Stereo Tool

I hope you talk about AGC1. i'll wait final version and make changes to preset.

- Gating 2nd and 3rd band in AGC,
- gating multiband (?) not answered. Why not?
- hmm .. louder Highs in V5, talk from erlier ..- last 2 or 3 bands in multiband need better HF Clipper.
- ITU from AGC, cuts loud highs slow-back! slower then when is off, loud "S", loud instruments on high bands, Don't know why.
- "Hard Limit" knows knows to do same! too slow-back (check songs with loud "S" and only hard limit ON). This is from long time ago.

Yes, both AGC's actually. There was a bug in the processing (which caused the band 2 volume to drop too much in case of loud bass, Beyonce), and I've changed the up speed behavior to be independent from volume levels.

Unless I receive complaints about how it works now, I think it is finished (I haven't made any changes to it in the last week or so, only cosmetic changes such as the different values that are displayed now).
Should follow band 1, but I'll check if it really does TODO Not going to do that for now, maybe later. I think there are more - and more important - improvements possible in multiband. (Here also the up speed depends on volume!)

In fact, I think I can improve Multiband such that you can use much higher input volume levels for Multiband - and if that's indeed the case, the quality of the AGC will be less important. And (other remark) loud "S" sounds would also be handled (much) better by Multiband!

(Currently, higher Multiband input levels cause a faster 'up speed', so more aggressive compression. So songs with much highs get more aggressive compression on the highs. If I make the same change that I made in the AGC, the input level wouldn't matter anymore - or at least much less - so that wouldn't happen anymore. Sound for high input volumes would get more-or-less similar to that of lower input volumes - although higher input levels would still allow bigger changes between bands to occur.) Hm... Doesn't band 3 fix that? Is that with BETA8 or BETA? ??? It responds immediately, and only has a very short lookahead/lookback time. So the volume increases back within 20 ms or so. (Except for that, HARD LIMIT is not intended to handle loud sounds, so at least Final Limiter or Loudness should be enabled to get a decent sound out of it).

Darn that song!

I just cracked out "Abba Gold" and... that's irritating! Now I have a new test track, though

This is for Bojcha QA (Quality Assurance).

I've checked all questionable songs I have in my library, and see no issue with fast drops anymore with this preset.

@ SuperH, hvz, Bojcha and others

Maybe you are talking about (and listening to) apples and oranges.
Just to say "ABBA - Gimme! Gimme! Gimme!" is not precise.
What version (build ? ) of this song ? http://en.wikipedia.org/wiki/Abba_Gold

http://www.imarc.net/communique/403-loudness_war

Somewhere I found this comment about a newer ABBA Gold release:

michi95: I actually had thought about this... What if he had a different master? That was irrelevant I think, as my master was exhibiting the behavior explained by Bojcha. Mine is VERY dynamic.

Remasters (usually) are not good. A notable exception is the recent Beatles remaster.

Michi: Bojcha and I were talking about the original (or at least it looks like that).

The difficult thing is to configure the processing such that the not only the volume stays more or less constant, but also the RMS output volume doesn't deviate too much from other tracks.

Guys ... Unfortunately betta8 test until I am convinced ... Bass excellent! But the mids are not expressive. I was not able to handle the high frequencies that I need ... I'll try to check back in the studio. I still really like working version of 28 May. But there is a slight swing bass frequencies

Achieved by using high-frequency Band 2 .. Not bad .. I would be tested for transmitter.Sound wonderful.

The relation between mids and highs has not changed, and between bass and the rest should not have changed much.

However, the output volume is different (lower) at the same settings. Which means that Multiband receives a much lower input level, and has far less effect on the sound than before.

For a similar sound as before, you'll need to set the target output level about 35% higher! (That's assuming a Pre Amp of 24.00. For different Pre Amp values the percentage is different.....)

Details:
Original (BETA): new_volume = curr_volume * (1 - upspeed) + upspeed
Current (BETA8): new_volume = curr_volume * (1 - upspeed) + (target_volume * upspeed)

Say, curr_volume = .1, target_volume = .2 (input volume is 50% lower than the wanted volume).

Say upspeed = .01

Old version: .1 * (1 - .01) + .01 = .099 + .01 = .109 (9% increase)
New version: .1 * (1 - .01) + .2 * .01 = .099 + .002 = .101 (1% increase)

Now say curr_volume = .01, target_volume = .02 (again volume is 50% too low)

Old version: .01 * (1 - .01) + .01 = .0099 + .01 = .0199 (99% increase)
New version: .01 * (1 - .01) + .2 * .01 = .0099 + .0002 = .0101 (1% increase)

As you can see from these numbers, the increment goes much slower in the new version. In the old version, the increment very strongly depended on the current volume level (at 10% it was 9%, at 1% it was 99% which means that it's VERY difficult to properly configure the up speed).

In the new version, the volume increment is always the same for the same input, regardless of the (absolute) input level.


To get a similar sound, you might need to increase the up speed quite a lot though, or, if you don't want compression (and for an AGC you probably don't), increase the target output level to compensate for the slower volume increments.