GRANULIZER 2 Released | Inertia Sound Systems

VST, AU, AAX, etc. plug-in Virtual Instruments discussion
User avatar
InertiaSoundSystems
KVRist
216 posts since 9 Jan, 2017 from Athens, Greece

Post Thu Nov 26, 2020 11:07 am

ThomasHelzle wrote:
Thu Nov 26, 2020 3:32 am
Tempted, but get massive regular CPU spikes here that lead to crackles. Haven't seen something like this in any other plugin for years:
Granulizer2_DSP_Spikes.jpg
There seems to be one "stream" of these spikes per note played.
Not usable on my system as it is. :-(

Bitwig Studio (but it's the same in other DAWs), 64 GB RAM, GeForce RTX 2080 TI, i7 6 core @ 4.1 Ghz, Focusrite Scarlett 18i20, Windows 8.1 x64, all with latest drivers and updates.

I sent this to support some days ago but didn't get an answer yet.

And I miss a scalable interface...

Cheers,

Tom
Hello ThomasHelzle,
It seems that your buffer size is set at 128 samples. It is normal that for high polyphony and demanding settings the CPU will be stressed. We have a dedicated FAQ section for CPU and RAM consumption. There is also a detailed section in the manual that explains this case as well. Please feel free to contact our support for any other issues you may have, or reply here!
Thank you for your feedback!

User avatar
ThomasHelzle
KVRAF
5615 posts since 9 Dec, 2008 from Berlin

Post Thu Nov 26, 2020 11:30 am

I contacted your support several days ago and didn't get an answer ;-)
I would be okay with "normal" DSP load, but your plugin does something weird that leads to those regular spikes and that prevents me from even using it with something like two voices in a normal project context.
This shouts bad programming to me and I have other granular tools that don't have any such problems, so thank you, but I'm not interested anymore.
512 samples latency? No, thank you very much.

Cheers, over and out,

Tom
"Out beyond the ideas of wrongdoing and rightdoing, there is a field. I’ll meet you there." - Rumi
ScreenDream|Thomas Helzle 8) Twitter

User avatar
InertiaSoundSystems
KVRist
216 posts since 9 Jan, 2017 from Athens, Greece

Post Thu Nov 26, 2020 11:36 am

ThomasHelzle wrote:
Thu Nov 26, 2020 11:30 am
I contacted your support several days ago and didn't get an answer ;-)
I would be okay with "normal" DSP load, but your plugin does something weird that leads to those regular spikes and that prevents me from even using it with something like two voices in a normal project context.
This shouts bad programming to me and I have other granular tools that don't have any such problems, so thank you, but I'm not interested anymore.
512 samples latency? No, thank you very much.

Cheers, over and out,

Tom
We are sorry and we really apologize to have missed that! We will update our FAQ with the further details included in the manual so that users with similar demands from the plugin can figure out things more intuitively.
Cheers and thank you for the feedback.

User avatar
InertiaSoundSystems
KVRist
216 posts since 9 Jan, 2017 from Athens, Greece

Post Thu Nov 26, 2020 11:56 am

chrisstiles wrote:
Thu Nov 26, 2020 3:48 am
Yeah same. After demoing it I saw similar - and what seems to be a single play head is somewhat of a limitation.
Hello chrisstiles,
Each voice has 2 playheads one for each channel. The amount of grains per second produced cannot be visualized through the playhead here. You can try and tweak the magic button 1st to add some randomness in the read process. This will instantly reflect to the visual feedback.You can also tweak density and speed to see how this affects is, and also map that to the sound that you hear.
Hope i got this correctly :)
Cheers!

WasteLand
KVRian
1383 posts since 8 Jun, 2018

Post Thu Nov 26, 2020 12:38 pm

it is not my intention to go in discussion, but i also bought spacecraft. and have other granular synths, or with granular capabilities. and demoed quanta. i am not an expert on granular synthesis (composition...), i am reading a book about it, that goes really in depth. in way grains can lead to compositions, or it is an archaic term for a new type of synthesis, or better an approach to music.

so that aroused my interest. spacecraft, the names says it all, 128 grains playing at the same time, 60% cpu usage (one core by the way)... it works with MPE, at 512 samples. (with unarmed tracks, because it can play itself, of course in reaper and cubase, they preprocess, and buffer load drops a lot.)

