Hi all,
A bunch of people over at GS have shown some enthusiasm over some comments I made about pitch shifters. I have some pitch shifting technology in Shimmer that sounds pretty good, that I would like to put into a dedicated plugin at some point. I have a bunch of goals for the plugin, and some of them may clash with other goals:
- The pitch shifter needs to be able to perform the "doubling" tasks that most people use an H3000 for. Generally speaking, H3000s tend to be stuck on one of the Multishift patches, with the left and right slightly detuned from each other. So 2 pitch shift voices is a minimum.
- The pitch shifter should be simple to use. Think VintageVerb, as opposed to ÜberMod.
- The pitch shifter should be able to generate complex sounds. 4 voices is better than 2. Note that this goal is in conflict with the above goal.
- The pitch shifter should act as a backwards delay. The pitch shift code in Shimmer can do this already.
- The pitch shifter should not be a ripoff of existing plugins. I don't want to release "Crystallizer, but cheaper!"
- I like the drum sounds on Bowie's Low, so whatever is needed to get this sound has to happen.
- I like "dumb" pitch shifters. I feel that the intelligent, format preserving autoharmonizers are well taken care of by AutoTune, Celemony, and the stuff that is being put into DAWs nowadays. This pitch shifter would be a classic granular time domain pitch shifter, with my own special sauce squirted in there (ewww).
So, open thread on this topic. Please keep any "I'd rather that you work on an xxx plugin" to another thread - feel free to open up an xxx thread if you like.
Thanks for your input! And thanks for scrolling all the way down to this forum. It's my own damn fault for not starting my company name with an "A."
Sean Costello
Let's talk about pitch shifters!
-
- KVRAF
- Topic Starter
- 3426 posts since 15 Nov, 2006 from Pacific NW
-
- KVRAF
- 12352 posts since 7 May, 2006 from Southern California
I love granular pitch shifters.
Particularly, those that allow for ridiculously long grain sizes. Diatonic pitch shifting like on my Evetide Pitch Factor is also most welcome (though I assume that requires pitch tracking).
The ability to freeze the buffer and still be able to adjust the pitch and grainsize would be nice.
+/-48 semitones (at least), please!
Particularly, those that allow for ridiculously long grain sizes. Diatonic pitch shifting like on my Evetide Pitch Factor is also most welcome (though I assume that requires pitch tracking).
The ability to freeze the buffer and still be able to adjust the pitch and grainsize would be nice.
+/-48 semitones (at least), please!
-
- KVRAF
- 5140 posts since 22 Jul, 2006 from Tasmania, Australia
I like playing with granular pitch shifters too.
I am familiar with 2 methods.
1- write to a buffer, and read from it at a different rate
2- chop a grain, and resample it.
Both methods suffer from volume oscillation where u fade the grain ends.
How did u plan to do it Sean?
The best trick I have implemented is starting the grain on a 0 crossing of the original signal. This way there needn't be a fade on grain start.
I'm having headaches with the end though.
Perhaps if I can do well with the end,
there would be minimal amp modding.
All success
I am familiar with 2 methods.
1- write to a buffer, and read from it at a different rate
2- chop a grain, and resample it.
Both methods suffer from volume oscillation where u fade the grain ends.
How did u plan to do it Sean?
The best trick I have implemented is starting the grain on a 0 crossing of the original signal. This way there needn't be a fade on grain start.
I'm having headaches with the end though.
Perhaps if I can do well with the end,
there would be minimal amp modding.
All success
-
- KVRAF
- 3021 posts since 6 Nov, 2006
i'd rather you worked on.... 
i love pitch shifters. the weirder the better. deceptively simple is always a bonus. rhythm trickery w/skew and envelope type controls over grains is always fun too.
i love pitch shifters. the weirder the better. deceptively simple is always a bonus. rhythm trickery w/skew and envelope type controls over grains is always fun too.
-
- KVRian
- 771 posts since 9 Jan, 2012
yes, don't neglect the modulation section of your pitch shifter. audio sidechain{s}, MSEGs, lfos, step sequencers.
would be kool if there was some time stretching options in there, or pitch-shifting-without-preserving-tempo type algos.
would be kool if there was some time stretching options in there, or pitch-shifting-without-preserving-tempo type algos.
-
- KVRAF
- 4014 posts since 29 Jun, 2011 from USA
I'm in favor!
