I had a request for my "KRC Default" and "Drone" patches demoed in this video, so I added a link to those in the description. ☝I can't redistribute the "Calling" .utu, but I've modified these so they use one of the factory partial files. Just open your own partial files and get your drone on! 🙂
I really liked it when you said that you also bought Sumu because you want to support this kind of development. It makes me feel less silly because I have similar motives often.
@HJPhilippi, thanks for watching! Along similar lines, between recording this video and my last one, I subscribed to Chris from Airwindows's patreon at the $6.66/month level as, now that I've discovered him, I think what he's doing is fascinating and I've gotten at least that value from his plugins already. (Related: Anybody that wants to show similar appreciation for what I do is welcome to purchase KRC Mathwaves whether they are actually interested in the world's largest collection of wavetables or not! 😆)
Thanks for watching and for the kind comment! I didn’t even really get to the pulse generator, scope, and other stuff. In a future video though, so like and subscribe if you haven’t yet!
Sumu looks dope! I especially love the unique modulation options. I’ve been thinking a lot lately about resynthesis/additive approaches. I love my Make Noise Spectraphon on the hardware side of things but I haven’t really found a resynthesis tool I really like in vst format. I’ll definitely be downloading the trial to play more with sumu. Also a big air windows fan. Thanks for running through this Keith 🤘🏻
I think it’s super interesting and though I’m kind of (ab)using it like a traditional polysynth in much of the video, it’s more like a selected set of hardware modules. The voices are actually free-running and always computing, hence the high CPU usage (and this is covered in the manual).
@@Phantom_Communique I suspect plugins like this are single-threaded. I don't know how it might respond to something like threadripper, but I'll be interested to hear about it!
A very interesting piece of software that I plan to check out - thanks! I find Vutu absolutely astonishing, to be honest. I work in software but I'm not a DSP expert. I've written code to do Fourier processing and stuff like that, and I know how much number crunching is involved. Seeing Vutu analyse samples in a second or two and have them ready to play made my jaw hit the floor. It's like witchcraft! HOW do they do it that fast?
@@InsidesAndOutsides check out the link to the Loris resynthesis technique used in Sumu/Vutu. It’s pretty clever and I agree that it’s surprising how well and how fast it works. (Part of the magic is that it makes no attempt to be storage efficient. The analyzed version of a sample is larger than the audio .wav file, not just because of the transform from binary to JSON/text, but because it’s like a description of how all of the harmonics are moving over time.) Anyway, it’s a totally different approach to representing sound that combines temporal and harmonic information.
@@kcrosley Thanks for the tip - I will put that on my reading list! It's hit-and-miss with me and technical papers. I do have a degree in maths, but that was in the 1980s and I'm extremely rusty :) I'll give it a go, though.
@@kcrosley By the way, have you seen a plugin called VISCO? Polarity Music reviewed it recently, and I can't help wondering whether it's using a similar approach.
Replied (briefly)! Look for Keith in your inbox. We should chat soon. Look for email from Keith. (Sorry you had to ping me here to get my attention. I’m just that sort of person. 🤷♂️)
I’m demoing on a high-spec but not state-of-the-art, laptop. Note that I’m abusing Sumu in a sense in this demo as I’m sometimes (mostly) using it as a standard polysynth. But it’s not that… it’s oscillators are free-running and it consumes CPU even when not making audible sound. Keep in mind that this sort of technology is traditionally used OFFLINE and not in real-time!
I was excited by the idea of Sumu and downloaded it as soon as I received the email that it was available for testing. Sadly, it became clear that it was even more of a CPU killer than Myth. I was able to keep Myth under control if voices were generally limited to 12 and I didn't overdo the FX side of things. Sumu killed my processor using just 4 voices and it doesn't really have much of an FX side as it is. If, as you suspect, there is limited room for improving CPU efficiency, then Sumu is just not an option for me.
Hey @mdavid200ify, thanks for watching! I suspect that we might see some sort of eco or low-res mode come to Sumu so that at least one could do higher polyphony stuff and get an idea of what one is doing. But yeah, it seems that this is definitely something that needs freezing/printing in most cases. I think in the current version, the oscillators are (often? always?) computing sort of like they are free running? (I'm not sure...)
