Trying to find software that can "blend" two samples together

  On 9/19/2013 at 9:18 PM, John Ehrlichman said:

so have you guys had a chance to use Kyma? I'm curious how much the CDP spectral morphing compares..

 

Played around with kyma years ago in school, but never dove into all the options, can't say how it compares.

I like this demo:

http://www.youtube.com/watch?v=nt9tXXaXRrM#t=53

  On 9/19/2013 at 9:18 PM, John Ehrlichman said:

so have you guys had a chance to use Kyma?

 

Can't say I have. Unfortunately my need to utilise programs like that has diminished over the years, at least straight up musically. Might have a look into it though, strikes me as something that would be amazing for sound design.

I'm sure someone here gave me this a while ago - but god knows who, and god knows where I've archived it:

Loris was originally used as the basis for much of Kyma's spectral resynthesis and non-Tau morphing algorithms (the Tau morphs require the hand-drawing mentioned above and lots of trial and error but can ultimately sound the best). There are Csound opcodes, as well.

If you don't mind code maybe you can get some interesting results:

http://www.hakenaudio.com/Loris/

http://csounds.com/manual/html/loris.html

interesting, had no idea about Loris. I tried Tau out a few times but could never get it sound good as the standard method, for me it ended up sounding more like camelaudio alchemy.

really nobody mention spectral repair in izotope rx? i wonder if rx3 has improved the algorithm. it's pretty fancy results but its not realtime

you have two waves pasted into the same file with a silent space inbetween them which spectral repair will attempt to bridge with FFT and noise data

beatwife convwife?

also http://www.martin-brinkmann.de/plugs.html

plugin's called 'morfl'

  On 9/21/2013 at 12:34 AM, RadarJammer said:

really nobody mention spectral repair in izotope rx? i wonder if rx3 has improved the algorithm. it's pretty fancy results but its not realtime

 

you have two waves pasted into the same file with a silent space inbetween them which spectral repair will attempt to bridge with FFT and noise data

interesting, are there any tutorials or demonstrations of this? I have RX3 but didn't manage to find that feature.

  On 9/21/2013 at 3:45 AM, skibby said:

also http://www.martin-brinkmann.de/plugs.html

 

plugin's called 'morfl'

its a decent plugin, sounds similar to the Prosoniq morph plugin posted earlier in the thread. It still has that typical convolution 1024 FFT sound though, pretty much anything that uses standard FFT algorithms is going to sound more or less the same, like an IR reverb or convolution effect.

i "really" want to know if anybody has tried the beatwife convolver, or any of the other beatwife plugins, they look and sound amazing on the youtubes.

  On 9/21/2013 at 5:06 AM, John Ehrlichman said:

 

  On 9/21/2013 at 12:34 AM, RadarJammer said:

really nobody mention spectral repair in izotope rx? i wonder if rx3 has improved the algorithm. it's pretty fancy results but its not realtime

 

you have two waves pasted into the same file with a silent space inbetween them which spectral repair will attempt to bridge with FFT and noise data

interesting, are there any tutorials or demonstrations of this? I have RX3 but didn't manage to find that feature.

 

its really easy to do, of course what i described isn't the intended purpose so there isn't any tutorials for that. just use the standard feature but instead of having the same sound on both sides of the audio gap have different sounds and mess with the fft and noise rules to get the best sound

see at 1:20

  On 9/21/2013 at 9:56 PM, skibby said:

i "really" want to know if anybody has tried the beatwife convolver, or any of the other beatwife plugins, they look and sound amazing on the youtubes.

i tried it after reading this thread, my opinion is that it sounds a lot like Native Instruments Vokator, it sounds almost between a traditional vocoder and an FFT convolver, very good for percussive sounds controlling other sounds (as shown in the demo video above).

  On 9/22/2013 at 5:28 AM, RadarJammer said:

 

  On 9/21/2013 at 5:06 AM, John Ehrlichman said:

 

  On 9/21/2013 at 12:34 AM, RadarJammer said:

really nobody mention spectral repair in izotope rx? i wonder if rx3 has improved the algorithm. it's pretty fancy results but its not realtime

 

you have two waves pasted into the same file with a silent space inbetween them which spectral repair will attempt to bridge with FFT and noise data

interesting, are there any tutorials or demonstrations of this? I have RX3 but didn't manage to find that feature.

 

its really easy to do, of course what i described isn't the intended purpose so there isn't any tutorials for that. just use the standard feature but instead of having the same sound on both sides of the audio gap have different sounds and mess with the fft and noise rules to get the best sound

 

see at 1:20

 

ok i gave this a go. The room reverb reducer and the general background noise removal stuff in it is top notch.

Here is what the spectral repair feature sounds like (lol)

RX3testlolz.wavFetching info...

i purposefully dropped out the audio from :06 - :11 and this was the result. Sounds interesting, might work much better on other types of sounds. I'd be curious to try it on a sort of ambient field recording.

http://www.eisenberg-audio.de/einklang

Or Iris.

