Stereo Tool

Is there a way to use Stereo Tool with multiband expander functionality ?

From my naive point of investigation I think it is not possible to do it.
Or have I overlooked something ?

There is the "Multiband compressor / limiter" section.

But it is not possible to use negative ratio values
(would be necessary to use it as an expander !? - correct ?).

Hans if you find the time, please add a multiband expander functionality to this section, to make Stereo Tool complete.
It should not be that difficult to add (to program) negative ratio range.
Instead of "Multiband compressor / limiter" section it would be fantastic to have a "Multiband compressor / limiter / expander" section.

Now I am using AudioProc - the multiband expander only.
It works the way I need it.
But:
Stereo Tool overall audio quality (except the missing expander functionality) is much better than AudioProc, which comes with terrible presets, so that I already was close to deinstall it, before I started from scratch (multiband expander only).

And AudioProc introduces extreme latency and is limited to 44,1 KHz processing.
That's why I have to reduce the sampling frequency from 48 KHz (german DVB standard) to 44,1 KHz to use it for my DVB-S Radio (de-)processing.

So it is really ironic, that the professional Stereo Tool is almost free for home users (without broadcast purpose), but this old fashioned AudioProc is $39.99 USD.
Hans you are to good for this world !


Why do I (we) need expander functionality ?
I finally came to the conclusion, that there is already to much compression, limiting terrorism (using other processing tools than Stereo Tool ) in this world today.
From the point of view of an audiophile 99 percent of all radio stations (streams) are absolute junk.
http://www.dynamicrange.de/

@ all you radio guys
Please don't use StereoTool as a weapon of ear destruction and be aware about the dynamic range (RMS) you are using.
There are still some persons that do not use cheap PC desktop speaker systems !

off topic:
I hope that all these spammers like Votlooste (the persons behind) go to hell !!!

Hi michi95,

Does that really work for badly compressed audio? I have thought about making something like that, but I never did because I thought it would be completely useless.

Looking at compressed sound, I cannot determine whether the input volume was '10' or '100' - it always comes out at the maximum volume. I can make soft sounds softer (or loud sounds louder), but does that really improve anything?

First of all I have to say:
It is absolute impressive how fast you respond to almost every question in this forum.
This behaviour is outstanding
Besides Stereo Tool I use (know) many free VST plugins and also try many trial versions of commercial plugins.
There are many people (even audio engineers) who do not believe in dynamic expansion of a complex (!) mix.

During the last months (almost a year) I have done serious investigation to this topic and now I know what is possible and what not.
I am using a special hosting setup (including wrappers) with two VST plugins instances of the TT DR Meter sandwiching all other processing on-the-fly while playing all kind of music sources.
This way I hear and see what I am tweaking.

So I don't need this multiband expander option desperately in Stereo Tool.
But it would be easier and would open a door for me to create some usefull Stereo Tool presets for other (home) users to "reduce the pain" of many radio stations and streams.

When I have more time, I will write a detailed guide for chaining VST Plugins (all freeware !) and will show some audio examples.
Of course it is not possible to rebuild an original dynamic range (that would be magic).
But it is possible to redefine the dynamic characteristics for many (not for all and very extreme) sources.
From narrow limited back to a vivid dynamic punch (without pumping).
It is not simple, but I have spent hours of trial and error fine tuning chains of plugins and finally found a good approach to this kind of audio reprocessing.

Every limiting and compression have an impact to the spectral balance.
Though the results of processing can sound very different there is a characteristic spectral fingerprint to do some kind of "reverse audio engineering" !

Interesting... To 'turn' the behavior, I could still determine all the levels as I do now, but instead of multiplying by them, divide by them. I'm not going to add this to the next version (I wanted to release it in november and I'm still working on it ) but I'll do some tests when I have some time left to see if I can get this to work, and the results are usable.

Here is a little preview of my dynamic expansion setup.
It is still work in progress:
Wrapper's Delight !

The left TT DR Meter shows the signal before all processing (red range !).
The right TT DR Meter is placed at the end of all processing (green range !).

Companding is something I've always done in hardware to a degree. This is the principle I was trying to explain with two AGCs, with your top-down (compression) algorithms.

This is a combination of compression and expansion. This is very useful around the AGC (I used to force this effect with Stereo Tool's 5.10 AGC) especially. "Diff" can be used to expand to an extent with Stereo Tool.

You set a threshold, and have a level, Vth = _dB.
If you exceed that threshold (above), you attack (compress) at rate _dB/sec. If you go below that threshold, you attack (expand) at a rate of _dB/sec. This rate is logarithmic, of course, based on your outer limits (Vmax = _dB and Vmin = _dB). Exceeding those limits will only allow attack (compress or expand) at the rate which you have set at your max, which are configured by attack time.


Example
So, Vth=-18dB, Vmax=-10dB, Vmin=-26dB, RateAttackCompress=-6dB/sec, RateAttackExpand=6dB/sec. Any input above -10dB will attack at a rate of -6dB/sec, and if it drops between -18 and -10, it will attack with a reduced function (inverse), at a magnitude less than -6dB/sec. Any input below -26dB will attack at a rate of 6dB/sec, and if it elevates between -26 and -18, it will attack with a reduced function (inverse), at a magnitude less than 6dB/sec. This principle is often applied to multiband as well.