Hey @eren3390, thanks for watching! Totally understand cost concerns -- when I went to purchase Sumu, Madrona offered (and still does) a bundle deal on all of their synths at a steep discount and I was like... 🤔... but haven't pulled the trigger on that! If you haven't snagged Airwindows Consolidated yet, it's "pay what you want/can", so you could at least get that! 🙂 www.airwindows.com/
It’s in “early access” and you can install it in demo mode and/or license it to remove the watermarking (it fades everything to noise every few seconds in demo mode). I talk about all that in this video…
Did you watch the part of my video about Vutu? It works great for vocal samples, but it’s helpful to understand what the controls do. (See also link in description to Madrona’s own Vutu walkthrough.)
Wow, I couldn't use two instances of this synth at the same time without slowing everything to a crawl. I hope they are able to work on it's development to where it could be a 'usable instrument' to, you know, write some serious music. This, in opposition to it being a neat sounding toy to simply grab a sample or two from...which is how it feels, and works for me in its current state.
@rm3950, thanks for watching! I do think we'll see some sort of low-res/eco mode for Sumu. It looks to me like the focus with Sumu to date has been simply getting the thing working in a stable fashion and various niceties and small additions are the next phase of the roadmap. (I'm not implying that "making the instrument useful for a wider audience" is simply a "nice to have", of course!)
@@kcrosley Seems odd, that for $120, they're selling it in a 'non-working unstable fashion'. If, indeed, they are "getting the thing working in a stable fashion". Whatever the case, selling it in its current state kind of leaves a bad taste in my mouth, due to it being next to unusable. Especially when use with other VSTs like Hexel, or sequencers, let alone 10 other tracks. This video and your Dawesome/Myth video sort of confuse me. As you speak of Myth as 'needing more', though it works perfectly for many people. In all honestly, I had Myth running fine with highly demanding VSTs, such as Slate and Ash instruments. So you boast a non-working VST, and demand more from a working VST. I wonder if Sumu would sound as good if it was as usable/workable as Myth? I also wonder if some compromises had to be made with Myth? As you could probably surmise, i'm looking at VSTs to make music, as opposed to how they are resynthesizing samples. I look at VSTs in the same way I look at real instruments. They have to work, for me to be able to work. IF not, out the window they must go.
@@vvcv__00 I totally get where you're coming from, but it's not that Sumu "isn't working". It's extremely CPU intensive. I misread your original comment a bit, interpreting what you said as "two voices", not "two instances". As you can see in the video, I wouldn't be able to run two instances of Sumu simultaneously (without freezing/printing one) on my machine either (neither do I think my machine can run 16-voice Sumu patches). It's totally fine if that makes Sumu a non-starter for your use cases. Personally, I find it extremely interesting and, like I say, I'm interested in encouraging unusual synthesis developments like this, so I voted with my wallet as they say. The techniques being explored in Sumu haven't really been available natively before. There are things like Kyma, but that requires dedicated (and rather pricey) hardware. It's definitely pushing the limits of what is possible on modern hardware. (But ya know, Moore's law and blah, blah, blah.) I can only speculate about what further optimizations might be possible as Sumu is further developed. One thing I didn't really talk about in this video is that Sumu is less like a traditional synth voice and more like a curated set of (weird? funky? perhaps "WTELF?") modular synth components (which seems to be the case with all the Madrona Labs stuff, AFAICT). And I also did not point out this -- and it's a very important point -- concept from the manual, which reads: "I’m not playing any notes, so why is Sumu eating my CPU time? Sumu has *free-running oscillators and processors* that are updated whenever your DAW is processing audio, so those eat up CPU at a rate that matches your chosen number of voices in the INPUT module. Just like on a modular synth, you can simply turn up the level dial on GATES to hear the oscillators, even if no notes are playing." So, this thing *is not* like a "normal" polysynth VST, and if we want to use it in that way, it does take us back a bit to those early days of in-the-box production, when you'd have to freeze/print a track before moving on to the next thing. Again, it's OK if that makes it a non-starter for what you want to do with it. As for Myth, to understand a little more about where I'm coming from with respect to that instrument, see my reply to @BF-up5xw elsewhere in the comments here. (I tried to get a link, but don't see a way to do that. Thanks, Google! And sorry BF-up5xw for the coincidental "at"!) In terms of "boosting" one thing in favor of another, I use this channel to just talk about things that are on my mind/find interesting and it's not really a "review" channel -- it's more about discovery plus a healthy dose of self-promotion. (Though people seem to enjoy the overview-type of content quite a bit and I find that gratifying!) Thanks for watching (and thanks for attending my Ted Talk 😉)!
