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

Stereo Tool 6.22 BETA
https://forums.stereotool.com/viewtopic.php?t=4102		 Page 31 of 39
Author:  michi95 [ Thu Mar 29, 2012 4:36 am ]
Post subject:  Re: Stereo Tool 6.22 BETA
:?:
:arrow:
Quote:
But right now I think I should really focus on the multiband section. And for that I need new parameter handler code. So that's what I intend do first.
:!:
Author:  phoenix [ Thu Mar 29, 2012 6:15 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Looks like I am able to reproduce the issue that probably Bojcha was pointing to. I hear it happening even at moderate loudness but only when bass is very powerful. I don't always hear the distortion with Bass Boost disabled (depends on MB calibration) but once enabled it's quite prominent.
In some cases, even without Bass Boost enabled.

Attached here is the preset and sample file (bass boost disabled). I could hear distortions (even with Bass Boost off).
Take 0:48 ... "Bring back those nights when I held..."
Tested on last beta VST.
Author:  Brian [ Thu Mar 29, 2012 7:43 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Quote:
Looks like I am able to reproduce the issue that probably Bojcha was pointing to. I hear it happening even at moderate loudness but only when bass is very powerful. I don't always hear the distortion with Bass Boost disabled (depends on MB calibration) but once enabled it's quite prominent.
In some cases, even without Bass Boost enabled.

Attached here is the preset and sample file (bass boost disabled). I could hear distortions (even with Bass Boost off).
Take 0:48 ... "Bring back those nights when I held..."
Tested on last beta VST.
Absolutely, positively NOT, at least not with the DSP version, and with no other software being used in the processing chain.

First, the preset came with bass boost off. I enabled it, since you said it was "quite prominent" with Bass Boost on.

Second, if you listen to the track in bypass, then compare it with being processed, all you are really hearing is the same sound made louder. The proportion of increase in volume of that passage seems to be in line with the overall increase in RMS over the unprocessed sound.

Now, there is indeed a higher volume bass note at that point. You can see the clippers in multiband become active in the bass bands. There is also a clap / snap just slightly ahead of that point.

What you need to really look at is the AGC. There is a bit of clipping a few seconds before, while "the nights are so unkind" is happening. The AGC has a look-ahead, so the clip in the AGC is actually the passage you think there is a problem with.

Now, compare this test preset with what I posted for HypersonicV7_Declipper. Compare the difference in what happens in AGC, and also focus on the amount of clip in multiband at that point with my 220 Hz band and your 250 Hz (bass boost on).

What you should see from all of this, is multiband is being driven by too aggressive AGC up speeds. Hans stated some time ago that multiband did not handle higher input levels as well as it should, so why is it some shocking surprise that higher input going into multiband yields less-than-desirable results?

The AGC settings in this (MX 20-based) have a slower down speed and a faster up speed than what you posted in Mirage MX12. You also have set 1 dB additional target level (-6 vs. -7 in MX20), and have lowered HPF to 20 from 25 in MX20. Then there are even more settings changes in Limiting & Loudness, like the Singleband compressor/limiter, a different phase rotator, and then you crank up all those changes by ~3 dB more than MX20.

The distortion you state that you're hearing (IMO, you have golden ears, or the VST version is different, or you have some other processing) has to do with bass. The vocal is actually clean. The problem is that the 100-800 (ish) region is shared with vocals and bass kicks and/or bass harmonics. This is being driven by settings changes.

If you set out to prove that you can change settings and make something distort / sound different ("bad"), then yeah, that's not something I'd dispute. I would however disagree that the product should be able to clean up bad settings to make them sound good. If that is the case, then I should be able to do whatever I want and have it come out sounding like an award-winning masterpiece, and I hope that's not what people are expecting.

You might take a read through EQ by the octave. That is for recording, but it is still somewhat pertinent. The other thing I generally look at is the Interactive Frequency Chart.

Bottom line for me is I still do not hear what you all are claiming is a problem, and that's why I'm so adamant about not spending additional time chasing a potential non-problem, when working on an improved multiband and possibly an improved (less fast rising) AGC would provide far more "bang for the buck".
Author:  DJ-DOGGY [ Thu Mar 29, 2012 9:14 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Ok You guys
I found difference in the output between the left and right channel and responsible for this is AGC .
Hans just put some ...lets say 40 Hz into two channels and check their levels in Spectra lab or some kind of VU meter .It`s about a decibel difference.
Test made with Stand alone beta35 and Normal output .

P.S: difference is 1.5 db
Author:  Brian [ Thu Mar 29, 2012 10:05 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Another thing to keep in mind here is that Un-Break My Heart is a vibrato-laden track. The track name is "Un-Break My Heart", not "Un-Do My Vibrato".

The real flaw here is the fact that you have a global compress setting in multiband, and thus the vocals above about 300 Hz are being softened due to that's what the compress setting tells the software to do. That is exactly the predicament I've run into with trying to define a strong bass region AND keep the vocals strong enough. Part of the vocals are being robbed away due to how multiband currently functions, thus the total signal has less mids hanging around to mask distortion of the bass...which is why you have to lower the "Always clip bass below" setting in loudness.

It is not a loudness issue, so all these tinkerings with loudness equate to treating the symptom, not the cause.
Author:  Valiant [ Thu Mar 29, 2012 10:12 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Image
Could it be that on different systems there is a slight chance of quality differences between users? ;)
Author:  hvz [ Thu Mar 29, 2012 11:05 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Quote:
Ok You guys
I found difference in the output between the left and right channel and responsible for this is AGC .
Hans just put some ...lets say 40 Hz into two channels and check their levels in Spectra lab or some kind of VU meter .It`s about a decibel difference.
Test made with Stand alone beta35 and Normal output .

P.S: difference is 1.5 db
Hm, I'm unable to reproduce this one... Can you send me your settings?

I've ran a sweep, using 1, 2 and 3 bands, with and without ITU-1770 protection, and with different Remove remaining peaks values - in all cases left and right are 100% identical...
Author:  hvz [ Thu Mar 29, 2012 11:11 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Quote:
Could it be that on different systems there is a slight chance of quality differences between users? ;)
Now where is the smiley that shows me rolling over the floor laughing?
Author:  Brian [ Thu Mar 29, 2012 11:33 pm ]
Post subject:  Re: Stereo Tool 6.22 BETA
Image

Image

Image
Author:  DJ-DOGGY [ Fri Mar 30, 2012 12:10 am ]
Post subject:  Re: Stereo Tool 6.22 BETA
Not any special settings but only AGC
settings in the attachment..

Video of difference link :
http://www10.zippyshare.com/v/15553529/file.html

Attachments:
Doggy.rar [3.64 KiB]
Downloaded 276 times
Page 31 of 39 All times are UTC+02:00
Powered by phpBB® Forum Software © phpBB Limited
https://www.phpbb.com/