Tone2 Rayblaster: OUT NOW! (demo version available)

[Ingonator]

I see Markus finally implemented a good decay shape! Any chance to make the same thing in ElectraX? I really liked ElectraX, but Decay was gated to audibly. Looking forward for this improvements in other Tone2 products.

Markus mentioned that he'll maybe add the load and save feature for arp patterns to Saurus, same for the fixed Sine LFOs from the mod matrix.


Ingo

Well done spokesman for ignoring the post and replying with irrelevant info

How does this my post mean ignoring him i just answered to the second part of his post:

Looking forward for this improvements in other Tone2 products.

Ingo

maybe you need to learn to read, he said:

Looking forward for THIS improvements in other Tone2 products.

Your posts in this thread are a riddle for me. I don't get if you are really interested in informations or just want to upset me.
Could be also that you just wait for me posting something wrong or not really precise so you could start your attacks like this again...

Igro could also respond for himself. I don't think he needs your help for that.


Ingo

[Ingonator]

deleted

[Igro]

[

Hi Ingo. I'm talking about the Decay gating, which is obviously improved in Rayblaster. In other words, the Decay tail is now muted at higher resolution, so I cannot hear any audible "Clicks" at the end of the tails.

Hi,

i agree that the Decay in Rayblaster is nice but as mentioned in my post before i doubt that the envelopes will be changed in other products as this could seriously affect existing presets. Anyway maybe there is a way to implement a different shape but keep compatibility (e.g. old/new shape selection) or to implement changes in a way that those presets do not sound too much different.


Ingo

I think that it will not change the sound of existing presets. Markus gated the decay shape at early position just for the sake of CPU usage, that's all.

[Ingonator]

deleted

[Igro]

[

Hi Ingo. I'm talking about the Decay gating, which is obviously improved in Rayblaster. In other words, the Decay tail is now muted at higher resolution, so I cannot hear any audible "Clicks" at the end of the tails.

Hi,

i agree that the Decay in Rayblaster is nice but as mentioned in my post before i doubt that the envelopes will be changed in other products as this could seriously affect existing presets. Anyway maybe there is a way to implement a different shape but keep compatibility (e.g. old/new shape selection) or to implement changes in a way that those presets do not sound too much different.


Ingo

I think that it will not change the sound of existing presets. Markus gated the decay shape at early position just for the sake of CPU usage, that's all.

TLhis is nothing i could answer easily as i don't really know what the problem with the Decay in other synths is.
Would be nice if you got some example/test presets for Rayblaster and other synths to show what you mean. I got all Tone2 synths except Firebird for testing.


Ingo

I just explained you what the problem is. There was an another thread where I showed you everything about this with pictures and audio examples, but AFAIK you didn't understand the point. Don't think it's worth to do this again.

[Ingonator]

I just explained you what the problem is. There was an another thread where I showed you everything about this with pictures and audio examples, but AFAIK you didn't understand the point. Don't think it's worth to do this again.

I am sorry but don't remember it now. and no also from your explanation here i don't get the problem. Could you post a link for that thread?

UPDATE:
Found it, was back in April, don't think you could call this "just explained..."... I guess you talk about this:
http://www.kvraudio.com/forum/viewtopic ... t=#4887757
Have to take a little time to read trough it again but it looks like i din't really agree to you at that time.


Ingo

[LeVzi]

Kriminal, your pathetic trolling of this thread, and the relentless telling people to stop speaking for Tone2 is really boring. Seriously, wind your neck in, stop being such a pathetic, jealous little boy.

As for the synths sounding alike, if you program a synth to make a certain sound, it sounds similar no matter what synth you use. A moog emu sounds like a moog, regardless of who developed it. A juno emu sounds like a juno, but also can be made to do stuff the moog can, so is it fair to say they are the same synth reskinned ? Hardly.

Try the demo for RB, and see how you get on with it, if you like IMS then you'll like it. If you want something more standard, then try something more traditional like Saurus, or Gladiator or ElectraX.

[Kriminal]

Seriously....

....its way past your bedtime.

[Kriminal]

trolling

stating facts

telling people

one person

jealous

of?

[MaxSynths]

Ignore him. (And the other one... )

[pdxindy]

Playing around with the demo again... I notice it is easy to crush the cpu... Haven't figured out which parameters hit the cpu most, but one preset I am editing, on high quality mode, with unison set to 2 takes 40% of a core for 1 note...

I read the Rayblaster manual, but it does not really explain the visual references of the osc window and how the different settings are affecting the sound... It is a complicated trial and error to figure things out... I think some more explanation would be useful for users...

[MaxSynths]

Concerning the CPU use, these are some tips from Markus:

If you notice a high CPU load you should tweak this:

1) Use the Ensemble or Chorus effect instead of 4x unison
2) Go to setup and reduce the number of voices. For most sounds 8 is enough.
3) Reduce release time in the volume envelope. Use the Delay effect for Pads instead.
4) Avoid very low formant frequencies. This has heavy impact on the CPU!
5) Go to setup and select a lower quality mode
6) Set Wave 1/2 mix to 0. Otherwise both waveforms have to be calculated.
7) A very high pitch can also increase the CPU

[Sendy]

Playing around with the demo again... I notice it is easy to crush the cpu... Haven't figured out which parameters hit the cpu most, but one preset I am editing, on high quality mode, with unison set to 2 takes 40% of a core for 1 note...

I read the Rayblaster manual, but it does not really explain the visual references of the osc window and how the different settings are affecting the sound... It is a complicated trial and error to figure things out... I think some more explanation would be useful for users...

As far as I can tell, there's a sample which instead of being looped is triggered at "pitch" frequency (which will be audio rate as that's our note) and each grain is pitched according to "formant" frequency, which has the effect of moving the spectral content of the wave up and down without changing pitch. This is what gives you moving resonances and stopbands, etc.

