Fathom Synth Development Thread

VST, AU, AAX, CLAP, etc. Plugin Virtual Instruments Discussion
Post Reply New Topic
RELATED
PRODUCTS
Fathom Synth

Post

FathomSynth wrote: Mon Nov 12, 2018 5:47 pm The wave draw expanded view partials dial goes up to 400 in FathomMono64.dll, I just tried it myself.

Which version are you running?

Go to the main page and click on the Fathom logo in the upper left corner and report here the version?
[/quote
Fathon 2.4.0
32 bit
Eyeball exchanging
Soul calibrating ..frequencies

Post

ENV1 wrote: Mon Nov 12, 2018 4:17 pm
FathomSynth wrote: You can't insert a VST instrument into another VST instrument, since this is impossible.
Yes, its possible.

ImageLines DirectWave can do it and DiscoDSPs Highlife can do it too.

(I have the latter and it works great.)


EDIT: Plus there are other examples. Phrazor, besides the standalone, also comes in a VST and VSTi variant and being a host it can (obviously) host plugins. Same for FLStudio, energyXT, and various chainer plugins. So its definitely possible to host plugins within plugins, its been done numerous times.

(DirectWave and Highlife are the only two i know that support hosting specifically for the purpose of sampling the hosted instrument though.)
Also sugabytes obscurium is a vst that can host another vst inside it.
Melda Production & United Plugins
Surface Studio = i7, 32gb, SSD.
Windows 11. Bitwig, Reaper, Live. MTotal.
Audiofuse, Adam Audio monitors + sub, iLoud MTM.
Polybrute, Summit, Pro 3, Tempest, Syntakt, AH2.
Ableton Push 2, Roli Seaboard Block.

Post

jmg8 wrote: Also sugabytes obscurium is a vst that can host another vst inside it.
Also VAZ Modular has a module specially for hosting VST/VSTi plugins.

With this you can use any 'foreign' synths or effects as modules within the modular structure.

Very handy.

Post

ENV1 wrote: Mon Nov 12, 2018 8:13 pm
jmg8 wrote: Also sugabytes obscurium is a vst that can host another vst inside it.
Also VAZ Modular has a module specially for hosting VST/VSTi plugins.

With this you can use any 'foreign' synths or effects as modules within the modular structure.

Very handy.
Another is Nora which can host VSTi's.
For VST midi fx being able to host a VSTi's simplifies the routing in most DAWs.

Most Modular hosts which are VSTi's Oor VST fx can host VSTi and VSTfx
such as Bidule, energyXT and many more.

Post

Version 2.4 ? We are currently on version 2.22

Your Fathom is 18 versions old, please go here and get the latest:

https://www.fathomsynth.com/download/

OK, I stand corrected on holding VST's inside other VST's, however I'm not planning on adding this capability to Fathom any time soon. The sampler will be able to automatically load a folder full of samples, but users will have to handle exporting samples themselves. I'm not saying it's a bad idea, and I can see it would be useful, but there are just much higher priorities currently.

Post

FathomSynth wrote: Mon Nov 12, 2018 9:16 pm Version 2.4 ? We are currently on version 2.22

Your Fathom is 18 versions old, please go here and get the latest:

https://www.fathomsynth.com/download/

OK, I stand corrected on holding VST's inside other VST's, however I'm not planning on adding this capability to Fathom any time soon. The sampler will be able to automatically load a folder full of samples, but users will have to handle exporting samples themselves. I'm not saying it's a bad idea, and I can see it would be useful, but there are just much higher priorities currently.
O.K. just did , limited partials are solved now , but now lots of filters are missing ,polar morph , Zero delay etc...I swear these were available in the previous version (MONO)

Anyway , I still think all the filters lack any bite ...
Also ,I can only load the 32 bit monoMT version ...the other one just crashes (the non MT version )
Does MT stand for microtuning ?
When closing and re-opening the gui always takes some time to load and sucks resources (old laptop here don't ask ) , the gui could be a lot better ..IMHO that speckled surface is not really inviting , and the blue metal led knobs are hideous , especialy when downscaled :cry:
Anyway , I tried to like it ..but I can't ..something just doesn't feel right , my opinion
Last edited by gentleclockdivider on Mon Nov 12, 2018 11:24 pm, edited 1 time in total.
Eyeball exchanging
Soul calibrating ..frequencies

Post

MT is the statically linked build, nothing to do with microtuning. If the normal build does not run on your system and the MT build does then you are running the right version. This is often the case with 32 bit systems which is why I provide the MT build.

