Dawesome MYTH

[vurt]

I lost my heart in Taumatawhakatangihangakoauauotamateaturipukakapikimaungahoronukupokaiwhenuakitanatahu

that came up in a crime show the other day.
some messed up shit too.

[Ah_Dziz]

So it sounds like you're saying (I realize you didn't say this), using a proprietary analysis method, the sound is split into 4 FFT (or something similar) arrays, which represent 4 features that your ML recognizes, or picks out, as layers. During the analysis and "splitting" the overall fundamental pitch is estimated and flattened. which are then resynthesized via iFFT (or something similar) in a way optimized way to interact with the iris controls and the envelope system you are using. Making each layer a 1024 bin static FFT which is enveloped via the 500 point envelope.

Is there a direct control to mute or unmute the layers individually? All of the iris level controls had effects on the sound which were very similar to relatively straightforward manipulation of an fft array, and I'm guessing some adjust the relative levels of the four layers, but listening to each "layer" in isolation and what's happening at that level as each "iris control" is manipulated would be very useful.

I'm personally not a big fan of "black box" type systems for my sound design. I usually like to go in with a plan in mind as I write and do my sounds simultaneously. I did get some interesting results dropping random short sounds on the iris and using the built in resampling of modulations based on the default sine wave. This was the only way I could get things working in a way that felt "controlled". I do very much appreciate the description above as ( my speculative rambling aside) It did better at explaining some of the behavior of the iris that was a bit puzzling for me, than each of the bits of info would have alone.


I guess I'll probably end up with this not really being for me. Having the ability to change the analysis (splitting rules) and resynthesis settings or even methods is more my speed. It's definitely a very capable engine. I'll go back to it on my real studio rig sometime and see if I feel differently when it's not maxing out my poor laptop's cpu.


Edited because I'm dumb.

[revvy]

I lost my heart in Taumatawhakatangihangakoauauotamateaturipukakapikimaungahoronukupokaiwhenuakitanatahu

that came up in a crime show the other day.
some messed up shit too.

It blew chunks at the time, but I'm pretty much over it now.

[databroth]

So it sounds like you're saying (I realize you didn't say this), using a proprietary analysis method, the sound is split into 4 FFT (or something similar) arrays, which represent 4 features that your ML recognizes, or picks out, as layers. During the analysis and "splitting" the overall fundamental pitch is estimated and flattened. which are then resynthesized via iFFT (or something similar) in a way optimized way to interact with the iris controls and the envelope system you are using. Making each layer a 1024 bin static FFT which is enveloped via the 500 point envelope.

I'm not sure if pitch is estimated or flattened"
this is why C3 works well, because I found other pitches (especially other octaves) would end up shifted as a result, so I can't speak to any pitch estimation or flatening
also I believe it's more than 1024 bins, I'm not exactly certain on the specific number, I just know it's more than 1000 (I think it's less than 10k too)

Is there a direct control to mute or unmute the layers individually? All of the iris level controls had effects on the sound which were very similar to relatively straightforward manipulation of an fft array, and I'm guessing some adjust the relative levels of the four layers, but listening to each "layer" in isolation and what's happening at that level as each "iris control" is manipulated would be very useful.

the "WTF" control lets you crossfade between layers, there is no way to mute just one layer, but you can sorta "solo" one layer with "WTF".

I'm personally not a big fan of "black box" type systems for my sound design. I usually like to go in with a plan in mind as I write and do my sounds simultaneously. I did get some interesting results dropping random short sounds on the iris and using the built in resampling of modulations based on the default sine wave. This was the only way I could get things working in a way that felt "controlled". I do very much appreciate the description above as ( my speculative rambling aside) It did better at explaining some of the behavior of the iris that was a bit puzzling for me, than each of the bits of info would have alone.

Totally understand, but you do understand that every synth is "black box" to some extent
especially things like reverbs and chorus don't give you full control over each individual APF and delay line.
I do get wanting further control over the oscillators, but this is for most people very tedious
could you imagine thousands of partials with individual levels and adjustments
there's wavetable synths out there, and Melda has a synth that gives you loads of control over each and every frequency.
But for Myth, the Iris are only a starting point, they are cool and quite powerful
but it's not meant to be the sole and only purpose, there is a whole modular environment below with many interesting effects and modules.
it's not that the Iris aren't fun, they are, and are quite cool, it's just only the beggining.