@@kcrosley "Ted Talk".. ahaha, too funny!. I gotcha, regarding your point of view behind your content. But, to let you know, from a viewer perspective, once you put content 'out there', it no longer belongs to you. I look at it as a musician/artist putting their work out 'into the public'. Once that happens, the public brings all of their baggage/experience/wants/etc. to the content, be it a painting, live show, or a video. As an example, I was interested in your videos from the perspective of a musician, not interested in a lot of tech stuff, looking for a little extra something 'behind' the GUI of the instruments i'm interested in. So, you may/may not, want to consider information less techie people want to get/take away from your content. As i'm sure they'd tune in. Moreover, though your videos may contain much information on the tech side of things, there are certainly some crossover bits of information that are interesting for those coming from the tech, AND musical perspectives. Again, once your video is out, your control over who watches, and what they bring to the table, is out of your hands. :D Anyway, i'll try to catch your video once Sumu is fully developed.
Myth doesn't aim to reproduce samples through synthesis. It's use of resynthesis is intended to create new sounds. You might be better off comparing Sumu's resynthesis to Novum; though even then the aims are different. Myth and Novum are created by the same developer, Dawesome. Really, though, none of these synths are trying to reproduce the original samples so much as use them as complex starting points for synthesis. Thanks for the exploration of these synths, though, even if I have reservations on these points.
Hey @BF-up5xw, thanks for watching! One of my issues with Myth (that I covered in my original overview, ruclips.net/video/Q7cvoHqZSSs/видео.html) is that Dawesome's own materials about that synth (both marketing and documentation), says stuff like: "MYTH offers a paradigm shift: drag and drop audio and it will be re-synthesised as an IRIS. The inherent variations and richness of the sample can now be exploited with the TRANSFORMER dials." But in practice, the resulting variations have a tendency to be quite "samey", and in an unpleasant way. Experimentation there feels like *work*, rather than *fun*. They go on to say stuff like, "Or you take a sample and then browse through presets to put them over your sample like a dress." My take on that is it's more like Myth takes a strong pimp hand to your pretty little sample and puts it in a rather ratchet dress, and then you can cover up the bruising with some very pretty FX. I just feel that the best parts of that synth are substantially weakened by the core sound generation concept (and when I say "concept", I mean more the actual execution). I'm not alone in that feeling. 🤷♂️ There's nothing wrong with liking Myth if it does things that you like, and it does have some very interesting features (like the MODAL filter), which I covered in a separate video, as it basically gets no love in the factory patches (which is a little mind-boggling). For anyone who's not seen it, that video is here: ruclips.net/video/Jm69svhy43M/видео.html. There's also always the looming specter of pricing and value-for-money. I've got no issue with the economics of "boutique" or more-or-less indie developers needing to price products in a way that affords them a living. However, looking at something like Myth, which to my mind is a letdown in terms of its primary feature (the whole Iris thing), when offered at $179... Well, what's left?... A pretty simple VA type synth and exciter/resonator type physical modeling and some admittedly great-sounding FX. When compared to something like Korg modwave native, available for $149, which at the end of the day is an excellent, fully specced synth architecture (the best since E-mu's E4/module architecture) that's great for VA (but with infinitely-variable waveforms), also does wavetable stuff, has **sampling**, and its own complement of great FX (like Myth, only available in the voice, not as an FX-mode plugin), and **successful** deep/experimental features like the mod sequencer, it's no contest. (If you're not into wavetables, you could make a similar argument for wavestate.) (☝️ Aside: Man I hate RUclips's reply formatting.) That's why one of my main bits of feedback about Myth was (and again, I made those comments as a **customer**) "just add a sample playback module to this thing". It needs something more to be competitive over what it offers right now. (Also, "who is this for?" is a key question with any product. Myth really has no point of view in that department.) I also note that (perhaps this is an error) Myth has unceremoniously dropped from the Traktion home screen. 🤷♂️ My channel isn't really a "review" channel, so I don't go deep into issues like that. And in the current Sumu video, since the thing is still in development, I don't talk much about any reservations I have about it as we don't really know what the ultimate product will be like. (Though most of my surface-level feedback is already addressed in their planned additions for the release version.) I just talk about my own products, synthesis topics that I find interesting, and touch on how one might use my products with those interesting things.