Post

Clicking on the gear icon/cog wheel does nothing
No possibility to change any setting.
Eyeball exchanging
Soul calibrating ..frequencies

Post

FathomSynth wrote: Mon Nov 12, 2018 9:16 pm OK, I stand corrected on holding VST's inside other VST's, however I'm not planning on adding this capability to Fathom any time soon. The sampler will be able to automatically load a folder full of samples, but users will have to handle exporting samples themselves. I'm not saying it's a bad idea, and I can see it would be useful, but there are just much higher priorities currently.
How about an auto-sampling feature that uses the DAW for the routing. Fathom would not need to host the VSTi. It would send out the correct midi notes at the correct velocities and record the input.
Then save it all as a multisample.
Melda Production & United Plugins
Surface Studio = i7, 32gb, SSD.
Windows 11. Bitwig, Reaper, Live. MTotal.
Audiofuse, Adam Audio monitors + sub, iLoud MTM.
Polybrute, Summit, Pro 3, Tempest, Syntakt, AH2.
Ableton Push 2, Roli Seaboard Block.

Post

@Fathomsynth

Expanding on your sampler plans....

Since Fathomsynth is already FFT/iFFT based and the Wave editor is already able to do an FFT/iFFT transform from any handdrawn or Bezier drawn oscillator shapes, it is probably a small step to do an FFT transform from any signal / audio source.

That would open the door to importing/extracting Spectra from a variety of sources. Like samples (single cycle wavforms or segmented or "averaged" longer sample material) or even audio input (external, digital or analog via ADC) or even from other digital audio sources/plugins. Without having to save all (sample/audio) source material.

So not so much an audio sampler as in the more common implementations, but an FFT/Spectrum-sampler.

Supplying an FFT/Spectrum library would be a natural extension of this.

And of course an FFT/Spectrum librarian would be very helpful with options to manage/export/import/audition the Spectrum library....

Post

@Fathomsynth

Expanding on the suggested FFT/iFFT Spectrum "sampling" options....and more general towards the amount of control over harmonic content/partials/spectrum...

You could even so far as to give dynamic control over the amplitude (or even phase) of individual partials or groups of partials. That would open the door to true additive synthesis.

For dynamic control over the amplitude of each individual partial you would need as many envelopes as you have partials. Doubling that if you want control over the pahse as well. Managing that kind of complexity (while programming sounds) is quite a daunting task. Even without control over the phase of the partials.

But you could start with linking (groups of) partials to as many Envelopes as you have available and just control the amplitude. This is how my now ancient Kawai K5 Additive Synthesizer (1987) solved reduction of complexity. Mainly given the computational and other technical constraints of its time, by the way (probably running the whole machine from the NEC V30 processor that I found inside).

http://www.vintagesynth.com/kawai/kawaik5.php

The Kawai K5 only works by modulating the amplitudes of the (2 x 64) harmonics (as far as I know). It uses this also for the "filtering" and "EQ". By modulating in the amplitude domain. Thus, you can "line-draw" your own "filter" or "EQ" shapes. You can create filters of "any" slope. The steepest slopes nearing 90 degrees. A "bandpass" can then sweep over individual harmonics.

You could also introduce a skew factor in the time domain of the modulation. In that way you can use a single envelope (e.g. ADSR) to make high frequencies decay faster then lower frequencies (the amplitude of partials that are linked to that envelope). A common phenomenon in many natural sounds ("emulating" a "closing" low pass filter). Or the other way around (emulating a high pass filter moving "upwards"). You could even do non-straight-lineair skew shapes (like a V or upsidedown V or other shapes or shape control) to create even more options.

All this (with or without the skewing) again would also open the door to "true" resynthesis. Not only extract a static spectrum from a source, but also "an approximation of" the amplitude (and phase) modulation of (groups of) partials over time.

