/44.


Some good suggestions here. I may have time to try what Page suggested as well as what Drew said.

I only suggest that method because I've tried the volume envelopes and it annoyed me to hear the transition no matter how many times I tried. I find the envelopes best for transition between sources or where you temporarily need half of the sbd signal in a matrix/mix and don't want to junk the entire track or have the transition be jarring.

Izotope has a VST plugin for limiting that sounds sweet even when you blast on it, but it won't do the soft limiting effect that I described in step 2. That's the one plugin ability that I think audacity still holds value (for me). I suppose you could accomplish the same result using a 2:1 compression ratio on the top 3db but I find it easier to use the residue function there and for small db amounts I find it sounds fine.

Best of luck.

[Gutbucket]

Can you describe this 'residue function' in more detail?  I don't quite understand what it is or is doing. I'm curious because correcting for overly loud applause this is a pretty important issue on many of my recordings.

Personally, I try to set recording levels for quieter material so that the applause is just under clipping.  That does mean the recordings benefit from the effort of raising the level of the musical passages later, but recording at higher levels and clipping applause requires more radical correction work to sound natural in my experience.  Even for very low-level passages and silent sections, doing that has not yet raised the noise floor of my equipment above that of the room.  Plus, I find clipped applause sounds offensive to my ears even when lowered in level relative to the music.  It still sounds clipped, just not as loud. 

Regardless, I've made plenty of recordings that have the applause clipped, and have wondered about trying a 'clipped peak restortation' tool on those sections after lowering them to non-offenive levels in an effort to restore a more natrual 'unclipped' sound to the applause, regardless if they were limited, compressed, or enveloped to reduce their dynamics.

[page]

Can you describe this 'residue function' in more detail?  I don't quite understand what it is or is doing.

yeah, i'll post a picture when I get home of a drawing. In my experience, it essentially takes the difference between no limiting and hard limiting and averages them based on the residue percentage and that's your final result of limiting.