Stereo Tool

How about command line version?

These changes will be available in the next beta:
- Fixed bug in previous version (startup in 2048 mode, then switch to 4096, Loudness sounds horrible)
- Removed Alternative mode
- Added 2nd slider for time spread - for testing purposes
- Bass shape now also works if you turn Advanced Bass Distortion Protection off; moved sliders to the left part of the panel.
- Stereo Boost disabled for band 1 (bass, below approx. 140 Hz) - hint from Bojcha
- Icon visible in stand alone version
- De-esser range extended, highest value = OFF.

While this is all good and stuff, mids and highs are getting squished by... :drumroll:

...the AGC.

If someone would like to explain to me the "proper" AGC settings so that a track like Kelis - "Brave" doesn't have the synthesizers and vocals significantly reduced in places, I'd be very grateful... Here's a hint: Band 2 MUST go above band 1, but the AGC logic clamps down too hard on Band 2 in those specific areas of the track (Kelis - "Brave").

Here's even a huge head start for anyone wanting to try. Try this preset, as it is in the file, and then reduce preamp to x1.0 and turn off the AGC, and notice how everything stays up...

The biggest polluters of the sound are AGC and then multiband. I could work around the problem with the AGC, if multiband let me take some bands above a "10.0" soft limit WITHOUT over-EQing.

Edit: Oh, and the extra-loud presets out there don't have to deal with this, because they're extra-loud in the mids and lacking in the bass that I'm going for, so, no, Bojcha, you probably never run into this because of the high RMS you use.

Edit 2: Actually, Bangkok Dynamic does suffer from this same issue, although I think that has lower power than another test preset that was posted.

- Fixed bug in previous version (startup in 2048 mode, then switch to 4096, Loudness sounds horrible)
- Removed Alternative mode
- Added 2nd slider for time spread - for testing purposes
- Bass shape now also works if you turn Advanced Bass Distortion Protection off; moved sliders to the left part of the panel.
- Stereo Boost disabled for band 1 (bass, below approx. 140 Hz) - hint from Bojcha
- Icon visible in stand alone version
- De-esser range extended, highest value = OFF.

Stand alone: http://www.stereotool.com/download/ster ... 31-012.exe
DSP: http://www.stereotool.com/download/dsp_ ... 31-012.exe
VST: http://www.stereotool.com/download/vst_ ... 31-012.dll

+1. Agreed! Bass improved.

TODO's:
- check sound with bass/mids/highs distortion > 0 - Seems ok?? There was something weird going on between 200 and 800 Hz - O not just that. Problem at 500-800, 3000-4000 and below 400 if voice vibrations filtering is used. FIXED
- remove 'alternative method' from clipper code
- compare cpu load against 6.30 - A few % higher with the same settings; with new 6.31 settings (certain filters disabled, clipping less strict) it's now a bit lower.
- smoothe_mids_width behavior does not make any sense. Changed behavior for lower latencies. Bojcha's proposed new settings doesn't make sense, have to ask him what he's doing... Especially to set this thing at '1' seems far too low - 2 MIGHT be ok... - no, replace by 4/.30!
- can 'Allowed highs distortion - strictness' be reduced to 1 now?
- update presets
- change version #
- release

I think I've figured this out. Still testing, but basically, it plays out like the Star Trek: Next Gen episode "Booby Trap". Had to look at things differently...

The bass is good!

I noticed that in my notebook (AMD L325 dual core) CPU usage per core is uneven, usually consumption in the core 1 is twice the core 2. This enabled the Multicore option, this is correct?

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

Hm, interesting. I've tried what would happen if I remove *all* traces of 'alternative mode' from the clipper. Earlier that seemed to have a bad effect (lower output level). But now it seems i'm getting both louder output and stricter clipping... (0.03 dB difference). Which is not that interesting, but it also (slightly) lowers the cpu load and code complexity.

Ti maar sure that there are no bad effects i'll build 2 different versions tonight.

Some minor bug fixes in the clipper.

Normal clipper:
Stand alone: http://www.stereotool.com/download/ster ... 31-013.exe
DSP: http://www.stereotool.com/download/dsp_ ... 31-013.exe
VST: http://www.stereotool.com/download/vst_ ... 31-013.dll

With 'alternative method' stripped completely:
Stand alone: http://www.stereotool.com/download/ster ... 1-013A.exe
DSP: http://www.stereotool.com/download/dsp_ ... 1-013A.exe
VST: http://www.stereotool.com/download/vst_ ... 1-013A.dll

I'll run a longer comparison between these two later tonight. Expectation: 013A is clipped slightly tighter, no noticeable audio difference.
Edit: First measurement shows: With Hard Limit on, the RMS level of 013A is 0.02 dB higher... Still need to check audio.

Edit: No big difference. Compared CPU load of 6.31 against 6.30; with the same settings it's slightly (~ 3%) higher, with the new optimized settings (less strict clipping etc.) it's slightly (~ 6%) lower.