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

Low latency secondary input (microphone)
https://forums.stereotool.com/viewtopic.php?t=853
Page 67 of 75

Author:  hvz [ Sat Apr 03, 2010 2:32 pm ]
Post subject:  Re: Low latency secondary input (microphone)

Quote:
This Robotized Voice caused bu Phase Rotatator .. other processors have this side-effect.. some more, some less...
Curently in ST this is not so dangerous and certainly less then other processors with phase rotator.

I know many peoples who actually loves this effect.

btw, i think in US Phase Rotating is Forbiden.
Maybe that's why some processors have always on Phase rotating and another button/phase rotator On/Off. :)
Ow ok. I was afraid that it would be caused specifically by my filter (you wrote something about poles and bands a few days ago, well, if I were to have to express my filter in poles and bands I would need hundreds of both :shock: - this filter is really different from what I found online).

What do you mean about forbidden? That wouldn't make any sense... And I've never heard about that?

Author:  hvz [ Sat Apr 03, 2010 3:08 pm ]
Post subject:  Re: Low latency secondary input (microphone)

New version uploaded as BETA3A again. Robot voices problem should be reduced, for the tracks I measured with the symmetry of the output is about equal to that of the previous version.

Author:  Luke [ Sat Apr 03, 2010 7:06 pm ]
Post subject:  Re: Low latency secondary input (microphone)

Quote:
New version uploaded as BETA3A again. Robot voices problem should be reduced, for the tracks I measured with the symmetry of the output is about equal to that of the previous version.

Hey Hans, I like robot voices! Reminds me of myself when my wife says (tells) me something :mrgreen:

Why should phase rotation be illegal in the US???? If you are within the specified bandwidth everything should be ok I think.

Author:  Bojcha [ Sat Apr 03, 2010 7:09 pm ]
Post subject:  Re: Low latency secondary input (microphone)

i have one voice sound at 700-800Hz .. and it's not totaly fixed...

is that because you moved start freq. from 500 to 1000 Hz ?

Author:  hvz [ Sat Apr 03, 2010 7:32 pm ]
Post subject:  Re: Low latency secondary input (microphone)

Quote:
i have one voice sound at 700-800Hz .. and it's not totaly fixed...

is that because you moved start freq. from 500 to 1000 Hz ?
Does this mean that it's at least better than the previous version? If so, when I've made things configurable (tonight?) you should be able to raise the start frequency further, or make the effect smoother.

Oh wait: Not totally fixed -> do you mean that there's distortion now, or robot voice?

Anyway, in this version I changed 3 things, 2 of which might affect both issues (maybe even 3 if the robot voices are gone, because that could _also_ be caused by the removed pre-ringing filter) - so when it's configurable you can check which of the two causes it :-)

Author:  hvz [ Sat Apr 03, 2010 8:47 pm ]
Post subject:  Re: Low latency secondary input (microphone)

Configurable version uploaded as BETA3A (see remark below!)
4 new parameters in the Loudness/Limiting window:
- ON - turns it on
- Start Freq - start frequency, sounds below this frequency are not changed
- Start Speed - change speed at the start frequency. Doesn't really mean anything yet (the value is incorrect), but higher values means more aggressive rotations.
- Smoothness - this is a VERY interesting setting. Higher = less aggressive. But even at 99% I don't really see a reduction in the effectiveness (and the old value in the software was 95%). So setting this higher might improve things, but I still need to check it on multiple recordings.

I've only updated the Winamp and Stand Alone SSE2 versions. (no time, have to leave NOW...)

Author:  Bojcha [ Sat Apr 03, 2010 10:07 pm ]
Post subject:  Re: Low latency secondary input (microphone)

Quote:
i have one voice sound at 700-800Hz .. and it's not totaly fixed...
is that because you moved start freq. from 500 to 1000 Hz ?
Yes, it's because start freq 1000Hz .. best start freq. is 800Hz :)

Author:  hvz [ Sat Apr 03, 2010 10:42 pm ]
Post subject:  Re: Low latency secondary input (microphone)

Quote:
Yes, it's because start freq 1000Hz .. best start freq. is 800Hz :)
Great! Does increasing the smoothness (_maybe_ combined with a slightly lower start freq) reduce the robot voices issue? (I haven't heard it yet so I cannot easily check it).


Note: I'll post an image with an explanation of the phase rotation filter in a moment here to clarify what each slider does. That might help understanding things better.


Edit: With a start freq of 800 Hz I'm getting the best results with a smoothness of 98%. (I haven't changed initial speed)
Edit #2: Hm, start freq of 700 Hz seems to be better (with this smoothness level)

Author:  Bojcha [ Sat Apr 03, 2010 10:51 pm ]
Post subject:  Re: Low latency secondary input (microphone)

Not robot voices for now ...
and yes, beter selection for initial speed is 1400

You forgot to enlarge the window :) you know already that

Author:  hvz [ Sat Apr 03, 2010 11:02 pm ]
Post subject:  Re: Low latency secondary input (microphone)

I quickly drew something in Paint :oops:

Image

What you see here is:
- At frequencies below Start Freq, nothing happens
- At Start Freq, a full rotation occurs at a certain speed (should be related to the 'speed' parameter, but there's a bug in the calculation so setting it to 1500 Hz does not mean that the first rotation ends at start freq + 1500, but a lot sooner. I'll fix that later.)
- From that point on, the rotation speed goes DOWN very rapidly.

(Note that -180 degrees equals +180 degrees; there's no jump at the sides of the image)

This would give a very rough transition at the start freq (which in turn causes horrible pre-ringing etc.). To fix that, I've added a smoothing parameter that slowly follows the values that are drawn in this picture. So instead of a steep angle at the start, it starts very smoothly (which also means that very little happens near the start frequency). Setting smoothing higher means that the following is done more slowly, so the same phase rotation is reached at a higher frequency.

- Setting the start frequency higher means that lower frequencies (where a similar shift in phase means a much bigger difference in time) are left unaltered, so the sound will be more natural. But you risk getting more distortion.
- Setting the speed at a higher frequency means that rotation is SLOWER (!!!) - hence also that there's less difference compared to the original sound, but more chance of getting distortion.
- Setting smoothing higher also reduces the effect (especially at lower frequencies), so that also makes the sound more similar to the original, but you might risk getting more distortion. Setting it to lower values causes horrible amounts of pre-ringing (probably due to the steeper start).

Page 67 of 75 All times are UTC+02:00
Powered by phpBB® Forum Software © phpBB Limited
https://www.phpbb.com/