but it is indeed strange, or not strange, granulizer 2, only uses 10-15%, i think. it is not always the cpu usage that determines how fast a buffer can be filled.

i see it with other plugins that put a lot strain on the buffer, and in cpu percentage it isn't very high (and i think, i can be wrong, a not so high cpu usage, does not mean, a slower cpu can do it, with a higher cpu usage, the ISP is still important even on a load of 10 or 20%. (at 100% core frequency of course..)).

i can't determine of granulizer 2 can be more optimized. and as i already said, i am no expert, don't have all granular synths..
but i before i read the FAQ, and i saw what granulizer 2 is capable of, i thought, that's gonna strain my DAW..

each have their own priorities, i can get it, that you want to stay at 128. (i can't 256, is the minimum, my soft modulars would break.. and i wanted to stay at 256... but i must let go.. personal.)

so this post is based about the performance, and the decision not to buy it when it needs 512 samples, but not for a discussion, more insight, why it puts so much strain on the buffer. quanta 100 grains per voice, 1000ms.
spacecraft 128 grains... 60% cpu usage!

it is a good idea, inertia!, to put that in the FAQ and/or manual. it is not uncommon..

quanta for instance isn't that cpu intensive, but it works quite differently, almost identical to the granular oscillator in biotek2, also capable of MPE. quanta, i repeat, it can make 100 grains of 1000ms per voice, but it works differently than for example biotek2 granular osc, that can also do per voice, but how many grains? so the differences? i am not new to granular synthesis, but new to the in depth workings of granular. the approach of quanta is quite different.

i can only add this: over a month, when working with granular synthesis, more in depth and reading the book, rereading it, and read part 2... i can say more about it. so why say things now. it are questions.

it seems that unlike wavetable synths, that struggle with the same limits, granular synths are quite different in their approach, or granular effects (like melda's, glitchmachines (also instruments), unfiltered audio).

so i don't really understand it, but i already can judge a little by what a granular synth can do, what it costs, for your buffer..

(i am curious though which other granular tools Tom from Berlin means. and i repeat, don't get me wrong, understanding this particular synth better, also how it works under the hood, gives me more possibilities to work with it. so Tom from Berlin, i use your post as a trampoline, as you see in how i write... and i must jump of..)
take me as apple as reed-son as flesh-icon, immediately
shits the wise man. hero and blind. take me as
stick as tongue as dog... https://soundcloud.com/sada-exposada

User avatar
ThomasHelzle
KVRAF
5615 posts since 9 Dec, 2008 from Berlin

Post Thu Nov 26, 2020 12:58 pm

The ones I like most are Granite and all the grain modules in Falcon.
I also enjoy The Mangle (up to 32 voices and 1000 grains per voice) and several Reaktor Ensembles.

But I don't want to clutter this thread with other software so will stay out of here from now on.

Cheers,

Tom
"Out beyond the ideas of wrongdoing and rightdoing, there is a field. I’ll meet you there." - Rumi
ScreenDream|Thomas Helzle 8) Twitter

User avatar
InertiaSoundSystems
KVRist
216 posts since 9 Jan, 2017 from Athens, Greece

Post Thu Nov 26, 2020 1:10 pm

WasteLand wrote:
Thu Nov 26, 2020 12:38 pm
it is not my intention to go in discussion, but i also bought spacecraft. and have other granular synths, or with granular capabilities. and demoed quanta. i am not an expert on granular synthesis (composition...), i am reading a book about it, that goes really in depth. in way grains can lead to compositions, or it is an archaic term for a new type of synthesis, or better an approach to music.

so that aroused my interest. spacecraft, the names says it all, 128 grains playing at the same time, 60% cpu usage (one core by the way)... it works with MPE, at 512 samples. (with unarmed tracks, because it can play itself, of course in reaper and cubase, they preprocess, and buffer load drops a lot.)

but it is indeed strange, or not strange, granulizer 2, only uses 10-15%, i think. it is not always the cpu usage that determines how fast a buffer can be filled.

i see it with other plugins that put a lot strain on the buffer, and in cpu percentage it isn't very high (and i think, i can be wrong, a not so high cpu usage, does not mean, a slower cpu can do it, with a higher cpu usage, the ISP is still important even on a load of 10 or 20%. (at 100% core frequency of course..)).

