Stereo Tool
https://forums.stereotool.com/

Low latency secondary input (microphone)
https://forums.stereotool.com/viewtopic.php?t=853		 Page 63 of 75
Author:  hvz [ Wed Mar 31, 2010 1:27 am ]
Post subject:  Re: Low latency secondary input (microphone)
Quote:
yes :)
In breakaway not all filters are turned off in bypass mode, like that phase rotating.

I am getting all songs fixed too with iZotope Alloy VST plugin. In limiter sectiong there is Phase rotating 0-100% and best is at 98%
Does that fix the vibration/distortion when bass sounds are present, if you feed it through Stereo Tool afterwards? I had the impression that I needed to get rid of the spikes altogether, but in the image that you posted I still see those spikes, only now they are on both sides. (Which should at the very least be _better_, but it the distortion really gone this way?)

If this is really enough, then it should be possible to fix it the following way:
- Use NO phase rotation by default
- At very short time intervals, check if the high frequencies are off-center
- If so, at those places introduce some phase rotation; return to no phase rotation when the high freqs are ok again
Which should be relatively easy to implement.

Alternatively, I could measure the minimum spike level (top or bottom) and clip the highs at that level, so spikes on one side will be removed. That's basically what v4.22 did, but I think it was often filtering too strongly.

Edit: Or even better: Do that, then measure back which frequencies have been reduced, and then just reduce those frequencies. That might avoid causing distortion, while still removing (most of) the spikes.
Author:  Bojcha [ Wed Mar 31, 2010 1:30 am ]
Post subject:  Re: Low latency secondary input (microphone)
good point, i'll test it!

I don't know how hard is this to implement it .. but this VST plugin fixes all this phase errors.
and in this song James Last - Dance The Night Away (which is worst what i heard) there is no distortion anymore )

wait .. i'll upload ....
uploaded: http://www36.zippyshare.com/v/84200784/file.htm
You can try both with ST.
Author:  hvz [ Wed Mar 31, 2010 1:41 am ]
Post subject:  Re: Low latency secondary input (microphone)
Amazing! I'm sold! The difference is really huge!

Ok, now I know how to solve it :-) Thanks!

(This also means that my understanding of the cause of the distortion was completely wrong - so I was looking in the wrong direction. It wasn't spikes per se, but the fact that there's an offset from the center.)
Author:  Bojcha [ Wed Mar 31, 2010 1:51 am ]
Post subject:  Re: Low latency secondary input (microphone)
:)
Also .. i think there is no DC offset anymore caused by loudness, but i am still not sure.
So, where will be best place for this filter.. i think just before Azimuth .. after AGC (?)
Author:  hvz [ Wed Mar 31, 2010 1:58 am ]
Post subject:  Re: Low latency secondary input (microphone)
Quote:
Phase rotating 0-100% and best is at 98%
Woohaha! Oh man. This filter is already present in Stereo Tool! And yes, if I turn it on it does indeed improve the trumpets a lot.

Ok, now you're probably confused so... Let me explain:

1. You just triggered me by taking about a constant phase rotation. Before I thought that every frequency needed a different rotation.
2. Here we go: Enable 'Stereo Image' in the Stereo window, set (keep) everything at the default values. But set 'angle' to -84 degrees. Now the RIGHT channel is rotated 84 degrees (I'm assuming that 100% in your VST plugin would mean 90 degrees). And the trumpet on the right channel sounds much stronger and has no distortion left. If I turn everything else off, the spikes are going in 2 directions now. (Left channel is unchanged).
Quote:
So, where will be best place for this filter.. i think just before Azimuth .. after AGC (?)
I don't think it really matters. Just before Final Limiter could already be good enough, but actually it can be placed anywhere (it has almost no effect on the other processing steps).
Author:  Bojcha [ Wed Mar 31, 2010 2:13 am ]
Post subject:  Re: Low latency secondary input (microphone)
lol
First i think it changes only in that direction (angle) .. but i "flip" same song and play again .. and again is in center :D yeah
Not so good as this plugin but really really close.

btw .. angle around 85-105, not same for all songs.

edit:
Further testing shows...
This phase rotating makes more peeks and less RMS. My oppinion is that place for this filter is somwhere before multiband,
Also i think bass in some cases is bit "phater" which is also good.
One bad side is i notice slighly more voice vibrating when specific low bass is present.
Author:  hvz [ Wed Mar 31, 2010 3:48 pm ]
Post subject:  Re: Low latency secondary input (microphone)
Quote:
lol
First i think it changes only in that direction (angle) .. but i "flip" same song and play again .. and again is in center :D yeah
Not so good as this plugin but really really close.

btw .. angle around 85-105, not same for all songs.

edit:
Further testing shows...
This phase rotating makes more peeks and less RMS. My oppinion is that place for this filter is somwhere before multiband,
Also i think bass in some cases is bit "phater" which is also good.
One bad side is i notice slighly more voice vibrating when specific low bass is present.
1. If you use the same filter 4 times you should get the original signal back. (If you run it twice, it's inverted). That's for 90 degrees.

2. About the more peaks and less RMS: Is that with my filter (Stereo Image - Angle) or with the iZotope filter, or both? This would confirm my initial idea that I should only use this filter when it's needed.

Actually this would make sense - at least for already compressed songs you'll definitely get more peaks this way. I'm not sure though if that would translate into lower output levels after Loudness. If songs are already processed such that the phase rotation is optimal, then you would indeed loose some RMS volume by performing an extra rotation (because you're moving away from the optimum).

I Googled for this iZotope plugin, they are talking there about 'dynamic phase rotation' - does dynamic here also mean that they do it when needed, or just that you can dynamically set the rotation level?
Author:  Bojcha [ Wed Mar 31, 2010 4:59 pm ]
Post subject:  Re: Low latency secondary input (microphone)
1. yes.
2. Happens with both but bit more with ST.

'dynamic phase rotation' - I am not shure but i have same feeling that something there is dynamic or "when needed" or combination of both, in ST is more static. This maybe explains bit more peeks with ST.
Author:  hvz [ Wed Mar 31, 2010 5:17 pm ]
Post subject:  Re: Low latency secondary input (microphone)
2: ST is completely static (that was the purpose of this filter!)

Anyway, if it happens with both I should just avoid adding a phase shift when it's not needed. Hence I'll go for the dynamic solution. (A small disadvantage is that at the start of a trumpet-like sound, the first 100-200 ms might still distort a bit. I'll have to see how fast I can move between phase shift and no phase shift without getting artifacts - and I think it won't be very fast. When I detect that there's probably really bad distortion I can of course move a bit faster.)
Author:  hvz [ Wed Mar 31, 2010 6:53 pm ]
Post subject:  Re: Low latency secondary input (microphone)
It's not the static part - the waveform that you've mailed me looks really better than what Stereo Tool's "Angle" filter makes of it.

Basically, Stereo Tool divides the spikes in an upper and a lower part, which are equal in height. So the peak level is divided by 2.

In your waveform I see 2 spikes on one side with a dip to the other, essentially dividing the peak level by 3. Which of course sounds better.
Page 63 of 75 All times are UTC+02:00
Powered by phpBB® Forum Software © phpBB Limited
https://www.phpbb.com/