Stereo Tool

Soon after the purchase you came up with beta 101, so I guess it's something to take pride upon.

I am a little stuck at MB calibration. So per your previous reply, I am doing it with single band AGC.
If I am not mistaken, when we are adjusting MB for flat output, the peak levels for all center frequencies should be same, right?
But on the other hand median values, if made consistent for all bands, merely represent consistency in compression ratios. How does that account for consistent peak levels for all band after MB?

Something very important for my work...Is there any way to establish a mathematical co-relation of attack / release time in milli-secs with attack / release speed scale of Stereo Tool's MB? I tried pouring over previous posts for any discussion of this order, but no results. If there's such a relation, it would be helpful for many of forum members to render desired signature to the mix.

EDIT: Please ignore the first question. Think I got my answer. It's just that for a given AGC level, the peak levels of first 4 bands are different than the remaining 6 bands despite setting the same attack(0.0005) and release(0.00005) speed for each one of them at soft limit threshold of 5. Bass band and mids show higher peak levels than the rest. Remaining bands are consistent at the same peak level.

_________________
visit website

Perhaps time to revisit that preset I sent you that uses the multiband clipper on those first two bands.

I made a further change to the clipper that seems to move further towards the sound of clipping without protection - on top of other things, the mid frequency volume drop when loud bass sounds occur seems to be reduced by about 1 dB, the maximum bass peak level is also slightly higher. BUT that's only the case with bass clipping set to 100% - the problem is that the bass band now runs upto 400-700 Hz by default instead of 150-250 Hz.

I'll just upload this new version anyway for testing purposes and to get feedback, but remember that it works best with the bass clipping level set to 1.0.

What I think I need to do (but it's too late now - already past 4 AM and I *really* need to get to bed), is to clip the bass band that is clipped at 100% further at a lower level, and then only take the lowest part of that. Hm, actually that's getting pretty close to what the original filter did - but a lot better. Will try to do that tomorrow.


Furthermore, in this version the Loudness display is fixed, and the RDS code has been cleaned up (it's now also in the Omnia 9).

Is it me or is something wrong with the Windows command line version? I've tried several from 101 to 096 and they all crash. I'm running Win7x64. I have an older beta (033) and that one seems to work okay.

Thanks!

Hans,

I don't understand one part of the Advanced Bass Protection. It says "Clip from 0 upto (250 Hz)" and then it says "then drop to no clipping at 400 Hz".

What happens to 251 - 399 Hz?

The amount of clipping drops. Things will change in the next beta (now building), so you might want to wait for that version now.

NOT FOR BROADCASTING - currently only works well at bass clipping level 1.0!
Winamp DSP plugin: http://www.stereotool.com/download/dsp_ ... 02TEST.exe
Stand alone version: http://www.stereotool.com/download/ster ... 02TEST.exe
The command line version is indeed crashing and the VST version didn't build after all the RDS changes.

Anyway, here's the story about the new sliders.

The problem that occurred in previous versions was this:

Say, you have a bass tone at 100 Hz and a harmonic at 300 Hz. Say the 100 Hz tone is at 300% and the 300 Hz tone is at 50%.
Now by using 0...150-250 Hz as bass band, and clipping it separately, the 100 Hz tone gets dropped to AT MOST 100%, while the 300 Hz harmonic is untouched - still 50%. This will sound very crappy; basically what it comes down to is that the 1st harmonic is 3 times louder than before, relative to the main frequency. And it gets worse: When the two are added the harmonic will push the original bass sound down even further.

The solution: A large overlap area between the two bands, which is partially ignored.

At the default settings, this version does the following:
- Take bass from 0...700-1200 Hz
- Clip
- Take the part from 0...400-700 Hz

- Take mids from 150-400 Hz upwards
- Clip
- Take the part above 400-700 Hz

This allows far less sudden changes in response between frequencies, and hence stays much closer to the original without protections.


What's wrong here? Well, if you now set bass clipping to - say - 70%, then the whole part upto 400-700 Hz gets reduced. That's now what we want. So I'm going to add an extra bass clipping stage where the whole 0...400-700 part gets clipped again at a level below 100% (first step will be AT 100%), but only the lowest part of the stronger clipped result will be used. This kinda resembles what I had earlier in the old protection filter (2 different bass bands) - I might even add a clipping level for the 0...400-700 portion just like before, but this should work better.

What part is now also in the Omnia 9? Is this the Always clip bass below slider?

RDS encoding. [/quote]
Yes. And also the dynamic reduction slider.

That might work good, with one change. Since you first clip it, then you want to clip it again for final level... do not clip it, just LIMIT, fast as possible. That will not ruin bass buy two times clip, it's already clipped once. Later will test b102.

For now, beta 101 works very well for L/R output. With FM still testing...