Aiynzahev-sounds
Sound Designer - Soundsets for Pigments, Repro, Diva, Virus TI, Nord Lead 4, Serum, DUNE2, Spire, and others
Sound Designer - Soundsets for Pigments, Repro, Diva, Virus TI, Nord Lead 4, Serum, DUNE2, Spire, and others
-
- KVRAF
- 2193 posts since 25 Dec, 2005
let me throw this in,it's not granular pitch shifting but i feel this must be under my foot.
i like this demo very much and i think digitech did an awesome job with this pedal.
never liked digitech too much to be honest.
also i would love to go software only with this one for pedantic reasons.
http://www.youtube.com/watch?v=FMBNdqdvULA
i like this demo very much and i think digitech did an awesome job with this pedal.
never liked digitech too much to be honest.
also i would love to go software only with this one for pedantic reasons.
http://www.youtube.com/watch?v=FMBNdqdvULA
-
- KVRAF
- 1548 posts since 12 Jan, 2010 from Copenhagen
The above sounds interesting!nix808 wrote:I like playing with granular pitch shifters too.
I am familiar with 2 methods.
1- write to a buffer, and read from it at a different rate
2- chop a grain, and resample it.
Both methods suffer from volume oscillation where u fade the grain ends.
How did u plan to do it Sean?
The best trick I have implemented is starting the grain on a 0 crossing of the original signal. This way there needn't be a fade on grain start.
I'm having headaches with the end though.
Perhaps if I can do well with the end,
there would be minimal amp modding.
All success
I'm all for the thickening type effect that Sean mentioned in the first post. I also like creative use of pitch shift and agree that so called ugly/extreme wide pitch shift is interesting to explore. I'll pull out Uhbik G and explore what it has to offer as well and report back..
waves break, but somehow it all makes sense.
-
- KVRist
- 151 posts since 24 May, 2008
+10E6!!!valhallasound wrote:
- I like the drum sounds on Bowie's Low, so whatever is needed to get this sound has to happen.
Also, I'm not quite sure how to describe this idea, but it seems the ability to map shift amounts to MIDI notes would be cool for some microtonal/beat frequency/Phil Niblock weirdness.
Love your stuff, can't wait to see where this one goes.
Patrick
-
Sampleconstruct Sampleconstruct https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=191286 - KVRAF
- 16142 posts since 12 Oct, 2008 from Here and there
The Valhalla Pitchworx should have 4 voices definitely, it would be great to have controls for coarse tuning (semitones) and finetuning (cents) for each voice and controls for positioning them in the stereo field. And this should be Sean's first plug which has a LFO (or better two) assignable to the pitches of each voice and their pan positions, bipolar and unipolar modes, selectable waveforms (sample+hold please), syncable to host and free running.
If there would be a matrix where one could determine the feedback paths for each voice - e.g. voice 1 is fed back onto itself and also into voice 2 - that could make for some incredibly crazy sounds. HP and LP filters would also be welcome.
Ah, and a cappuccino with some cinnamon powder please...
If there would be a matrix where one could determine the feedback paths for each voice - e.g. voice 1 is fed back onto itself and also into voice 2 - that could make for some incredibly crazy sounds. HP and LP filters would also be welcome.
Ah, and a cappuccino with some cinnamon powder please...
Last edited by Sampleconstruct on Fri Mar 08, 2013 8:07 pm, edited 1 time in total.
-
- KVRian
- 1222 posts since 2 Dec, 2008 from Finland
Maybe make it multiband while you're at it?
-
- KVRAF
- 35276 posts since 14 Sep, 2002 from In teh net
-
- KVRian
- 511 posts since 19 Jul, 2008
...and it should be able to cook the coffee,
and to clear the dishes away...
Now seriously, for me it has to sound good (mean smooth, with a minimum
of artefacts) - it would at least need to stand up against my Digitech pedal.
Low latency as possible, smooth transition, if I shot up a note 2 Octaves up -
no garbling (or the less as possible).
Something as "Doubling" I think is less a problem, as the detune amount needed
is rather low.
Don't know about the tech behind it (as you mention "granular domain") - I think, there's already a lot of "experimental" or "Lo-Fi" sounding Plug-Ins available, but I still look out for some quality FX, which can compare to the stompboxes I got.
My 2 cents,
Cheers
and to clear the dishes away...
Now seriously, for me it has to sound good (mean smooth, with a minimum
of artefacts) - it would at least need to stand up against my Digitech pedal.
Low latency as possible, smooth transition, if I shot up a note 2 Octaves up -
no garbling (or the less as possible).
Something as "Doubling" I think is less a problem, as the detune amount needed
is rather low.
Don't know about the tech behind it (as you mention "granular domain") - I think, there's already a lot of "experimental" or "Lo-Fi" sounding Plug-Ins available, but I still look out for some quality FX, which can compare to the stompboxes I got.
My 2 cents,
Cheers
-
- KVRist
- 279 posts since 13 Feb, 2007
I have been hoping this subject would come up. A Valhalla implementation of something like the H3000 shift algos would be great. A simple UI would be OK but if there were some way to access more of the guts of the algos via extended tags in the preset file and plugin automation that would be even better.