I guess I'll probably end up with this not really being for me. Having the ability to change the analysis (splitting rules) and resynthesis settings or even methods is more my speed. It's definitely a very capable engine. I'll go back to it on my real studio rig sometime and see if I feel differently when it's not maxing out my poor laptop's cpu.

well, I suggest trying it, and this is all why I don't think it's important to fixate on the Iris
as Myth is more than just the Iris
I feel nervous about saying this, as I don't want to be attacked again
I'm not saying the Iris don't exist, or they aren't important at all
I'm only saying they are not the whole of Myth, and Myth has much more to offer beyond that
sorry to be so tedious with my language, I just want to make sure I am in no way misconstrued

I think picking analysis types and splitting results on an atomic level would make Myth a completely different synth, it'd distract from the fun of exploring modules and playing with the various Iris transformers

the Iris are an interesting way to get unique oscillator tones from your existing sample collection
the reason I stress (or I guess "crusade") that Myth isn't a sampler, is because there is an impression that the iris are meant to be a way to manipulate your samples or sound exactly like the source material.

in this sense, it is reminiscent of wavetables, MANY MANY samples will not sound good when resynthesized as a wavetable, but a wavetable can also get you fun new oscillators from your samples. The Iris' being based on bins rather than harmonics allows for them to replicate atonal sounds and a different set of samples compared to wavetables.

if it's not for you, totally understand, but I would suggest taking a look at everything else that is offered in Myth and seeing if you might have overlooked other features

sorry for the long reply, please don't take the length of it as any sign of anger or argument, just trying to address your concerns

[Ah_Dziz]

Words, etc.

Hey man. I very much appreciate your input.

As far as my "technical interpretation", estimating and flattening the pitch definitely seems to happen. Not a single sound I dropped in an iris had any drift in pitch or even a noticeable drift of pitch between layers, and the "estimation" is just picking the fundamental for playback. The big " use this control if we didn't estimate the pitch correctly" control would seem to point to that happening. The rest of my blathering was just using FFT in place of the many other forms of analysis that use linear frequency. I do realize I'm a bit of an outlier wanting to have control over analysis/ resynthesis size, windowing (may not even be a resynthesis consideration here as the spectrum is static outside of amplitude scaling). Anyway this is all just my personal opinion/ preference etc. I think it's a cool synth, but isn't really my cup of tea.

I gave it a go for the better part of a day. The synthesis engine and modularity are both right up my alley and I like the use of "convolution" (or whatever particular method is used) in the resonators. I dig the effects, filters, routing, support for MPE, etc etc. It's a very nice synth. I just already have many many nice synths and even without the Iris in use it's pretty heavy on my CPU.

I'm also trying to be clear that I'm not shitting on the synth at all. A different resynthesis approach is always something I'm interested in and I gave it a go. The flexible engine is always something I love and it delivered great sounds. Also the presets are very very nice, but that's not really something I need. I'll definitely give it another go on my big rig instead of a laptop. I'd also recommend it to folks who want a big complex synth with lots of presets. I've been piling up VSTs since they were released and at this point, I'm trying to be a bit pickier about buying stuff that's not filling a direct need I have and myth doesn't at the moment. I did finally demo and buy Novum though so, bonus.

I definitely appreciate your input on here though. You're doing a good job for Dawesome. Don't let any of the lunatics around here get you down.

Peace
JJ

[databroth]

I'm also trying to be clear that I'm not shitting on the synth at all.

Totally, didn't get that vibe at all, just had to be concise for others.

and yes, I think it is important to be picky about plugin choices, as a reviewer, I cover everything, but that is not to suggest everyone needs everything, it's just to give a fair shake to each option.
(I can't stand reviewers who only cover a few synths, that all happen to be the "best ever". How can you say it's "the best" when you haven't even covered alternatives?)

I also like to get into the nitty gritty with synths, and fine tune controls to get thorough sound design. I even asked for some more advanced editing, but it was important to Dawesome to keep things streamlined and fun. Already, as the Iris are, they are a bit much for some users, so I believe this was a wise choice, even if it alienates some more advanced outliers.

[kraster]

One thing I'd like to see:


I'd love to have a DAW start sync for the LFO. Right now it's either retriggering with each note or just free running.

If you have an ARP running the LFO will retrigger with every note. If you turn retrigger off then it's just at a random.

One use case would be you had an arp pattern but you want to modulate a parameter over time (e.g. Like a Filter cutoff with a triangle wave). As it is the LFO just retriggers or just radomly picks up when you pres play on the daw.

[Ah_Dziz]

Well, since you work for/ with them now, see if they would be into adding an "advanced import" right click option or something. It's crazy how different things can sound when analyzed at a different size or even window overlap amount.

But I'll just leave this be until I have other questions about Myth in it's current state.

[databroth]

The ability to turn off the auto snap on the LFO when you get near the bottom or top of the Y axis when editing points.

hmmm? I don't seem to have this behavior, there is no snap for me at top/bottom
only in the middle, which you can bypass by deactivating bipolar mode.

are you sure it is snapping to the top and bottom areas?

I'd love to have a DAW start sync for the LFO. Right now it's either retriggering with each note or just free running.

