Stereo Tool

Not so scary

Perhaps, but in my opinion, and like I said, people love (live?) to tell me I'm wrong, but in my opinion, a lot of these spikes are self-inflicted due to aggressive up speeds and high band 2 / band 3 values. I have not seen anyone that is actively making presets use only 1-band AGC. I only did it temporarily while I was trying to figure out how the AGC functioned.

@Brian: High up speeds make it worse, but there are always situations where the new filter is really needed. Of you have Leif's torture test, try the track at 32:30 (When I Grow Up), it's MUCH louder than anything in the minutes before it - without this fix there's no way to get the volume to behave nicely. Even with infinite down speed it's not good.

I do not have that. Is it available for download?

Edit:

Found a download.

The thing that comes to mind here is something I've noticed for a long time - the current AGC implementation never lowers the gain unless it has determined that something in its' look-ahead timeframe will be too loud. The problem, as I see it, is that the "quiet" intro portion of "When I Grow Up" is actually too loud if you use preamp > 1. What about adaptively lowering the gain based on the same look-ahead, factoring in the preamp?

Hm, it used to be. I just searched but I couldn't find it. Does anyone know if it's still available somewhere?

See my edit... Also, the sweeping tests happening at 36:58 need a warning, telling you to turn the volume down... lol

http://www.totradio.com/mpx/torturecd/f ... _wave.flac

_________________
by GAP
"Less is More" (Bob Katz)

I tried the track of "torture" with the 6.21 version without the fix, with my preset (download/file.php?id=400) and resolves without major problems the AGC, but the fix is much better.

_________________
by GAP
"Less is More" (Bob Katz)

Will be a lot better in the next BETA... Thanks to input from Bojcha I found an extra filtering method that helps a lot here.

Hello Hans and Bojcha - thanks both, I will try it out tommorow. I had meant to post some example tracks, so sorry for not doing so.

I had found (as of a couple of Betas ago) that sensible changes to Advanced Extreme Highs Protection made little difference to the problem.

A good example track I can suggest is Anne Murray - A Little Good News. It's not an excessively bright track, and has reaosnable dynamics. Especially about 27s immediately around the vocal "some senator was squawking about the bad economy". Lots of hole punching there around the vocal, using the preset FM TRANSMITTER SOFT US STEREO. Not surprising that the steeper 75us pre-emphasis curve better demonstrates the effect. The only way I got rid of a lot of it was a combination of reducing the top few compressor bands eq, and dialing a fair amount of clipping in the compressor, again more heavily on the top 3-4 bands. But it was far too much to think about putting on air.

Playing that track through BBP in the non-pumping presets Zenith, CGSmooth etc, no hole punching, even with the Rock preset which has a lot of top end boost.

Later I did find a ST user preset that shows none of this pumping; Bojcha's x_test_orban_v08. This preset (by purposeful design) clips the audio a fair bit, especially the top end. I can't use it on air, but negligible pumping to be found around that part of the Anne Murray song - it's very good. But as soon as you choose a more open preset, the problem returns. It's clear the main ST limiter/clipper is capable of processing top end audio exceptionally well, as apart from the hole punching effect, it's easy to hear that it passes and processes the top end extremely well, and notably better than some other leading processors

I gather that track maybe similar in effect to what Bojcha sent you, but hopefully the Anne Murray track maybe another useful quick test before you release the next beta. Thank you both for looking into this - I greatly look forward to hearing it! Cheers AJ