@@kcrosley Thank you for such a great reply. You have a lot of really interesting thoughts here. It is certainly true that Myth works better with very gnarly samples and that the resynthesis algorithm doesn't do as much of interest with cleaner, tamer ones.
@@kcrosley I think you are really harsh on Myth and I don't know really know why as someone who is clearly very knowledgeable and enthusiastic about synths. Even the product description that you cited is pretty much accurate. Put a rhythmic sample or non periodic sample in and you can indeed exploit the variations in it in any number of ways. Not every sample will work but a lot do. It's only "samey" because you are expecting the Irises to do all the heavy lifting. Saying it's "work" and "not fun" is really subjective. I think Myth is tremendous fun. Something being unpleasant on its own is par for the course with synths. A high pitched pulse wave or FM based oscillator on its own can make your eyes bleed. But taming it with some other processing can lead to great results. The Irises are there to extract temporal information from a sample that can further modified by the wealth of processors downstream from it. It's not like you're just stuck with the Irises on their own. The "who is it for" is a weird take. Who is any synth for? Who is Sumu for? People who want to experiment with a particular feature set. I feel like you're being really dismissive of the synth because you are disappointed in one aspect. There's a lot there. Myth has an array of sound generating and sculpting features, arps, chord generators, pitch quantizers, Various weird FM modulators, comb filters with different algorithms, resonators and modal filters, Multiple filters, AM, unique distortions, extensive and easy to use modulation, routing options and great fx. Taken together they form a pretty formidable toolset. And if you can't get unique and weird sounds from it I don't know what to say. It can do a form of resynthesis, VA style and physical modelling. It's very in keeping with Dawesome's general oblique takes on instruments. All of his Synths are a twist on various technologies. I think the mistake is thinking that it is exclusively a resynthesis based synth. It uses a form of resynthesis as a component in the overall sound design with a metric ton of options to sculpt and incorporate to something bigger. I think it's a wonderful synth.
@CapriciousBlackBox, while there's almost always small opportunities for optimization, I suspect high CPU usage is the nature of this particular synthesis beast. It's sort of amazing this is available as a real-time process at all. Basically, each voice is something like 64 sine and noise oscillators (representing each partial) and amplitude envelope(s), so there's a lot of computation going on. And I didn't really mention this, but the control data generated by Vutu is very extensive. (Some approaches to resynthesis are essentially ways of compressing musical/sound data, but not in this case, where the approach is a relatively high-fidelity reproduction. For example, that "A Simple Break!" sample is 233K as .wav, but 767K as .utu.)
@@kcrosley Very interesting...and if I understand it correctly I would agree with you re: amazing that this is available as a real-time process at all. I'll have to look into this VST for sure. Cheers
@@kcrosley cool things for writing I highly suggest you check out Tone2 synthesizers since they have a lot of options including resynthesis and they have a different approach hope to see you in the next video
I had a request for my "KRC Default" and "Drone" patches demoed in this video, so I added a link to those in the description. ☝I can't redistribute the "Calling" .utu, but I've modified these so they use one of the factory partial files. Just open your own partial files and get your drone on! 🙂
I really liked it when you said that you also bought Sumu because you want to support this kind of development. It makes me feel less silly because I have similar motives often.
@HJPhilippi, thanks for watching! Along similar lines, between recording this video and my last one, I subscribed to Chris from Airwindows's patreon at the $6.66/month level as, now that I've discovered him, I think what he's doing is fascinating and I've gotten at least that value from his plugins already. (Related: Anybody that wants to show similar appreciation for what I do is welcome to purchase KRC Mathwaves whether they are actually interested in the world's largest collection of wavetables or not! 😆)
NICE video! Cool synth. 😍
very good and patient explanation. Subscribing.
@@cachelesssociety5187 thanks for watching!
Dude… I love your channel!
Thanks so much for watching!
This is great synth and a great video. Thank you! 👏🏻👏🏻👏🏻
This tutorial should be official-most excellent and clear walkthrough!
