Sonic Charge Synplant 2 Announced

[Caine123]

somehow strange how all come out with AI at the same time around now? did they wait until recently or what?

[gentleclockdivider]

I would love to see the A.I. doing it's magic on synths like Zebra or Tranzistow ...

Or a modular synth that covers all (most) synthesis methods

Tranzistow

[martiu]

I would love to see the A.I. doing it's magic on synths like Zebra or Tranzistow ...

Or a modular synth that covers all (most) synthesis methods

Tranzistow

what is that ?

[sqigls]

somehow strange how all come out with AI at the same time around now? did they wait until recently or what?

technology has advanced enough to make it viable. Neural engines of Macs for example.
as Magnus said, with his modest machine learning rig, the AI took several months to learn Synplant. And now it takes only a few seconds for Synplant to resynthesise a sound. When he first succeeded in getting Synplant to resynthesise a sound he said it took about half an hour.

[Shabdahbriah]

As fascinating and good sounding it is ... I'm not sure if just 'rolling a dice' is as satisfying as creating your own sound by skills you've learned and gathered over time.

The same goes with AI.

Will I be happy with something that only needed a handful of prompts (if at all)?

but it's just a sound, surely what you then do with the sound is also important.
if the sound is good, does it matter how it was arrived upon?
i love making my own sounds, but if i hear a preset that works for me, then ill use that.

you could think of it as a synthesiser of found sounds

Indeed, I very much enjoy the serendipity factor.

[Dalle]

imagine the same AI tech on 1000 partial additive synth

You don't need AI to do resynthesis on an additive synth, just analyze the partials. The tech for that has been available for a long time. Making a non-additive synth approximate sounds in an automated way is quite a different process I imagine.

[Caine123]

somehow strange how all come out with AI at the same time around now? did they wait until recently or what?

technology has advanced enough to make it viable. Neural engines of Macs for example.
as Magnus said, with his modest machine learning rig, the AI took several months to learn Synplant. And now it takes only a few seconds for Synplant to resynthesise a sound. When he first succeeded in getting Synplant to resynthesise a sound he said it took about half an hour.

thx, just strange you read AI everywhere now, well Terminator when?

anyways, looking forward to upgrade!

[martiu]

oh yeah, i forgot about resynthesis, i never used it on additive synths, may have to try it on Loom later today

[Shabdahbriah]

As fascinating and good sounding it is ... I'm not sure if just 'rolling a dice' is as satisfying as creating your own sound by skills you've learned and gathered over time.

I probably will improve my sound design skills with Genopatch. I will probably learn methods I would not have learned with my traditional tweaking.

I don't know that ones skills would necessarily "improve" overall as for example when Magnus goes back into the editing page (oscillators, filters, envelopes, etc) to attempt to essentially reverse engineer what was created with Genopatch, even if you could nail it, it will still be specific to Synplant. Though there is certainly fun to be had in the exploration of how this works.

For those who missed the previous page post, there is around a 30-minute intro to the new features here:

[peterdh]

@Shabdahbriah or any other betatester: I want to collect samples for Genopatch these months. Do they have to have a specific note number? Or does Genopatch automatically figure the pitch of the sample out?

[LFO8]

As fascinating and good sounding it is ... I'm not sure if just 'rolling a dice' is as satisfying as creating your own sound by skills you've learned and gathered over time.

I probably will improve my sound design skills with Genopatch. I will probably learn methods I would not have learned with my traditional tweaking.

I don't know that ones skills would necessarily "improve" overall as for example when Magnus goes back into the editing page (oscillators, filters, envelopes, etc) to attempt to essentially reverse engineer what was created with Genopatch, even if you could nail it, it will still be specific to Synplant. Though there is certainly fun to be had in the exploration of how this works.

For those who missed the previous page post, there is around a 30-minute intro to the new features here:

That looks very promising! Imagine if in e few years most synth plugins come with an AI preset generator like this.

[Shabdahbriah]

@Shabdahbriah or any other betatester: I want to collect samples for Genopatch these months. Do they have to have a specific note number? Or does Genopatch automatically figure the pitch of the sample out?

No you don't 'need' specific (clearly defined) notes to get a reasonable (at times remarkable) proximity/representation, but it may expedite the process if you are wanting to get as close as possible. Though samples with complex overtones/harmonics generate interesting results, as those harmonics/overtones seem to get/be 'isolated' and then replicated or 'bent' by the algorithms in Genopatch. Then, combining/merging the original sample with the generated version, creates something uniquely 'complex'. Rinse and repeat.

Again, I'm having good results with short (2-4 second) mono (16/44.1) .wav samples, rather than loading longer samples, and using the selection bars (starts here/ends here) portion of the sample, within the sample waveform view window, which are a little small for my eyes.

[MrJubbly]

Again, I'm having good results with short (2-4 second) mono (16/44.1) .wav samples, rather than loading longer samples, and using the selection bars (starts here/ends here) portion of the sample, within the sample waveform view window, which are a little small for my eyes.

Question: Can you drag and drop .wav samples from the DAW browser or external audio editors onto the GenoPatch UI?

Or do you have to manually browse for samples, within the GenoPatch UI itself? (as shown in these teaser videos).

Because, user workflow would be greatly improved and more efficient with support for drag and drop of samples into the plugin.

Also, if you do currently have to use just the internal Synplant browser, does it at least also have an audio preview option whenever you click each sample (i.e. before loading it into GenoPatch?)

These are some common workflow improvements that will need to be added to this synth, if not supported upon the forthcoming v1.0 official release.

[dayjob]

Also, if you do currently have to use just the internal Synplant browser, does it at least also have an audio preview option whenever you click each sample (i.e. before loading it into GenoPatch?)

in that long video recently posted i think he previews samples from within the browser of synplant 2 and even says so while he's doing it.

edit: it's on this page 5 posts up.

[Shabdahbriah]

Again, I'm having good results with short (2-4 second) mono (16/44.1) .wav samples, rather than loading longer samples, and using the selection bars (starts here/ends here) portion of the sample, within the sample waveform view window, which are a little small for my eyes.

These are some common workflow improvements that will need to be added to this synth, if not supported upon the forthcoming v1.0 official release.

Neither will have to be 'added'.

[edit] Sorry I misread one thing as previewing the sample within Genopatch (which it does), so the sample 'auditioning' within the sample 'browser' prior to loading it, is being worked on, as it is highly requested.