glokraw wrote: Fri Aug 07, 2020 6:48 pm
But guys like Fathom, Urs, and Dimitri shape the future of
wallets
yeah, including their own, its called business
Fathom Synth Development Thread
-
- KVRAF
- Topic Starter
- 1584 posts since 25 Mar, 2017
StardragonEX, In regard to your question upgrading Fathom's sound.
That is the critical question and it has been on my mind now for about a year.
This is complicated.
At the risk of sounding arrogant, technically speaking there is no other synth that has higher "quality" sound than Fathom, in the mathematical sense. Fathom uses wave buffers of 16384 samples which is insanely large. Then Fathom uses fourth order double polynomial interpolation between sample points and to my knowledge it is the only synth in existence which has this. So the high frequency noise floor in Fathom is lower than any other synth in existence.
This shown here:
https://www.fathomsynth.com/home#/audio/
The image is actually for Fathom's third order polynomial interpolation and the recent addition of fourth order double polynomial interpolation means the noise floor is even lower than in that older diagram.
The comparison is difficult because most synths do not allow you control over the max partials before filtering so there is no way to turn down the partials to see the high frequency noise as there is in Fathom. If you could, however, you would see that Fathom's high frequency aliasing is lower, in most cases by a lot.
However!
"Clean" sound does not mean "better" sound.
If you ask 100 people which synth actually has the "best" sound, 90 of them will say Dune 3, five of them will say uHe Diva and another five will say Omnisphere, with the more advanced sound designers probably selecting Omnisphere.
If you turn off everything in both Dune 3 and Fathom and just play a mono oscillator and nothing else, no detune, no filters, no effects, then Dune and Fathom will be indistinguishable, assuming you were to draw the Dune 3 waveform into Fathom's wavedraw oscillator.
But this is not the right comparison.
The reason Dune sounds so incredible, in my opinion, is their detune algorithm.
There is some serious DSP magic going on under the hood in Dune 3 as soon as you detune the oscillator.
Unfortunately I'm not exactly sure how they are doing it.
To my ears it sounds like they are using a Gaussian distribution for both the pitch separation and the width of frequency oscillation in each virtual voice. In addition they are probably using nonlinear oscillation for each virtual voice like Fathom does. You can program the same thing into Fathom, but Fathom only offers 8 detune voices. I think the secret in Dune is the efficiency since they are automatically generating a similar Gaussian / Non Linear algorithm automatically but for a significantly higher number of voices.
I plan on experimenting with this in Fathom in the near future but have not yet had time due to the priority of the Intel AVX parallel processing.
The very next thing I plan on doing in Fathom after the 3.0 release is Video card GPU processing which will give Fathom the ability to modulate individual partials in real time with zero aliasing. This is the holy grail of synth audio and to my knowledge Fathom will be the only synth to do it. Until then Fathom's perceived sound quality will not be at the absolute top in some people's opinion despite the fact that its high frequency aliasing is actually cleaner.
That is the critical question and it has been on my mind now for about a year.
This is complicated.
At the risk of sounding arrogant, technically speaking there is no other synth that has higher "quality" sound than Fathom, in the mathematical sense. Fathom uses wave buffers of 16384 samples which is insanely large. Then Fathom uses fourth order double polynomial interpolation between sample points and to my knowledge it is the only synth in existence which has this. So the high frequency noise floor in Fathom is lower than any other synth in existence.
This shown here:
https://www.fathomsynth.com/home#/audio/
The image is actually for Fathom's third order polynomial interpolation and the recent addition of fourth order double polynomial interpolation means the noise floor is even lower than in that older diagram.
The comparison is difficult because most synths do not allow you control over the max partials before filtering so there is no way to turn down the partials to see the high frequency noise as there is in Fathom. If you could, however, you would see that Fathom's high frequency aliasing is lower, in most cases by a lot.
However!
"Clean" sound does not mean "better" sound.
If you ask 100 people which synth actually has the "best" sound, 90 of them will say Dune 3, five of them will say uHe Diva and another five will say Omnisphere, with the more advanced sound designers probably selecting Omnisphere.
If you turn off everything in both Dune 3 and Fathom and just play a mono oscillator and nothing else, no detune, no filters, no effects, then Dune and Fathom will be indistinguishable, assuming you were to draw the Dune 3 waveform into Fathom's wavedraw oscillator.
But this is not the right comparison.
The reason Dune sounds so incredible, in my opinion, is their detune algorithm.