Thanks for watching and for the kind comment! I didn’t even really get to the pulse generator, scope, and other stuff. In a future video though, so like and subscribe if you haven’t yet!
@@kcrosleySubbed and looking forward to the follow up videos.
This was an EXCELLENT video and much needed to help make sense of this synth. The run through of Vutu was also really appreciated! Thank you!!!
Thanks for watching and I’m glad to hear that it was helpful!
Sumu looks dope! I especially love the unique modulation options. I’ve been thinking a lot lately about resynthesis/additive approaches. I love my Make Noise Spectraphon on the hardware side of things but I haven’t really found a resynthesis tool I really like in vst format. I’ll definitely be downloading the trial to play more with sumu. Also a big air windows fan. Thanks for running through this Keith 🤘🏻
I think it’s super interesting and though I’m kind of (ab)using it like a traditional polysynth in much of the video, it’s more like a selected set of hardware modules. The voices are actually free-running and always computing, hence the high CPU usage (and this is covered in the manual).
Lucky for me I just upgraded my pc to a threadripper pro workstation build … probably overkill for audio but Sumu might be a good way to test it out 😎
@@Phantom_Communique I suspect plugins like this are single-threaded. I don't know how it might respond to something like threadripper, but I'll be interested to hear about it!
Hey, it's my old pal Keith! Good to see this review....I had not previously heard of Sumu; thanks for bringing it to my attention.
Thanks! This was needed, its hard to understand this synth. You make it a bit easier
Thanks for watching and I’m glad I could help!
A very interesting piece of software that I plan to check out - thanks! I find Vutu absolutely astonishing, to be honest. I work in software but I'm not a DSP expert. I've written code to do Fourier processing and stuff like that, and I know how much number crunching is involved. Seeing Vutu analyse samples in a second or two and have them ready to play made my jaw hit the floor. It's like witchcraft! HOW do they do it that fast?
@@InsidesAndOutsides check out the link to the Loris resynthesis technique used in Sumu/Vutu. It’s pretty clever and I agree that it’s surprising how well and how fast it works. (Part of the magic is that it makes no attempt to be storage efficient. The analyzed version of a sample is larger than the audio .wav file, not just because of the transform from binary to JSON/text, but because it’s like a description of how all of the harmonics are moving over time.) Anyway, it’s a totally different approach to representing sound that combines temporal and harmonic information.
@@kcrosley Thanks for the tip - I will put that on my reading list! It's hit-and-miss with me and technical papers. I do have a degree in maths, but that was in the 1980s and I'm extremely rusty :) I'll give it a go, though.
@@kcrosley By the way, have you seen a plugin called VISCO? Polarity Music reviewed it recently, and I can't help wondering whether it's using a similar approach.
Good demo, thanks.
Thanks for watching!
Great video! Sent you some email! 😀
Replied (briefly)! Look for Keith in your inbox. We should chat soon. Look for email from Keith. (Sorry you had to ping me here to get my attention. I’m just that sort of person. 🤷♂️)
And I sent you a longer reply today. Thanks for reaching out!
Looks cute,yet from what I am reading, you will need a monster of a workstation dedicated to experiment with Sumu..
I’m demoing on a high-spec but not state-of-the-art, laptop. Note that I’m abusing Sumu in a sense in this demo as I’m sometimes (mostly) using it as a standard polysynth. But it’s not that… it’s oscillators are free-running and it consumes CPU even when not making audible sound. Keep in mind that this sort of technology is traditionally used OFFLINE and not in real-time!
I was excited by the idea of Sumu and downloaded it as soon as I received the email that it was available for testing. Sadly, it became clear that it was even more of a CPU killer than Myth. I was able to keep Myth under control if voices were generally limited to 12 and I didn't overdo the FX side of things. Sumu killed my processor using just 4 voices and it doesn't really have much of an FX side as it is.
If, as you suspect, there is limited room for improving CPU efficiency, then Sumu is just not an option for me.
Hey @mdavid200ify, thanks for watching! I suspect that we might see some sort of eco or low-res mode come to Sumu so that at least one could do higher polyphony stuff and get an idea of what one is doing. But yeah, it seems that this is definitely something that needs freezing/printing in most cases. I think in the current version, the oscillators are (often? always?) computing sort of like they are free running? (I'm not sure...)
sadly, it’s expensive :( will just use the demo for now 😂
also: great video!
