Stereo Tool

Independent of the question if my approach to use identical (!) downspeed settings for all (!) bands is useful or not, it has the benefit that adjacent bands use exactly (!) the same down speed (during editing the bands were linked in the MB GUI).
When I started with Stereo Tool I thought that this linking option in the GUI was only meant to have an option for faster editing.
I was sursprised when I found out that this linking option also has an influence to the characteristic (even if all bands already use identical speeds it can make sometimes an enormous difference whether a linking option is on or off !).

Comparing the sliders in the GUI the speed sliders have the most extreme fine parameter resolution.
Some might think that the changes using the mousewheel are too inefficent.
But I think that this fine parameter resolution is very useful.
Especially in conjunction with our shared approach to use/try down speeds close to zero.

If you adjust/test more of that concept, you will notice that sometimes even minimal changes can have a very deep impact.
IMO it could be useful/necessary for you and everybody else that use this concept only for two or three adjacent bands, to edit STS files manually.
Sometimes looking at the MB GUI it seems as if some speed sliders have identical positions/values, but it could be false positive !
Especially in conjunction with the stable voice problem it could be very important to use identical (down) speeds for the adjacent bands.

I have experienced sometimes enormous difference for the overall sound (voice) stability if I tried minimal changes/differences for adjacent (mid) bands.
So IMO it could be necessary for you to precisely adjust/synchronize those down speeds manually in your STS files.
Especially adjusting the speed sliders in that blind everything-is-zero range in the GUI reminds me of analog radio/synthesizer tuning.
In that context it means that using a GUI mimics an old-fashioned analog principle of calibration.
It could be a very fast editing process, but sometimes it leads to sub-optimal results, because you have not the digital precision.
So at the moment (the actual GUI behaviour) IMO it is necessary to combine both editing concepts for the best possible results:
Fast analog type GUI editing plus slow but precise digital STS editing.

I hope that the new GUI will have a much higher parameter resolution for EQ.
We could use already smaller steps via manually edited STS files, but then we lose control of the whole MB GUI (any further adjusting of any value in that sub-GUI would reset everything to the limited standards of GUI).

Yes it will have much more control (including compress vs. limit and more) per band.

@michi95: That 'link' checkbox really only affect the GUI...

See your PMs for discussion about what to do with this preset. Let me know and I'll proceed accordingly.

Updating preset for inclusion into 6.31 bundle.

This was a blend of the soft limit to 10 and downspeed to 0 settings that I had been working on. What's different is that I made the decision since the near-zero settings on the first 4 bands and the last band were non-standard edits of the STS file rather than using the GUI, I set the first 4 and last 1 bands' down speed to be the lowest GUI setting of 0.0001.

The primary benefit of doing what I've done here is that it almost completely resolves the issue of the mids / vocals from being dropped too much in certain tracks. The other things that got impacted by the mid dropping were synthesizers, and guitars for rock/metal, and possibly some strings/brass/woodwinds in classical music.

Also, due to not using the slower down speeds for bass, I left the deep bass boost in loudness enabled to partially compensate. I have discontinued using the Bass Boost filter entirely.

Beyond that, I've updated the strictness setting to -2 (down from 0), which is midway between default and the previous setting. I also enabled the bass shape at .50 strength, but max harmonics up to 200 instead of the default of 160, with the primary purpose of getting both 4x50 and 2x100 harmonics.

@Hans - take a listen, and then include whichever you like more. Use the name Purity Control, without the R2 (Rev 2), if you do decide to pick R2.

I think I want to change the Deep Bass Boost to at least 50% instead of the current 33% in R2, maybe as much as 75%. Not using the slower down speeds has cut into the bass a bit more than I first thought.

Thanks! I've downloaded it and increased Deep Bass Boost to 50%. Will compare the two versions when I get in my office tomorrow.

You do know that nearly every value is changed in this update??

If you're talking about me changing "nearly every value" in the preset, yeah, I know...

I changed my mind after having more sleep, and decided to go with Option 2.5, a blend of options 2 and 3.

There was no real reason to withhold the changes that I made to multiband going down to zero in the GUI. Those are absolutely zero, both in the GUI and in the underlying STS file, and the changes were made to address mids and the supporting highs for vocals from dropping too much.

The first 4 bands and the very last band is where the settings looked like zero, but were not. So, I decided to make them non-zero in the GUI, but the lowest non-zero allowed by the GUI, which is just barely .0001.

The only "problem area" is really the bass, and really the 100-440 area. I increased the soft limit values to the same soft limits I have in the near-zero down speed settings, but because the down speeds are higher, at .0001, I'm getting a bit more compression, which I had hoped that the deep bass boost and additional harmonics would offset enough.

R2b

Changes from R2:

- AGC up push strength changed from 0.4 to 0.5

- Moved 4th band (220 Hz) to soft limit of 10, down speed of 0. I was very close to this anyway with the near-zero preset development.

- Increased bass EQ settings slightly

- Altered multiband clipping settings slightly

- Increased multiband steepness to 150%

- Changed Final Clipper (I still consider it "Final Limiter") to 64%.

- Deep Bass Boost is set to 50%.

- Noticed that the two "time spread" settings were set to 0, so I set them to the default of 20%.

The end result is PCR2/R2b is about the same total power output, just distributed slightly differently, with R2b having a bit more bass.

R2b is 0.3-0.4 dB louder than the original Purity Control preset.

For tracks like Alejandro and others that have clipped input, enabling declipper does yield a bit better bass output. I don't enable it for a preset because what I make is geared towards non-commercial users. If commercial users want to use what I'm making, I'm greatly flattered, but they can also feel free to enable the paid-for items (declipper, more loudness, FM)...

A good track to test with is Neodammerung and/or Navras, from The Matrix:Revolutions.

The mids still drop some. This is not a perfect solution, but it's getting there. I could come up with a better solution, possibly, with some more time, but I guarantee I could come up with a better solution with a redesigned multiband.

Fundamentally, right now the AGC is acting as the primary setting, with multiband being used to fine-tune. Due to the limitations with multiband, I have not yet been able to fine-tune enough. It may be possible within the current design, as the last "aha!" moment I had was after posting the "frustratingagc" preset, when I realized that pushing more power through the AGC to multiband, then reducing the total ouput via the Final Clipper (Limiter) reduced/eliminated the mids problem.

Well, I did some tests, and the results vary. The old version has more highs and sounds a bit brighter. The new one sounds more open and dynamic / less smashed (might just be the extra bass, but it could also be what you are describing).

If I had space I would add them both, but - well you know the drill, new interface needed etc.

I think I'll go for the RC2 version, also because to me it sounds as if the quality is better (as if I'm listening on a better audio system, dunno how to exactly describe it).