I bet you are capable, curious and visionary enough to at least consider some of these suggestions for future versions of your allready marvellous and masterfully made machine.

Marvellous....reminds me....RIP Stan Lee (Dec 28, 1922 - Nov 12, 2018)

Post

@Kwurqx

I really like this idea.

It made me think how cool it would be to import the amplitude information from 1 sample and the pitch information from a different sample. This would create a sort of spectral vocoder morph of 2 samples.
Melda Production & United Plugins
Surface Studio = i7, 32gb, SSD.
Windows 11. Bitwig, Reaper, Live. MTotal.
Audiofuse, Adam Audio monitors + sub, iLoud MTM.
Polybrute, Summit, Pro 3, Tempest, Syntakt, AH2.
Ableton Push 2, Roli Seaboard Block.

Post

@fathomsynth

Another (I guess) technically not to hard to implement idea is to provide Vector Synthesis. Not much more then a mixer that can dynamically crossfade between multiple oscillators/signals. Something that can allready be achieved using multiple interlinked oscilator/modulator combos. But a modulatable mixer would greatly simplify the configuration. And a mixer could also take input from elsewhere in the signalpath/matrix (not just raw oscillators).
https://en.wikipedia.org/wiki/Vector_synthesis

And along the same line is Wave Sequencing. Triggering multiple oscillators in a sequence. With crossfade options. Could also be used as a sort of L/A style oscillator where Attack and Sustain are actually 2 seperate oscillators in sequence.

And more complex, but along the same lines. Morph multiple Spectra by interpolating on the partials level. This is different from just mixing. Imagine Spectrum 1 is a Saw wave spectrum (all partials present, amplitude 1 / rank) and Spectrum 2 is silence (an extreme example to illustrate). A crossfade would just reduce all harmonics in amplitude, lowering the overall volume (you would hear increasingly less of the Saw and more of the silence so to speak). An interpolated path on the harmonics level over the same time period would make it necessary for high amplitude partials to decrease in value faster then lower amplitude partials, therefore changing not only the volume but also the timbre. More generally speaking: different speeds/angles are needed to bridge different amplitude differences between both sets of partials in the same time. You would get a different kind of a timbre morph.

Post

@Fathomsynth

Today I was trying to figure out what the FM capabilities of Fathomsynth actually are. Is it even possible to create/emulate Yamaha DX style multiple operator algorithms? With multiple carrier/modulator combinations? And is an FM feedback loop even possible?

In other words: can I use the matrix as an FM matrix? Connecting oscilator outputs to inputs? I allready noticed that a feedback loop can not be "drawn" in the matrix. Not from/to the same oscillator (obviously conflicting with the current GUI usage/design) and not between oscillators. You'd probably need at least a sample delay on "re-entry" to determine the start and direction of the signal flow.

Another complication: In a full (or half) FM matrix, each operator could be both a carrier (output to amp) and a modulator (just modulating one or more other operators and/or itself). This could be solved by sending the oscilator "volume" to the amp (or next stage) and the modulation amount to the (next) connected FM modulation input(s).

There are probably some holes and/or inconsistencies in my FM proze..but FM has so many unique characteristics and possibilities. I'd love to have some (more) in Fathomsynth.

FM (actually most often Phase Modulation) is currently dominated by Yamaha in hardware and Native Instrument FM8 in software.

Some practical implementations may he derived from the free and open source FM (PM) synthesizer plugins:
- Dexed FM (a Yamah DX7 emulation)
- Oxe FM (a free FM matrix implementation)

It is is my understanding that FM/PM is computationally quite cheap...so some extra FM features are not necessarely bound to bog down the whole synth.

The bandlimiting already implemented in Fathomsynth (currently except for the FM part of the oscillators?) will probably also help restrain the digital artefacts and aliasing often associated with FM due to the complex sidebands and frequency mirroring (I hope)

Post

BTW...

@Fathomsynth

I'm not expecting any detailed reactions on any of my recent posts/suggestions. Instead, just keep up the good work and make all of us happy with future development/evolution/versions/functions of Fathomsynth

Post Reply

Return to “Instruments”