Hey @eren3390, thanks for watching! Totally understand cost concerns -- when I went to purchase Sumu, Madrona offered (and still does) a bundle deal on all of their synths at a steep discount and I was like... 🤔... but haven't pulled the trigger on that! If you haven't snagged Airwindows Consolidated yet, it's "pay what you want/can", so you could at least get that! 🙂 www.airwindows.com/
@@kcrosleythat‘s amazing, airwindows has some amazing stuff indeed! especially the filters are amazing :) thanks also for the new wavetables 😊
Has it been released I’ve Been waiting
It’s in “early access” and you can install it in demo mode and/or license it to remove the watermarking (it fades everything to noise every few seconds in demo mode). I talk about all that in this video…
So there’s no way to get a clean sample inside of sumu? I used vutu to try and import a vocal sample and it degraded it like crazy.
Did you watch the part of my video about Vutu? It works great for vocal samples, but it’s helpful to understand what the controls do. (See also link in description to Madrona’s own Vutu walkthrough.)
Oh. I would like to try sumu, but it crashes my daw.
Thanks for watching! That’s a bummer. What DAW/platform? (Def let support@madronalabs.com know about that.)
Wow, I couldn't use two instances of this synth at the same time without slowing everything to a crawl. I hope they are able to work on it's development to where it could be a 'usable instrument' to, you know, write some serious music. This, in opposition to it being a neat sounding toy to simply grab a sample or two from...which is how it feels, and works for me in its current state.
@rm3950, thanks for watching! I do think we'll see some sort of low-res/eco mode for Sumu. It looks to me like the focus with Sumu to date has been simply getting the thing working in a stable fashion and various niceties and small additions are the next phase of the roadmap. (I'm not implying that "making the instrument useful for a wider audience" is simply a "nice to have", of course!)
@@kcrosley Seems odd, that for $120, they're selling it in a 'non-working unstable fashion'. If, indeed, they are "getting the thing working in a stable fashion". Whatever the case, selling it in its current state kind of leaves a bad taste in my mouth, due to it being next to unusable. Especially when use with other VSTs like Hexel, or sequencers, let alone 10 other tracks.
This video and your Dawesome/Myth video sort of confuse me. As you speak of Myth as 'needing more', though it works perfectly for many people. In all honestly, I had Myth running fine with highly demanding VSTs, such as Slate and Ash instruments. So you boast a non-working VST, and demand more from a working VST. I wonder if Sumu would sound as good if it was as usable/workable as Myth? I also wonder if some compromises had to be made with Myth?
As you could probably surmise, i'm looking at VSTs to make music, as opposed to how they are resynthesizing samples. I look at VSTs in the same way I look at real instruments. They have to work, for me to be able to work. IF not, out the window they must go.
@@vvcv__00 I totally get where you're coming from, but it's not that Sumu "isn't working". It's extremely CPU intensive. I misread your original comment a bit, interpreting what you said as "two voices", not "two instances". As you can see in the video, I wouldn't be able to run two instances of Sumu simultaneously (without freezing/printing one) on my machine either (neither do I think my machine can run 16-voice Sumu patches).
It's totally fine if that makes Sumu a non-starter for your use cases. Personally, I find it extremely interesting and, like I say, I'm interested in encouraging unusual synthesis developments like this, so I voted with my wallet as they say.
The techniques being explored in Sumu haven't really been available natively before. There are things like Kyma, but that requires dedicated (and rather pricey) hardware. It's definitely pushing the limits of what is possible on modern hardware. (But ya know, Moore's law and blah, blah, blah.) I can only speculate about what further optimizations might be possible as Sumu is further developed.
One thing I didn't really talk about in this video is that Sumu is less like a traditional synth voice and more like a curated set of (weird? funky? perhaps "WTELF?") modular synth components (which seems to be the case with all the Madrona Labs stuff, AFAICT). And I also did not point out this -- and it's a very important point -- concept from the manual, which reads:
"I’m not playing any notes, so why is Sumu eating my CPU time?
Sumu has *free-running oscillators and processors* that are updated whenever your DAW is processing audio, so those eat up CPU at a rate that matches your chosen number of voices in the INPUT module. Just like on a modular synth, you can simply turn up the level dial on GATES to hear the oscillators, even if no notes are playing."
