Stereo Tool

This is a release candidate for V6.

The sound is very similar to the preset that Hans bundled with the 6.20 release, but it is around 1 - 1.5 dB lower in RMS, with a corresponding increase in Dynamic Range. Voices should be louder than before, as I backed off of the bass a little bit and increased mids.

I've got a few more test versions to go through to see if I can increase Dynamic Range a bit more, but the overall sound is likely going to remain the same, unless someone finds an issue with it.

Final (unless anyone finds a glaring issue) added.

Added the final release version.

Inside the zip file are both the Generic Preset and Web Preset versions.

As far as the sound profile, the difference between RC1 and the final is that I added a slight increase to highs to help the resonant frequencies / harmonics / breath of the increased mids in RC1.

Overall RMS and Dynamic Range is relatively unchanged, at +/- 0.2 dB from RC1 (typically -, but in any case, it's relatively unchanged).

Fair warning:

Previous final was not-so-final...

I'm almost done. The sound is better, in my opinion. I should be posting the "final final" within the next day. After that, I think any changes will have to wait until Hans redesigns multiband.

Mostly-final Final added. It will be final unless anyone can find a glaring issue. I don't think I can make further meaningful changes without changes to how multiband functions (RMS vs. Peak).

Basically the changes made were:

- Reduce highpass to 30 from 35.
- Change 70 Hz multiband band to 80 Hz.
- Soft limit changes to multiple bands.
- Slight downspeed increase on 40 Hz and 80 Hz bands.
- Matched upspeed for 40 Hz and 80 Hz.
- EQ changes to multiple bands.
- Multiband clipping changes to multiple bands.
- Deep bass boost in loudness to 35% from 33% (applies to web version only)
- Very slight increase in de-essing (applies to web version only)
- Increased peak detection steepness for mids to 0.20% (applies to web version only)
- Deform bass peaks to protect highs increased to 133% from 125% (applies to web version only)

Feedback is welcome, and I'll make a Final Final Final if needed.

To give an idea of the difficult tracks I've thrown at this, here's one Youtube video of one of the tracks:

DJ Orkidea - Metaverse (Gareth Emery Remix) http://www.youtube.com/watch?v=5mAaalEtuxU

What this does not have is the declipper enabled. I've made sure to keep these presets available to everyone. You might get a bit better performance if you enable the declipper. I can't judge that too well, because my system is too slow to keep the declipper on.

Other tracks:

Pink Floyd - Time
Pink Floyd - Echoes
Jimi Hendrix - All Along the Watchtower
Jethro Tull - Locomotive Breath
Van Halen - 1984 (entire album)
George Michael - Faith (entire album)
Alicia Keys - Karma (both original and Karmastition remix)
Blue Oyster Cult - Don't Fear the Reaper (you can make out the cowbell if you try)
Lloyd Banks - Rotten Apple
Rob Dougan - Clubbed to Death (Kurayamino Variation)
Prodigy - Mindfields
Rob Zombie - Dragula
(noticing a Matrix theme yet?)
Marilyn Manson - This is the New Shit
Juno Reactor vs. Don Davis - Mona Lisa Overdrive
Rossini - William Tell Overture (performed by The Boston Pops)
Enigma - Back to the Rivers of Belief
Information Society - Running
B52s - Loveshack
Enrique Iglesias - Tonight
Frankie Goes to Hollywood - Relax
Scorpions - Still Loving You
Talking Heads - Burning Down the House
Metallica - One
Q Lazzarus - Goodbye Horses (extended version) <--- put the lotion in the basket!

...and a slew of others.

For me it is still very irritating that especially for songs with wide dynamic range like "Scorpions - Still Loving You", etc. the processing more or less completely inverts the original dramatic mixing conception and leads to a surreal mismatch.

But this critic is not exclusively addressed to your presets.
IMO this happens with almost every Stereo Tool preset (except my presets ).

Comparing many Stereo Tool presets with Breakaway or even with FM broadcasts it is more or less impossible to find a similiar artificial processing behaviour (with that extreme degree of dramatic inversion).

IMO most presets use too much compression in Multiband (you know I prefer limiting) and more or less almost no relevant audio expansion (minimal using of Noise Gate/Expander module).

The great thing about Stereo Tool is that everybody can tune the settings in every detail to match individual taste (listening experience).

So there is no problem at all, because we can agree to disagree.

The main advantage of Stereo Tool is that is the best choice to control and balance the spectral characteristics for very different sources.
You cannot do that with Breakaway Live (you can only try to compensate its general artificial highs with a post EQ).

OT:
But I totally agree with every word you have written in this topic:
viewtopic.php?p=14106#p14106
Especially this statement: About that Bill Gates population growth/reduction debate:
It is a well-known phenomenon that people with a very brilliant mind (like you)
and a straight logical diction get often misinterpreted.
But you convinced me absolutely with every detailed argument (because I am a brilliant mind too ) !

Are you listening from a CD, or from a MP3 / FLAC? I have the CD somewhere, but what I have been using is a 192K MP3.

The only thing I can understand from what you're saying, based on what I hear, is that the vocals may be at a slightly different spot in the soundstage. This is directly related to me using the 1990's style stereo boost. Beyond that, if I have Breakaway RTA running while listening, the processed signal is only fundamentally different with respect to the volume (RMS) being higher and a little more bass. Fundamentally the shape of the graph is the same, at least on my system.

When comparing your bundled preset, the vocals are brought forth into the foreground more than what they are when I bypass processing, and your percussion is more muffled, particularly the intial impact of snares, hi-hats, and ride cymbals, especially when adjusted for the total RMS level. From looking at RTA, it seems as though your lower vocal range, from 150 to maybe 600, is too hot, which then suppresses the upper-mids and highs. I looked at your multiband settings after typing this, and it only reinforces what I just said. Going with that low of a soft limit on the highs and attempting to compensate with aggressive upspeed leaves the mid-to-upper midrange and highs suppressed, amplifying the issue with the overstated vocal range.
This is another fundamental misunderstanding you have. Limiting IS compression. It is simply a different type of compression. What limiting has is generally a higher ratio than "compression", thus squashing off more sound overages to maintain a consistent total volume. It can actually lead to a LOWER dynamic range.

To give you an idea, when I run my 192K MP3 though Winamp's disk writer, then use TT-DR Offline, using the last Web version preset here, I get:

Dynamic Range = 8.6 (L & R)
RMS = -10.9 L / -11.0 R

With your preset:

Dynamic Range = 6.9 L / 7.0 R
RMS = -9.6 L / -9.7 R

What I saw when I reduced RMS is I had a consistent 1:1 ratio on gains in Dynamic Range. If I assume that to be true when adjusting your preset down to my RMS, that would add 1.3 to the DR, yielding a final DR for your processing matched to the same RMS of 8.2 L / 8.3 R, which is still slightly less than mine.

I see what you're saying, but the data does not support what you're saying. It could very well be source-specific, I don't know. In any case, processing the same source adjusted to the same RMS gives a higher Dynamic Range with mine (although not by much), as well as a crisper sound.

Bottom line here is you like mids more than highs. You like the vocals to be warm and in the foreground. That's ok. I'm ok, you're ok.



As for the OT, thanks... Don't want to say much more about it out of respect though.