KVR Audio

kvaca · Post by **kvaca** » Mon Dec 22, 2014 6:05 pm

and without obtaining preecho? /=without use of linear phase design filter?/
If yes, are there any plugins capable of this??
Any info is welcome, because all plugins I have looked at so far obviously use LP design /like PHA979 etc./ or dont shift all frequencies by the same amount of phase/Phasetone/

cymazic · Post by **cymazic** » Mon Dec 22, 2014 9:55 pm

Is this any good to ya ? http://www.kvraudio.com/product/prefix_ ... nd/details

No_Use · Post by **No_Use** » Mon Dec 22, 2014 10:09 pm

http://audiogeekzine.com/tag/phasebug/

I don't know what kind of design this uses though.

sinkmusic · Post by **sinkmusic** » Mon Dec 22, 2014 10:22 pm

Would TB Flx and its autophase mode help ?

kvaca · Post by **kvaca** » Mon Dec 22, 2014 11:13 pm

sinkmusic wrote:Would TB Flx and its autophase mode help ?

interesting, but based on site description it seems to be a dynamic equalizer with some special features...but not a phase alignment capable plugin I´m searching
also their autophase wording is probably tied to some innovative equalizer design

TB FIX wrote: The phase response of the equalizer is dynamically adjusted to give the best of both worlds, continuously adapted to the input signal and the desired equalizer curve.

sinkmusic · Post by **sinkmusic** » Mon Dec 22, 2014 11:25 pm

Yes, my bad : i think it is related to linear-phase eq filtering (or other technique) instead of the "own's signal phase"...

camsr · Post by **camsr** » Tue Dec 23, 2014 2:50 am

Yes there is an IIR phase shifter (or Dome Filter). This was used in the original Bode Frequency Shifters. I don't think it's downloadable in a plugin though. In case you may want to try to implement it yourself: http://yehar.com/blog/?p=368

valhallasound · Post by **valhallasound** » Tue Dec 23, 2014 3:21 am

camsr wrote:Yes there is an IIR phase shifter (or Dome Filter). This was used in the original Bode Frequency Shifters. I don't think it's downloadable in a plugin though. In case you may want to try to implement it yourself: http://yehar.com/blog/?p=368

A "Dome Filter," or "phase differencing network," doesn't create a signal that is 90 degrees shifted from the input. Instead, it has 2 outputs, which have a 90 degree phase difference from each over (plus or minus a small amount of error) over a broad frequency range. The output signals usually have a fair amount of phase shift compared to the input (up to 1080 degrees for a 12th order phase differencing network).

An FIR Hilbert filter is usually linear phase, with a certain amount of fixed delay for the majority of the energy (there's gotta be a better name for this, but whatever the "peak" of the impulse response is). So, you could argue that an FIR Hilbert filter isn't shifted 90 degrees from its input, but it is shifted 90 degrees from the delayed input.

Sean Costello

camsr · Post by **camsr** » Tue Dec 23, 2014 4:34 am

The dome filter doesn't preserve the input's phase properties, so yes there is phase distortion from it's use.
As for IR "peak", that could be called something purposeful I guess, maybe origin or zero-time

But that would only apply to linear phase IRs.

kvaca · Post by **kvaca** » Tue Dec 23, 2014 8:49 am

valhallasound wrote: An FIR Hilbert filter is usually linear phase,..
Sean Costello

hi Sean,Im aware of that but all I want to know is if its possible technically to create 90 degree phase shift while NOT using linear phase design...or put it another way - I mainly want to get rid of that strong preecho /while maintaining good bass response/, if its possible

kvaca · Post by **kvaca** » Tue Dec 23, 2014 1:04 pm

No_Use wrote:http://audiogeekzine.com/tag/phasebug/

I don't know what kind of design this uses though.

tested now, uses linear phase design

valhallasound · Post by **valhallasound** » Tue Dec 23, 2014 4:36 pm

kvaca wrote:
valhallasound wrote: An FIR Hilbert filter is usually linear phase,..
Sean Costello
hi Sean,Im aware of that but all I want to know is if its possible technically to create 90 degree phase shift while NOT using linear phase design...or put it another way - I mainly want to get rid of that strong preecho /while maintaining good bass response/, if its possible

The Dome filter described above is IIR. Not linear phase, not minimum phase, just some sort of strange phase. Lots of synthesis languages will have Hilbert networks in them, and most of these are 12th order IIR phase differencing networks, so if you are familiar with Csound, MAX/MSP, Reaktor, etc., you can try these out.

Sean Costello

kvaca · Post by **kvaca** » Tue Dec 23, 2014 5:17 pm

valhallasound wrote:
kvaca wrote:
valhallasound wrote: An FIR Hilbert filter is usually linear phase,..
Sean Costello
hi Sean,Im aware of that but all I want to know is if its possible technically to create 90 degree phase shift while NOT using linear phase design...or put it another way - I mainly want to get rid of that strong preecho /while maintaining good bass response/, if its possible
The Dome filter described above is IIR. Not linear phase, not minimum phase, just some sort of strange phase. Lots of synthesis languages will have Hilbert networks in them, and most of these are 12th order IIR phase differencing networks, so if you are familiar with Csound, MAX/MSP, Reaktor, etc., you can try these out.

Sean Costello

no, sadly Im not familiar with these synths...
but another problem is I need to keep original audio untouched and only create 90 degree shifted version - preferably without preecho - for paralel processing and final summing...and if I uderstand your previous post correctly this will not be possible to obtain with Dome filter

valhallasound · Post by **valhallasound** » Tue Dec 23, 2014 7:18 pm

kvaca wrote:
valhallasound wrote:
kvaca wrote:
valhallasound wrote: An FIR Hilbert filter is usually linear phase,..
Sean Costello
hi Sean,Im aware of that but all I want to know is if its possible technically to create 90 degree phase shift while NOT using linear phase design...or put it another way - I mainly want to get rid of that strong preecho /while maintaining good bass response/, if its possible
The Dome filter described above is IIR. Not linear phase, not minimum phase, just some sort of strange phase. Lots of synthesis languages will have Hilbert networks in them, and most of these are 12th order IIR phase differencing networks, so if you are familiar with Csound, MAX/MSP, Reaktor, etc., you can try these out.

Sean Costello
no, sadly Im not familiar with these synths...
but another problem is I need to keep original audio untouched and only create 90 degree shifted version - preferably without preecho - for paralel processing and final summing...and if I uderstand your previous post correctly this will not be possible to obtain with Dome filter

What you are looking for MIGHT be physically impossible. All of the 90 degree phase shift things I know about have some sort of delay or phase lag in there (IIR phase differencing, FIR Hilbert, FFT w/phase rotation). I can't say for sure if it is impossible, as I don't have the math chops to prove or disprove this, but my gut tells me that it would be difficult for a filter to just add exactly 90 degrees of phase shift for all frequencies between its input and output.

Sean Costello

Ichad.c · Post by **Ichad.c** » Thu Jan 01, 2015 9:35 am

Sean is right, there will be either phase lag or delay. The main problem is that you need this to work in parallel; if you worked serially - a IIR Dome filter might be better - but is wholly dependent on how you use it in your signal chain. I don't know that much about FIR Hilbert transforms, but a IIR Dome filter will make notches in the spectrum if it were ever to meet its original copy, this is because of the phase lag -> a dome filter is actually just a specially tuned mono to stereo phaser.

Is it possible to shift audio by 90 degrees