  On 9/21/2013 at 12:34 AM, RadarJammer said:

really nobody mention spectral repair in izotope rx? i wonder if rx3 has improved the algorithm. it's pretty fancy results but its not realtime

 

you have two waves pasted into the same file with a silent space inbetween them which spectral repair will attempt to bridge with FFT and noise data

Really! I'll have to try that. I was going to suggest iZotope Iris:

http://www.izotope.com/products/audio/iris/

I use for blending samples Argeiphontes Lyre by Akira Rabelais. It's Mac only, but can create some bizarre and good sounding results. A lot of try-and-error as well, but you can do something at least with it coming close to other methods.

I use it since my musical beginning and it has never failed to amaze- it has an artsy fartsy GUI, but looks really decent.

and the best: it's free and offers also programs for pictures, video and sound………

nice, i remember you recommending that program to me Thorsten but never tried the morphing in it.

By the way anyone who's interested in serious artifact-free (relatively speaking) noise removal/clean up/audio forensic stuff. the new RX3 pack by Izotope is fucking amazing. The Spectral repair thing isn't that good, but the other features in the suite especially the 'room reverb' remover filter are very impressive. I feel like we're not too far off from software that can actually remove crowd noise and reverb from audience recordings and make them sound a lot better.

I saw it mentioned by Richard Devine and was kind of tempted but then I found out the advanced version with all those fancy pants features was $1,199 vs the 'standard' $349

The Spectral Repair is amazing in RX. I use it all the time. I was able to completely remove cicadas from an interview amongst a lot of other things. You can lasso any group of frequencies and listen to what you lassoed for glitchy type stuff. You copy and paste frequencies anywhere in the spectrum to make some really wild sounds.

interesting, i was trying to do mostly silly stuff with it like having it fill in gaps of silence (like in the demo) but i should investigate it more.

I've been looking for something that could do this for years, it doesn't seem like too odd of a thing to request either considering all of the insane plugins out there.

I don't know how many electronic musicians are consistently using Kyma, but I know Greg Hunter is probably the best. He used it for this Dub/Indian Music based album, which I think is 100% kyma. Vocals and instrument samples used in it became completely unrecognizable

[youtubehd]Wq1zpNDgZoA[/youtubehd]

[youtubehd]zzW0sxEvtRA[/youtubehd]

that metasurfing stuff is awesome, sounds like a lot of cross synthesis effects (having drums trigger sitars, etc) really digging it.

the electronic musicians I know for sure who use kyma are Thomas Dimuzio, Anthony Bisset, Amon Tobin, Richard Devine, BT , Brian Trifon, Matmos, and myself. Those are the only people I am 100% sure about, and to what extent Amon Tobin and Matmos or BT actually used it, i have no idea. With the others I've seen/know of them using it pretty extensively

and in direct response to what you said about being surprised there isn't a plugin out there that does this. In terms of processing power and available open source algorithms there easily could be at this point in time. The problem is that to make one that does it all automatically for you with a cross fader to morph between the sounds that actually SOUNDs like a real morph would probably be high level programming that unless someone is very very dedicated to the technique, they'd want to get paid for their efforts understandably. As Kyma stands right now you cannot simply load in 2 samples and go to town on a morph, I'd say it takes about an average of 10-20 minutes of prep-time (if you're lucky) for each sample you want to morph. This is fine if you don't mind having a set group of samples you'd prepared before hand (one you prep them they load instantly, because you aren't even loading in wave files anymore you load in spectrum files that Kyma spits out after prep is done) , but for a fluid and sort of 'unintentional' fuck around music making session the morphing on kyma doesn't really go in that toolkit.

I could see a company like Izotope or Celemony having the knowledge and resources to put something like this out, but would it make any money? And if it did come out everybody would try to use it and the effect might become gimmicky. It's a hard call. For now Kyma is still the best sounding way to go for the classic Terminator 2 people face morph equivalent for sound. You can get texture morphs all day on stuff like Prosoniq and others but with no time based morph it doesn't feel to our brains like what we understand as a 'morph', it's more just like fft phase vocoding

Uwe Schmidt (Atom tm) and autechre are also confirmed kyma users.

  On 10/16/2013 at 5:38 PM, Alcofribas said:

Uwe Schmidt (Atom tm) and autechre are also confirmed kyma users.

Pure brilliance.

that sounds awesome and like Kyma. Man the vocoder effects on kyma sound really fucking good. In a weird way they sound closer to the classic Meat Beat Manifesto/Kraftwerk vocoder sound than most new hardware and even newer analog vocoders. The problem is you have to have a very expensive souped up kyma (at the time I bought mine I would have had to spend at least $6,000) to get the vocoder to process in real-time. Mine has about a 200-300 ms delay for vocoding at the very basic level, any more complex and it's like 500-700ms. Huge cock tease for someone who didn't buy the sports car version of the capybara