MeldaProduction

vectorwarrior · Post by **vectorwarrior** » Fri Jan 15, 2016 10:16 pm

Yeah, that's what impressed me too. All the pitch detection I've used in the past has been pretty off and this sounds pretty darn good. Unfortunately I'm on PC, boo

Thanks for the info.

mesaone · Post by **mesaone** » Fri Jan 15, 2016 11:28 pm

vectorwarrior wrote:Yeah, that's what impressed me too. All the pitch detection I've used in the past has been pretty off and this sounds pretty darn good. Unfortunately I'm on PC, boo

Thanks for the info.

Can't MAutoPitch spit out MIDI from incoming audio? If not, Melodyne Editor and Vieklang2 can do it, but they cost money.

vectorwarrior · Post by **vectorwarrior** » Sat Jan 16, 2016 1:04 am

Interesting! I've not seen anything that suggests that it does (I just had a look in the manual and website), but I hope you're right. If anyone knows how to do this, please let me know!

(Sorry for the thread derail)

mesaone · Post by **mesaone** » Sat Jan 16, 2016 1:49 am

vectorwarrior wrote:I've not seen anything that suggests that it does

It turns out that I haven't seen anything that suggests it does, either... I just checked, and it turns out I have a faulty memory - MIDI input was added to MAutoPitch in version 9.14.

vectorwarrior · Post by **vectorwarrior** » Sat Jan 16, 2016 4:05 am

And that is how feature requests are born

MeldaProduction · Post by **MeldaProduction** » Sat Jan 16, 2016 10:12 am

hehe MTuner can actually spit MIDI, but it's very basic, it doesn't produce velocities and stuff.

Anyway thank you folks!

As for morphing vs. vocoding, this particular algorithm (which is probably quite similar to zmorph's) is actually highly based on it, yet kind of different

. It can definitely sound similar, but you can think of it as vocoding with many many bands, it's still different though

.

Jedinhopy · Post by **Jedinhopy** » Sat Jan 16, 2016 3:51 pm

It can't do pitch morphing.

MeldaProduction · Post by **MeldaProduction** » Sun Jan 17, 2016 1:15 pm

What do you mean by that?

Krakatau · Post by **Krakatau** » Sun Jan 17, 2016 1:35 pm

BTW, how long will last the introducing price for Mmorph ?

AudioTraveler · Post by **AudioTraveler** » Sun Jan 17, 2016 1:49 pm

I think Jedinhopy is referring to this https://09-lvl3-gcs-pdl.vimeocdn.com/vi ... 6a48ec92ed

at 3.50min

Edit Seems link the link isn't working anymore here's a working one: https://vimeo.com/116440506

Jedinhopy · Post by **Jedinhopy** » Sun Jan 17, 2016 2:05 pm

What i am trying to get with MMorph is this:

Simple morphing scenario:

Sound A is a 500 Hz sine wave.
Sound B is a 900 Hz sine wave.

The morphing transition between (Sound A) and (Sound B) should sound like the (Sound A) that is playing a 500 Hz sine wave is pitch bending into the 900 Hz frequency.

Complex morphing scenario:

For example pitch morphing a sawtooth waveform into a single sine wave will move all
frequencies towards the single frequency that the sine wave plays.

When morphing has gone from A to B.
All FFT paritals will play the exactly same frequencies because the sound B is only a solo monophonic sine wave.

During the transition between sound A and sound B.
Frequencies will bend upwards and downwards at the same time.

Totolitoto · Post by **Totolitoto** » Sun Jan 17, 2016 2:19 pm

AudioTraveler wrote:I think Jedinhopy is referring to this https://09-lvl3-gcs-pdl.vimeocdn.com/vi ... 6a48ec92ed

at 3.50min

Oh my... the road toward Kyma is still very long

vectorwarrior · Post by **vectorwarrior** » Sun Jan 17, 2016 2:49 pm

Jedinhopy wrote:What i am trying to get with MMorph is this:

Simple morphing scenario:

Sound A is a 500 Hz sine wave.
Sound B is a 900 Hz sine wave.

The morphing transition between (Sound A) and (Sound B) should sound like the (Sound A) that is playing a 500 Hz sine wave is pitch bending into the 900 Hz frequency.

Complex morphing scenario:

For example pitch morphing a sawtooth waveform into a single sine wave will move all
frequencies towards the single frequency that the sine wave plays.

When morphing has gone from A to B.
All FFT paritals will play the exactly same frequencies because the sound B is only a solo monophonic sine wave.

During the transition between sound A and sound B.
Frequencies will bend upwards and downwards at the same time.

Yeah, this is what I was hoping morph would do, but instead it is essentially a spectral based vocoder rather than truly morphing one signal to another.

I feel like melda has all the technology to make this happen:
Spectral analysis of signal a
spectral analysis of signal b
Determination of what transformation needs to happen to turn a into b (auto eqs have this)
Spectral transformation of a into b (as in mtransform)
This would happen every X samples as defined in the spectral settings.

It's a bit like an auto eq, but it takes very regular snapshots of comparison, and instead if eq it is a frequency transformation using mtransform.

It would probably sounds horrible but I would love to try this.

Krakatau · Post by **Krakatau** » Sun Jan 17, 2016 2:58 pm

A lame wanker called Krakatau wrote:BTW, how long will last the introducing price for Mmorph ?

I would say even more :

- how long will last the introducing price for Mmorph ?

...if you don't mind ?

Sampleconstruct · Post by **Sampleconstruct** » Sun Jan 17, 2016 3:13 pm

vectorwarrior wrote:
Jedinhopy wrote:What i am trying to get with MMorph is this:

Simple morphing scenario:

Sound A is a 500 Hz sine wave.
Sound B is a 900 Hz sine wave.

The morphing transition between (Sound A) and (Sound B) should sound like the (Sound A) that is playing a 500 Hz sine wave is pitch bending into the 900 Hz frequency.

Complex morphing scenario:

For example pitch morphing a sawtooth waveform into a single sine wave will move all
frequencies towards the single frequency that the sine wave plays.

When morphing has gone from A to B.
All FFT paritals will play the exactly same frequencies because the sound B is only a solo monophonic sine wave.

During the transition between sound A and sound B.
Frequencies will bend upwards and downwards at the same time.
Yeah, this is what I was hoping morph would do, but instead it is essentially a spectral based vocoder rather than truly morphing one signal to another.

I feel like melda has all the technology to make this happen:
Spectral analysis of signal a
spectral analysis of signal b
Determination of what transformation needs to happen to turn a into b (auto eqs have this)
Spectral transformation of a into b (as in mtransform)
This would happen every X samples as defined in the spectral settings.

It's a bit like an auto eq, but it takes very regular snapshots of comparison, and instead if eq it is a frequency transformation using mtransform.

It would probably sounds horrible but I would love to try this.

The new spectral morphing in Alchemy 2 can do this, depending on how much time you invest and how carefully you prepare the involved samples the results are pretty spectacular. But I doubt this can be a real time process for an FX plugin transforming an incoming audio stream, all necessary calculations (formant analysis and so forth) in Alchemy 2 is done when you re-synthesize the samples and I guess Kyma works the same way. So I assume Vojtech would need to release some sort of standalone app or morphing synth in order to do this kind of thing.

MeldaProduction

KVR Audio

Anything like Zynaptiq Morph?