i can't determine of granulizer 2 can be more optimized. and as i already said, i am no expert, don't have all granular synths..
but i before i read the FAQ, and i saw what granulizer 2 is capable of, i thought, that's gonna strain my DAW..

each have their own priorities, i can get it, that you want to stay at 128. (i can't 256, is the minimum, my soft modulars would break.. and i wanted to stay at 256... but i must let go.. personal.)

so this post is based about the performance, and the decision not to buy it when it needs 512 samples, but not for a discussion, more insight, why it puts so much strain on the buffer. quanta 100 grains per voice, 1000ms.
spacecraft 128 grains... 60% cpu usage!

it is a good idea, inertia!, to put that in the FAQ and/or manual. it is not uncommon..

quanta for instance isn't that cpu intensive, but it works quite differently, almost identical to the granular oscillator in biotek2, also capable of MPE. quanta, i repeat, it can make 100 grains of 1000ms per voice, but it works differently than for example biotek2 granular osc, that can also do per voice, but how many grains? so the differences? i am not new to granular synthesis, but new to the in depth workings of granular. the approach of quanta is quite different.

i can only add this: over a month, when working with granular synthesis, more in depth and reading the book, rereading it, and read part 2... i can say more about it. so why say things now. it are questions.

it seems that unlike wavetable synths, that struggle with the same limits, granular synths are quite different in their approach, or granular effects (like melda's, glitchmachines (also instruments), unfiltered audio).

so i don't really understand it, but i already can judge a little by what a granular synth can do, what it costs, for your buffer..

(i am curious though which other granular tools Tom from Berlin means. and i repeat, don't get me wrong, understanding this particular synth better, also how it works under the hood, gives me more possibilities to work with it. so Tom from Berlin, i use your post as a trampoline, as you see in how i write... and i must jump of..)
Hello,
One key difference is if you jump in the frequency domain or not. Than means that for every grain when the FX section is active, there is and FFT taking place, then several manipulations are applied to the spectrum of each grain (both magnitude and phase) and then the grains are brought back to the time domain before they overlap. And this happens for more than 100 / 300ms lasting grains / sec / voice depending on the granular mode (free or fixed).

WARP also extends the standard, using a grain based characteristics - feedback loop modulation.
If you like to see what the consumption is in the time domain strictly, you can turn down to 0 the WARP depth and the FX DRY/WET and see what you get. For medium range settings the CPU there is quite low. Even for highest time domain settings the CPU remains low when operating on the time domain (roughly 5-7% per voice with 128 samples buffer size at 44.1 kHz on a MacBook Pro (Retina, 15-inch, Mid 2014) in Ableton Live).

In contrast to the general belief the funny thing is that granular synthesis is not necessarily cpu - cheap. It can be very light at times, but it can also get extremely heavy depending on multiple engine parameters and on what core features - processing each manufacturer adds to the standard concept.

In terms of optimization, there has been much time spent on that and the CPU consumption has been reduced significantly since v.2.0.0, allowing the plugin to be more responsive and for more features to be added. Of course it goes without saying that there is always room for improvement! (i.e openCL has always been under discussion)

Here is the Halcyon preset with spectral processing and warping bypassed for 32 and 128 samples of buffer size with a single voice performing at maximum time domain settings
Screen Shot 2020-11-26 at 11.16.20 PM.png
Screen Shot 2020-11-26 at 11.15.51 PM.png
Cheers!
You do not have the required permissions to view the files attached to this post.

WasteLand
KVRian
1383 posts since 8 Jun, 2018

Post Thu Nov 26, 2020 1:28 pm

ThomasHelzle wrote:
Thu Nov 26, 2020 12:58 pm
The ones I like most are Granite and all the grain modules in Falcon.
I also enjoy The Mangle (up to 32 voices and 1000 grains per voice) and several Reaktor Ensembles.

But I don't want to clutter this thread with other software so will stay out of here from now on.

Cheers,

Tom
thanks for your answer. granite i know. there is another great granular synth i discovered, i must take notes, forgot the name. don't own falcon, and problably never will (personal opinion, great synth, no doubt, but not for me). the mangle i have read about, also on my list. but eventually i have chosen; spacecraft and granulizer2..

all are interesting, and have their merits. or are great. personal choice.

and yes of course reaktor, almost made for granular.. (and i have also granular stuff for M4L...)
in that department reaktor/M4L, i am good..
InertiaSoundSystems wrote:
Thu Nov 26, 2020 1:10 pm

Hello,
One key difference is if you jump in the frequency domain or not. Than means that for every grain when the FX section is active, there is and FFT taking place, then several manipulations are applied to the spectrum of each grain (both magnitude and phase) and then the grains are brought back to the time domain before they overlap. And this happens for more than 100 / 300ms lasting grains / sec / voice depending on the granular mode (free or fixed).