So, this thing *is not* like a "normal" polysynth VST, and if we want to use it in that way, it does take us back a bit to those early days of in-the-box production, when you'd have to freeze/print a track before moving on to the next thing. Again, it's OK if that makes it a non-starter for what you want to do with it.
As for Myth, to understand a little more about where I'm coming from with respect to that instrument, see my reply to @BF-up5xw elsewhere in the comments here. (I tried to get a link, but don't see a way to do that. Thanks, Google! And sorry BF-up5xw for the coincidental "at"!)
In terms of "boosting" one thing in favor of another, I use this channel to just talk about things that are on my mind/find interesting and it's not really a "review" channel -- it's more about discovery plus a healthy dose of self-promotion. (Though people seem to enjoy the overview-type of content quite a bit and I find that gratifying!) Thanks for watching (and thanks for attending my Ted Talk 😉)!
@@kcrosley "Ted Talk".. ahaha, too funny!. I gotcha, regarding your point of view behind your content. But, to let you know, from a viewer perspective, once you put content 'out there', it no longer belongs to you. I look at it as a musician/artist putting their work out 'into the public'. Once that happens, the public brings all of their baggage/experience/wants/etc. to the content, be it a painting, live show, or a video.
As an example, I was interested in your videos from the perspective of a musician, not interested in a lot of tech stuff, looking for a little extra something 'behind' the GUI of the instruments i'm interested in. So, you may/may not, want to consider information less techie people want to get/take away from your content. As i'm sure they'd tune in.
Moreover, though your videos may contain much information on the tech side of things, there are certainly some crossover bits of information that are interesting for those coming from the tech, AND musical perspectives. Again, once your video is out, your control over who watches, and what they bring to the table, is out of your hands. :D
Anyway, i'll try to catch your video once Sumu is fully developed.
Myth doesn't aim to reproduce samples through synthesis. It's use of resynthesis is intended to create new sounds. You might be better off comparing Sumu's resynthesis to Novum; though even then the aims are different. Myth and Novum are created by the same developer, Dawesome. Really, though, none of these synths are trying to reproduce the original samples so much as use them as complex starting points for synthesis. Thanks for the exploration of these synths, though, even if I have reservations on these points.
Hey @BF-up5xw, thanks for watching! One of my issues with Myth (that I covered in my original overview, ruclips.net/video/Q7cvoHqZSSs/видео.html) is that Dawesome's own materials about that synth (both marketing and documentation), says stuff like: "MYTH offers a paradigm shift: drag and drop audio and it will be re-synthesised as an IRIS. The inherent variations and richness of the sample can now be exploited with the TRANSFORMER dials." But in practice, the resulting variations have a tendency to be quite "samey", and in an unpleasant way. Experimentation there feels like *work*, rather than *fun*.
They go on to say stuff like, "Or you take a sample and then browse through presets to put them over your sample like a dress." My take on that is it's more like Myth takes a strong pimp hand to your pretty little sample and puts it in a rather ratchet dress, and then you can cover up the bruising with some very pretty FX.
I just feel that the best parts of that synth are substantially weakened by the core sound generation concept (and when I say "concept", I mean more the actual execution). I'm not alone in that feeling. 🤷♂️
There's nothing wrong with liking Myth if it does things that you like, and it does have some very interesting features (like the MODAL filter), which I covered in a separate video, as it basically gets no love in the factory patches (which is a little mind-boggling). For anyone who's not seen it, that video is here: ruclips.net/video/Jm69svhy43M/видео.html.
There's also always the looming specter of pricing and value-for-money. I've got no issue with the economics of "boutique" or more-or-less indie developers needing to price products in a way that affords them a living. However, looking at something like Myth, which to my mind is a letdown in terms of its primary feature (the whole Iris thing), when offered at $179... Well, what's left?... A pretty simple VA type synth and exciter/resonator type physical modeling and some admittedly great-sounding FX.
When compared to something like Korg modwave native, available for $149, which at the end of the day is an excellent, fully specced synth architecture (the best since E-mu's E4/module architecture) that's great for VA (but with infinitely-variable waveforms), also does wavetable stuff, has **sampling**, and its own complement of great FX (like Myth, only available in the voice, not as an FX-mode plugin), and **successful** deep/experimental features like the mod sequencer, it's no contest. (If you're not into wavetables, you could make a similar argument for wavestate.)