There is some serious DSP magic going on under the hood in Dune 3 as soon as you detune the oscillator.
Unfortunately I'm not exactly sure how they are doing it.
To my ears it sounds like they are using a Gaussian distribution for both the pitch separation and the width of frequency oscillation in each virtual voice. In addition they are probably using nonlinear oscillation for each virtual voice like Fathom does. You can program the same thing into Fathom, but Fathom only offers 8 detune voices. I think the secret in Dune is the efficiency since they are automatically generating a similar Gaussian / Non Linear algorithm automatically but for a significantly higher number of voices.
I plan on experimenting with this in Fathom in the near future but have not yet had time due to the priority of the Intel AVX parallel processing.
The very next thing I plan on doing in Fathom after the 3.0 release is Video card GPU processing which will give Fathom the ability to modulate individual partials in real time with zero aliasing. This is the holy grail of synth audio and to my knowledge Fathom will be the only synth to do it. Until then Fathom's perceived sound quality will not be at the absolute top in some people's opinion despite the fact that its high frequency aliasing is actually cleaner.
-
- addled muppet weed
- 111253 posts since 26 Jan, 2003 from through the looking glass
-
- addled muppet weed
- 111253 posts since 26 Jan, 2003 from through the looking glass
theres no actual price when you click on the "buy" page 
-
- KVRist
- 178 posts since 26 Apr, 2018
$45 here....
https://www.fathomsynth.com/products
or $65 here....
https://www.kvraudio.com/marketplace/fa ... weed-audio
Fathom Mono *is* free though here, but with some limitations (still bloody good though)....
https://www.pluginboutique.com/products/4309
https://www.fathomsynth.com/products
or $65 here....
https://www.kvraudio.com/marketplace/fa ... weed-audio
Fathom Mono *is* free though here, but with some limitations (still bloody good though)....
https://www.pluginboutique.com/products/4309
-
- KVRAF
- 9521 posts since 6 Oct, 2004
Indeed, "live long and prosper", all you great synth creators!
Making audio delights never before heard by human ears...so cool!!!
-
- KVRAF
- 19796 posts since 16 Sep, 2001 from Las Vegas,USA
No higher than you are.vurt wrote: Fri Aug 07, 2020 5:23 pm you just plus 1ed someone who was agreeing with you in the first place.
are you high right now?
Dead wrong but hey thanks for playing. I feel sorry for you that you don't understand how wrong you are......
Anyway back on Fathom. I own 2.33 is there an upgraded version from that ? I got in early with Fathom but kind of lost track of it as time went on.
None are so hopelessly enslaved as those who falsely believe they are free. Johann Wolfgang von Goethe
-
- KVRist
- 75 posts since 1 Jun, 2020
@FathomSynth maybe Fathom sound needs character like Sylenth1 or Dune 3, maybe you could add an option to add character (you have normal processing and list of different characters to give to your sound), additionally to the high quality of sound you want to achieve and optimization to use less processing power with the same or nearly the same results.
-
- KVRAF
- Topic Starter
- 1584 posts since 25 Mar, 2017
Companies like Synapse or Lennar who have serious DSP secrets don't publish them for good reason.
Once 3.0 is out I will be devoting more time to this and I have some ideas.
Once 3.0 is out I will be devoting more time to this and I have some ideas.
-
- Banned
- 10196 posts since 12 Mar, 2012 from the Bavarian Alps to my feet and the globe around my head
We need the best sounding synth with the best sounding imperfections, optimally perfected!
I know how difficult programming is, so I can only wish you the best!
I know how difficult programming is, so I can only wish you the best!
-
- KVRAF
- 9521 posts since 6 Oct, 2004
Dude, the whole post is obviously a joke!!! Thicken up that skin, a bit...Teksonik wrote: Fri Aug 07, 2020 11:21 pmNo higher than you are.vurt wrote: Fri Aug 07, 2020 5:23 pm you just plus 1ed someone who was agreeing with you in the first place.
are you high right now?
Dead wrong but hey thanks for playing. I feel sorry for you that you don't understand how wrong you are......
Anyway back on Fathom. I own 2.33 is there an upgraded version from that ? I got in early with Fathom but kind of lost track of it as time went on.
Cheers
-
- KVRAF
- 9521 posts since 6 Oct, 2004
since my sense of humour is known to be lousy, I'll try and make up for it with a noisy Fathom clip attached Cheers
You do not have the required permissions to view the files attached to this post.