If you have an ARP running the LFO will retrigger with every note. If you turn retrigger off then it's just at a random.

One use case would be you had an arp pattern but you want to modulate a parameter over time (e.g. Like a Filter cutoff with a triangle wave). As it is the LFO just retriggers or just radomly picks up when you pres play on the daw.

Yes, I requested transport sync as an option for this reason
it is tricky to add this in as another button, and changing the behavior of the "sync" button would potentially mess up people's previous work.

I'll mention this again and propose some potential solutions, it's hard to say if/when this will be added as there are many requests already

[kraster]

The ability to turn off the auto snap on the LFO when you get near the bottom or top of the Y axis when editing points.

hmmm? I don't seem to have this behavior, there is no snap for me at top/bottom
only in the middle, which you can bypass by deactivating bipolar mode.

are you sure it is snapping to the top and bottom areas?

I deleted that! You're too quick!

I noticed that it snaps to the equivalent on the other side. Kind of a mirror snap. (Editing a triangle wave, for example). Which is fine!

As for the LFO sync thingy I like how it's done in Minimal Audio's current. If it's synced to the host then it restarts with the daw. If it's not synced to the host it free runs. So you have the behaviour kind of "baked in" to the mode, if you know what I mean.

I think Phase Plant does something similar.

[databroth]

The ability to turn off the auto snap on the LFO when you get near the bottom or top of the Y axis when editing points.

hmmm? I don't seem to have this behavior, there is no snap for me at top/bottom
only in the middle, which you can bypass by deactivating bipolar mode.

are you sure it is snapping to the top and bottom areas?

I deleted that! You're too quick!

I noticed that it snaps to the equivalent on the other side. Kind of a mirror snap. (Editing a triangle wave, for example). Which is fine!

As for the LFO sync thingy I like how it's done in Minimal Audio's current. If it's synced to the host then it restarts with the daw. If it's not synced to the host it free runs. So you have the behaviour kind of "baked in" to the mode, if you know what I mean.

I think Phase Plant does something similar.

ahh my bad
yes, this is why I brought up "sync" mode in regards to transport
if we changed this behavior it could mess up previous projects/ presets
so it's rather "too late" for this unfortunately

I'll propose some ideas to Dawesome

[Ah_Dziz]

I don't have Myth in front of me but is there a "phase control on the lfo? If so, you could use a combination of key reset and phase offset via a cc to get the lfo behavior you want. A bit of a hassle but it would work. Having a robust selection of clocking and reset behavior for LFOs is definitely the move.

Edit: or just use automation/ host modulation if it's available.

[databroth]

I don't have Myth in front of me but is there a "phase control on the lfo? If so, you could use a combination of key reset and phase offset via a cc to get the lfo behavior you want. A bit of a hassle but it would work. Having a robust selection of clocking and reset behavior for LFOs is definitely the move.

Edit: or just use automation/ host modulation if it's available.

there is phase control, but currently there is no transport lock, which is what they are looking for
a few issues
A) there's no way to determine the "first" note
B) there's no way to determine when the Daw starts
C) the arps are sending midi to Myth, so there's no way to determine the difference between an arp note and a piano roll note

it is a contingency I've brought up with Dawesome before, and one I hope we find a solution for

[kraster]

I don't have Myth in front of me but is there a "phase control on the lfo? If so, you could use a combination of key reset and phase offset via a cc to get the lfo behavior you want. A bit of a hassle but it would work. Having a robust selection of clocking and reset behavior for LFOs is definitely the move.

Edit: or just use automation/ host modulation if it's available.

There is a clunky way to do it.

If loop and retrigger is off then it will start from the beginning of the LFO when you press play so you can automate the loop to come on once you hit play then it will be phase locked with the DAW's clock. But you have to disable it when you stop so it's a fudge.

The issue is that the LFO still runs when transport is stopped if it's in loop mode so when you press play it doesn't reset the phase. It will just randomly pick up at whatever stage of the LFO is on.

I could use an external modulator in Ableton but again it's not ideal.

It's an issue with a lot of synths, to be honest. Particularly synths with Arps and sequencers.

And for me personally, it's not some edge use case. I often want the LFOs to work independently and predictably while locked to the host's clock. e.g. A filter sweep using a saw wave or triangle that's locked to the tempo and always starts in the same place..

Minimal's Current does it well and Avenger 2 has good options for resetting the phase of the LFO based on the DAW start.

[kraster]

One use for Myth that I don't see mentioned a lot is that it's one of the best drum loop manglers in existence.

It's also extremely easy to sync a loop to tempo. Just use the "Create and use new LFO" command after dragging a loop to an Iris, set it to whatever bar division you want and adjust the start and end points so it loops in time with the DAW.

From there you can really go to town with the various processors, filters and FX.