Tags: posts polarity-music AI download FM-synths machine-learning oscillators Reverb soniccharge Sound-Design Sound-Generator SynPlant synthesizers Tutorial

SynPlant 2: The Future of Synthesizers

Tutorial | Oct 04, 2023

In my latest video, I discuss SynPlant 2, a groundbreaking synthesizer with the new GenoPatch feature that uses machine learning to create FM patches from samples. This feature allows users to generate multiple FM patches from a single sample, with options to adjust tuning, volume, and grow out different sound variations. I believe this is a game-changing innovation in the world of synthesizers and hope to see other synths incorporate similar features in the future.

You can watch the Video on Youtube - support me on Patreon

Questions & Answers

Maybe you dont watch the video, here are some important takeaways:

What is the main feature of SynPlant 2?

The main feature of SynPlant 2 is the new GenoPatch feature, which allows users to drop samples into the plugin and use machine learning to create a 2 oscillator FM patch from the sample. This innovative feature sets SynPlant 2 apart from other synthesizers and opens up new possibilities for sound creation.

How does the GenoPatch feature work in SynPlant 2?

To use the GenoPatch feature, you simply drop a sample into the plugin and it generates multiple FM patches based on the sample. You can then play the patches on a keyboard, adjust tuning and volume, and use the growing interface to create different variations of the sound. This feature utilizes machine learning algorithms to recreate samples with just two oscillators.

What makes the GenoPatch feature in SynPlant 2 unique?

The GenoPatch feature in SynPlant 2 is unique because it uses machine learning to recreate a sample with a 2 oscillator FM synthesizer, rather than using sine partials like other synthesizers. This innovative approach allows for more accurate and versatile sound creation, and has the potential to revolutionize the way synthesizers are used in music production.

Is SynPlant 2 worth the price?

While SynPlant 2 may be considered expensive for a synthesizer, the GenoPatch feature alone makes it worth the price. The ability to easily and accurately recreate samples using machine learning is a game-changer for sound design, and the possibilities it opens up are exciting for producers. If you're serious about music production and want to explore new sound creation techniques, SynPlant 2 is definitely worth considering.


This is what im talking about in this video. The text is transcribed by AI, so it might not be perfect. If you find any mistakes, please let me know.
You can also click on the timestamps to jump to the right part of the video, which should be helpful.