WARP also extends the standard, using a grain based characteristics - feedback loop modulation.
If you like to see what the consumption is in the time domain strictly, you can turn down to 0 the WARP depth and the FX DRY/WET and see what you get. For medium range settings the CPU there is quite low. Even for highest time domain settings the CPU remains low when operating on the time domain (roughly 5-7% per voice with 128 samples buffer size at 44.1 kHz on a MacBook Pro (Retina, 15-inch, Mid 2014) in Ableton Live).

In contrast to the general belief the funny thing is that granular synthesis is not necessarily cpu - cheap. It can be very light at times, but it can also get extremely heavy depending on multiple engine parameters and on what core features - processing each manufacturer adds to the standard concept.

In terms of optimization, there has been much time spent on that and the CPU consumption has been reduced significantly since v.2.0.0, allowing the plugin to be more responsive and for more features to be added. Of course it goes without saying that there is always room for improvement! (i.e openCL has always been under discussion)

Cheers!
thanks for your answer! this gives me more insight how it works.

the granular world is rich. and of course, i could quote william blake (and i actually read the whole poem, and actually i own all of his work, in facsimiles and text, several books, and a companion..).

o well why not:

To see a World in a Grain of Sand
And a Heaven in a Wild Flower
Hold Infinity in the palm of your hand
And Eternity in an hour

he foresaw the development of grains (irony mark..), it has quite a history, also with composers, it steadily evolved.. grains, granular. it is combination of technology and experimental composition, the development.

why mention poetry? and grains.. i have reason for it. but that is for another time, another place, and perhaps it will become available in the english language, someday..
that is the way i work, i try to understand, and then to feel, to make something.
take me as apple as reed-son as flesh-icon, immediately
shits the wise man. hero and blind. take me as
stick as tongue as dog... https://soundcloud.com/sada-exposada

User avatar
InertiaSoundSystems
KVRist
216 posts since 9 Jan, 2017 from Athens, Greece

Post Thu Nov 26, 2020 1:34 pm

WasteLand wrote:
Thu Nov 26, 2020 1:28 pm
ThomasHelzle wrote:
Thu Nov 26, 2020 12:58 pm
The ones I like most are Granite and all the grain modules in Falcon.
I also enjoy The Mangle (up to 32 voices and 1000 grains per voice) and several Reaktor Ensembles.

But I don't want to clutter this thread with other software so will stay out of here from now on.

Cheers,

Tom
thanks for your answer. granite i know. there is another great granular synth i discovered, i must take notes, forgot the name. don't own falcon, and problably never will (personal opinion, great synth, no doubt, but not for me). the mangle i have read about, also on my list. but eventually i have chosen; spacecraft and granulizer2..

all are interesting, and have their merits. or are great. personal choice.

and yes of course reaktor, almost made for granular.. (and i have also granular stuff for M4L...)
in that department reaktor/M4L, i am good..
InertiaSoundSystems wrote:
Thu Nov 26, 2020 1:10 pm

Hello,
One key difference is if you jump in the frequency domain or not. Than means that for every grain when the FX section is active, there is and FFT taking place, then several manipulations are applied to the spectrum of each grain (both magnitude and phase) and then the grains are brought back to the time domain before they overlap. And this happens for more than 100 / 300ms lasting grains / sec / voice depending on the granular mode (free or fixed).

WARP also extends the standard, using a grain based characteristics - feedback loop modulation.
If you like to see what the consumption is in the time domain strictly, you can turn down to 0 the WARP depth and the FX DRY/WET and see what you get. For medium range settings the CPU there is quite low. Even for highest time domain settings the CPU remains low when operating on the time domain (roughly 5-7% per voice with 128 samples buffer size at 44.1 kHz on a MacBook Pro (Retina, 15-inch, Mid 2014) in Ableton Live).