(☝️ Aside: Man I hate RUclips's reply formatting.)
That's why one of my main bits of feedback about Myth was (and again, I made those comments as a **customer**) "just add a sample playback module to this thing". It needs something more to be competitive over what it offers right now. (Also, "who is this for?" is a key question with any product. Myth really has no point of view in that department.)
I also note that (perhaps this is an error) Myth has unceremoniously dropped from the Traktion home screen. 🤷♂️
My channel isn't really a "review" channel, so I don't go deep into issues like that. And in the current Sumu video, since the thing is still in development, I don't talk much about any reservations I have about it as we don't really know what the ultimate product will be like. (Though most of my surface-level feedback is already addressed in their planned additions for the release version.) I just talk about my own products, synthesis topics that I find interesting, and touch on how one might use my products with those interesting things.
@@kcrosley Thank you for such a great reply. You have a lot of really interesting thoughts here. It is certainly true that Myth works better with very gnarly samples and that the resynthesis algorithm doesn't do as much of interest with cleaner, tamer ones.
@@kcrosley I think you are really harsh on Myth and I don't know really know why as someone who is clearly very knowledgeable and enthusiastic about synths.
Even the product description that you cited is pretty much accurate. Put a rhythmic sample or non periodic sample in and you can indeed exploit the variations in it in any number of ways. Not every sample will work but a lot do. It's only "samey" because you are expecting the Irises to do all the heavy lifting. Saying it's "work" and "not fun" is really subjective. I think Myth is tremendous fun.
Something being unpleasant on its own is par for the course with synths. A high pitched pulse wave or FM based oscillator on its own can make your eyes bleed. But taming it with some other processing can lead to great results.
The Irises are there to extract temporal information from a sample that can further modified by the wealth of processors downstream from it. It's not like you're just stuck with the Irises on their own.
The "who is it for" is a weird take. Who is any synth for? Who is Sumu for? People who want to experiment with a particular feature set. I feel like you're being really dismissive of the synth because you are disappointed in one aspect. There's a lot there.
Myth has an array of sound generating and sculpting features, arps, chord generators, pitch quantizers, Various weird FM modulators, comb filters with different algorithms, resonators and modal filters, Multiple filters, AM, unique distortions, extensive and easy to use modulation, routing options and great fx.
Taken together they form a pretty formidable toolset. And if you can't get unique and weird sounds from it I don't know what to say.
It can do a form of resynthesis, VA style and physical modelling. It's very in keeping with Dawesome's general oblique takes on instruments. All of his Synths are a twist on various technologies.
I think the mistake is thinking that it is exclusively a resynthesis based synth. It uses a form of resynthesis as a component in the overall sound design with a metric ton of options to sculpt and incorporate to something bigger.
I think it's a wonderful synth.
sumu is so cool but it bullies the hell out of my cpu
That’s a good way to put it. You can almost hear the Kyma people saying, “We told you you needed an audio coprocessor.”
@@kcrosley Is this something you think can be mitigated in future releases, or will it be "baked-in?"
@CapriciousBlackBox, while there's almost always small opportunities for optimization, I suspect high CPU usage is the nature of this particular synthesis beast. It's sort of amazing this is available as a real-time process at all. Basically, each voice is something like 64 sine and noise oscillators (representing each partial) and amplitude envelope(s), so there's a lot of computation going on. And I didn't really mention this, but the control data generated by Vutu is very extensive. (Some approaches to resynthesis are essentially ways of compressing musical/sound data, but not in this case, where the approach is a relatively high-fidelity reproduction. For example, that "A Simple Break!" sample is 233K as .wav, but 767K as .utu.)
@@kcrosley Very interesting...and if I understand it correctly I would agree with you re: amazing that this is available as a real-time process at all. I'll have to look into this VST for sure. Cheers
Sounds good but the interface is an eye sore.
Thanks for watching! It's definitely has a certain "look" to it, but I do like the way it scales nicely.
Icarus has better resynthesis
Thanks for watching! I'll have to check that synth out sometime. (BTW, KRC Mathwaves can be used in Icarus as well!)
@@kcrosley cool things for writing I highly suggest you check out Tone2 synthesizers since they have a lot of options including resynthesis and they have a different approach hope to see you in the next video