[00:00:00] Hey folks, welcome back to another video. Today it's about SynPlant 2 which just came out
[00:00:04] yesterday evening and in my opinion this is the future of synthesizers because we have
[00:00:14] now this new GenoPatch feature in here where you can just drop samples into this plug-in
[00:00:23] and it tries to figure out with machine learning how to create a 2 oscillator FM patch from
[00:00:30] this sample. So let's take here for instance, write the sample, you drop this in, then you
[00:00:42] say let's create some patches from it. You can see here CPU spikes pretty hard, 84%,
[00:00:53] only in the generation process, then it drops down of course which is nice. So now we generate
[00:00:58] it basically from this one sample we generate multiple FM patches. Just use this one here
[00:01:11] and then we can play it on the keyboard. You can also hear corrected tuning, so C is C
[00:01:23] and D is D, then we can normalize the volume. You can play around with velocity sensitivity
[00:01:39] and we can grow out these branches into so the sound becomes different. If you like a
[00:01:51] sound or some of these branches here we can just use the seed and plant it in the middle.
[00:02:13] So synth plant, if you don't know how it works, I mean this one is around for ages. Synth
[00:02:19] plant 2 with the GenoPatch feature just came out yesterday. So this one is basically just
[00:02:24] an, I mean just, I mean it's a really nice sounding 2 FM oscillator synth. We can dive
[00:02:31] into the DNA here and under the hood this is basically just envelopes, 2 oscillators
[00:02:37] with FM and filters and effects. This is what's going on under the hood. But to create these
[00:02:46] sounds you don't need to use here all the usual synthesizer elements. You can also just
[00:02:54] use this fancy interface here where you create a random seed in the middle, which is create
[00:03:01] a new random seed, right? You get a random sound. So all parameters of these FM synths
[00:03:07] are randomly generated and then you use this growing interface here to drag some of these
[00:03:15] branches out. And if you like the sound of one patch or one branch you can use the seed
[00:03:26] and plant it back into the middle and go from there. So this is the idea of synth plant,
[00:03:32] right? To have this kind of growing seed growing interface to create sounds. But with Gino
[00:03:39] patch we can just use samples, try to recreate a sample with an FM synthesizer, which is
[00:03:45] really nice. I think no one has ever done this before. I mean there are some recent
[00:03:50] synthesizers out there but they use usually sine partials. So they just take the overtones
[00:03:56] of a sample and then recreate these overtones with sine partials. But here we have just
[00:04:01] two oscillators and machine learning algorithm tries to figure out how to turn all these
[00:04:07] knobs on the synthesizer to recreate the sound with just two oscillators, which is really
[00:04:13] nice. And I really hope we see this in other synthesizers too. So just imagine you have
[00:04:19] here maybe a pulley synth in Bitwig, right? And you want to recreate some kind of sound
[00:04:26] and you have the sample, you drag the sample into the pulley synth and then the machine
[00:04:30] learning algorithm tries to figure out what knobs turn so pulley synth sounds exactly
[00:04:37] or similar to the sample. Really nice, would be really nice to have this instead of tweaking
[00:04:45] synthesizers here for hours just to have a bad representation of the sample in a synthesizer.
[00:04:54] So machine learning could really help. In my opinion this is also the right way of using
[00:04:59] machine learning or AI if you want to call it that way. In music production, right, you
[00:05:06] just make certain processes more easy, more accessible and faster instead of me sitting
[00:05:14] in front and tweaking all these knobs for hours, I just drag it in and the machine learning
[00:05:19] algorithm does it for me. So this is just my take on this. So I think this feature here
[00:05:27] is probably most innovative feature in the last maybe 10 or 20 years in synthesizers.
[00:05:35] I have no idea why no one else did this before. It's just I wanted this so many years before
[00:05:43] and I said, well, it's just machine learning, right? You have an input, you have an output,
[00:05:48] just test on the output of the input matches the output and if not, then just randomize
[00:05:52] some parameters until it fits the output. And then you can make a nice model that resynthesizes
[00:06:02] samples into synthesizer sounds. So in my opinion, every synthesizer should have these
[00:06:09] this kind of feature at some point in the future. Okay, so Simplan 2, like I said, is
[00:06:16] out at the moment. You can download just a Trial-Version. I'm not affiliated with the company.
[00:06:22] So this is just my personal opinion. I really like this. I want to play around with this
[00:06:26] here the next days and see how I can implement it in my daily producing, producing process.
[00:06:34] But I really like the idea behind this. The only downside is it's pretty expensive, I
[00:06:39] guess for synthesizer. It's one of the 50 bucks. But the feature is really, it's probably
[00:06:45] really worth it. So maybe look into this. Link is in the description below. And yeah,
[00:06:53] maybe let's try out some some different samples, I would say. So yeah, you have some piano
[00:06:59] samples on my hard drive from Skyance. So just use this here, drag it in. And yeah, just
[00:07:11] recreate the sound or try to recreate it. CPU spikes again pretty hard. It takes a while
[00:07:28] here. But I think it's it's already sounding nice here. It's kind of pretty good. So let's
[00:07:45] take this. So we can use now velocity here. We can say correct tuning because whatever
[00:07:53] I don't know what kind of tuning the original sample used. So I can correct here the tuning.
[00:08:00] I can normalize the volume, which is nice. Use a bit of velocity here. And we can use
[00:08:10] the mod wheel actually to grow these branches to turn it into a different sound while we
[00:08:18] are playing. This is super massive here. The free verb. So let's use the mod wheel.
[00:08:42] It's really nice that you can turn these sounds into some atonal mess here. But you can also
[00:09:00] bring down the atonality.
[00:09:25] More release. It's too much reverb. I have to experiment with this a bit more. But it's
[00:09:45] really nice to play around with this. Also here with the branches to create different
[00:09:52] versions of the same sound and drift into a tonality. I guess there's a lot of possibility
[00:09:59] there. I have to give it more time. So let me figure this out. So just a tip from me.
[00:10:07] Just head over to the website of synthplan 2. Link is in the description below. Try it
[00:10:13] out and let me know in the comments what you think about this new feature, GenoPatch.
[00:10:19] I think it's a future. Thanks for watching. Leave a like if you liked the video. Subscribe
[00:10:23] to the channel. See you in the next video. Bye.
[00:10:26] [BLANK_AUDIO]