In contrast to the general belief the funny thing is that granular synthesis is not necessarily cpu - cheap. It can be very light at times, but it can also get extremely heavy depending on multiple engine parameters and on what core features - processing each manufacturer adds to the standard concept.

In terms of optimization, there has been much time spent on that and the CPU consumption has been reduced significantly since v.2.0.0, allowing the plugin to be more responsive and for more features to be added. Of course it goes without saying that there is always room for improvement! (i.e openCL has always been under discussion)

Cheers!
thanks for your answer! this gives me more insight how it works.

the granular world is rich. and of course, i could quote william blake (and i actually read the whole poem, and actually i own all of his work, in facsimiles and text, several books, and a companion..).

o well why not:

To see a World in a Grain of Sand
And a Heaven in a Wild Flower
Hold Infinity in the palm of your hand
And Eternity in an hour

he foresaw the development of grains (irony mark..), it has quite a history, also with composers, it steadily evolved.. grains, granular. it is combination of technology and experimental composition, the development.

why mention poetry? and grains.. i have reason for it. but that is for another time, another place, and perhaps it will become available in the english language, someday..
that is the way i work, i try to understand, and then to feel, to make something.
We are really glad we are among your choices. As music producers ourselves we really like that there is so many great granular tools out there along with ours, each one of them with its own character, potential and flavor!

And of course poetry and grains go together!

This great artist and innovator has been our inspiration many years now, most probably you have heard of him!
https://www.granularsynthesis.com/hthesis/xenakis.html

Best!

cb8rwh
KVRist
447 posts since 25 Apr, 2005

Post Thu Nov 26, 2020 2:17 pm

I am trying to register so I can buy Granulaizer, but everytime I fill in the form and hit Register I get taken to a Forbidden page. I don't think I am doing anything wrong. Tried in Chrome and Firefox.

Any ideas?

User avatar
InertiaSoundSystems
KVRist
216 posts since 9 Jan, 2017 from Athens, Greece

Post Thu Nov 26, 2020 2:32 pm

cb8rwh wrote:
Thu Nov 26, 2020 2:17 pm
I am trying to register so I can buy Granulaizer, but everytime I fill in the form and hit Register I get taken to a Forbidden page. I don't think I am doing anything wrong. Tried in Chrome and Firefox.

Any ideas?
You got caught in the middle of trying to resolve the issue with our mail sender!

You can go ahead and register now. If you are using hotmail, we'll have to verify you manually though. Let us know if it works out, either here or directly through our email.

Sorry for the inconvenience this might have caused you.

cb8rwh
KVRist
447 posts since 25 Apr, 2005

Post Thu Nov 26, 2020 2:49 pm

Yeah thanks it is working now. Bought :tu:

Dunc
KVRist
443 posts since 17 Jun, 2003 from UK 'ish

Post Fri Nov 27, 2020 3:53 am

I registered on the site a few hours ago - still waiting on my email to confirm registering. Is there a delay I need to know about (didn't use a hotmail account)?

WasteLand
KVRian
1383 posts since 8 Jun, 2018

Post Fri Nov 27, 2020 4:04 am

Dunc wrote:
Fri Nov 27, 2020 3:53 am
I registered on the site a few hours ago - still waiting on my email to confirm registering. Is there a delay I need to know about (didn't use a hotmail account)?
i can't really answer this, but in my case, i contacted support, and indeed problems with hotmail and someother.

contact support, they will do it manually for you, of course, with same e-mail address..
in my case, within 15 minutes or so..

EDIT:

@inertia, yes xenakis! must read more and listen more. it will give me more inspirition. and learning now how to use granulizer 2...
take me as apple as reed-son as flesh-icon, immediately
shits the wise man. hero and blind. take me as
stick as tongue as dog... https://soundcloud.com/sada-exposada

User avatar
InertiaSoundSystems
KVRist
216 posts since 9 Jan, 2017 from Athens, Greece

Post Fri Nov 27, 2020 4:17 am

Dunc wrote:
Fri Nov 27, 2020 3:53 am
I registered on the site a few hours ago - still waiting on my email to confirm registering. Is there a delay I need to know about (didn't use a hotmail account)?
We keep poking our Mail Sender, but they haven't addressed it yet unfortunately. Feel free to reach to us directly at admin@inertiasoundsystems.com and we'll enable your account for you!

Sorry again for the inconvenience.

Return to “Instruments”