To cut out some noise and shape the sound further, a window function can be used on each grain, which can i.e. fade in at the beginning and out at the end, to silence and discontinuities that may be coming and going as the waveform "animates". Some of these windows feature kinks that'll distort the wave further, such as the comb types.

PW Sequence and Pulsewidth control controls the microtiming and polarity of the grains as they happen. For example the Square/Pulse PW sequence plays a positive impulse and then a negative impulse shortly after. The phase difference creates PWM effects. The other PW sequences are more complex and let you add subosc effects and general phasey goodness.

That's pretty much my observations of the synth's output from an enduser perspective. Maybe inside it's done differently, but this is my mental model of what happens based on analysing the output.

What happens when you load a loop or do resynthesis I'm not so sure, but it involves some kind of tweak to the setup above. You can also put an osc into sync mode, which fills in the gaps between the impulses with repeats of the impulse; in this case sweeping "formant" gives you sync sweeps which you can further mangle with the PW sequences and windows.

Hope that helps and if I'm wrong remember it's just my guesses from watching it do it's thing so don't take my description as gospel.

[contrary]

Playing around with the demo again... I notice it is easy to crush the cpu... Haven't figured out which parameters hit the cpu most, but one preset I am editing, on high quality mode, with unison set to 2 takes 40% of a core for 1 note...

I read the Rayblaster manual, but it does not really explain the visual references of the osc window and how the different settings are affecting the sound... It is a complicated trial and error to figure things out... I think some more explanation would be useful for users...

As far as I can tell, there's a sample which instead of being looped is triggered at "pitch" frequency (which will be audio rate as that's our note) and each grain is pitched according to "formant" frequency, which has the effect of moving the spectral content of the wave up and down without changing pitch. This is what gives you moving resonances and stopbands, etc.

To cut out some noise and shape the sound further, a window function can be used on each grain, which can i.e. fade in at the beginning and out at the end, to silence and discontinuities that may be coming and going as the waveform "animates". Some of these windows feature kinks that'll distort the wave further, such as the comb types.

PW Sequence and Pulsewidth control controls the microtiming and polarity of the grains as they happen. For example the Square/Pulse PW sequence plays a positive impulse and then a negative impulse shortly after. The phase difference creates PWM effects. The other PW sequences are more complex and let you add subosc effects and general phasey goodness.

That's pretty much my observations of the synth's output from an enduser perspective. Maybe inside it's done differently, but this is my mental model of what happens based on analysing the output.

What happens when you load a loop or do resynthesis I'm not so sure, but it involves some kind of tweak to the setup above. You can also put an osc into sync mode, which fills in the gaps between the impulses with repeats of the impulse; in this case sweeping "formant" gives you sync sweeps which you can further mangle with the PW sequences and windows.

Hope that helps and if I'm wrong remember it's just my guesses from watching it do it's thing so don't take my description as gospel.

That's an awesome explaination; But somehow I can't shake the feeling that somewhere in there , the words " and then a miracle happens" should be included!!!

Just my take ,YMMV

[Ingonator]

Concerning the CPU use, these are some tips from Markus:

If you notice a high CPU load you should tweak this:

1) Use the Ensemble or Chorus effect instead of 4x unison
2) Go to setup and reduce the number of voices. For most sounds 8 is enough.
3) Reduce release time in the volume envelope. Use the Delay effect for Pads instead.
4) Avoid very low formant frequencies. This has heavy impact on the CPU!
5) Go to setup and select a lower quality mode
6) Set Wave 1/2 mix to 0. Otherwise both waveforms have to be calculated.
7) A very high pitch can also increase the CPU

FWIW about CPU and RAM use i had posted this at page 20 of this thread:

Now the release is imminent, I'll speak. I was a demo tester for Rayblaster. It can do some amazing sounds. A lot of times it reminded me of an Andromeda with those gritty harmonics, other times a Juno with a creamy pad. I can tell you the 32 bit at least is as rock solid as anything. It took everything I threw at it and never winced. I also had @ 5 or 6 instances going at once and the computer barely registered 50% CPU (and this is with other fx and things in the mix) There were two presets I had issues with that really rocketed CPU and, as always, I'm sure it was related to release on the envelope. TBH - it was like ElectraX and Saurus had a mutant baby what with the being able to load waves and resynthesis.
I highly suggest you at least have a go with the demo. I, for one, can't wait for some of these talented sound designers to have a go at this.
Peace.

Rayblaster CPU use:

I used those two audio demos from me for testing under "real" conditions:
Ingo Weidner - Empire Of Rays - Rayblaster Demo 1
Ingo Weidner - Empire Of Rays - Remix 1

CPU use for the first audio demo with 6 instances of Rayblaster and one instance of Drumazon was 10-24% on a Core 2 Quad Q6600 (4 x 2.4 GHz) CPU.
The maximum on the "Remix" that uses FXpansion Geist for drums was 29%.

The maximum number of used voices for those demos including Unison voices should be around 19 voices (+ a few voices for the drum in Drumazon).

The CPU use of the presets, even polyphonic ones could be either quite low (6-10% on my Core 2 Quad Q6600 CPU) or a higher depending on features, parameter settings and Unison etc.
Anyway all presets sounds should be usable also with 6 notes or more and "high quality" setting.
Lead sounds (monophonic) in average are around 2-5% on my CPU with "high quality" setting and 1-2% less with "Low CPU" setting (from "setup" menu.
While there are a few monophonic presets that only need 1% there are some extreme lead presets with 4x unison and delay/Echo FX that need up to 10% in high quality mode (again on a 5 years old Core 2 Quad Q6600 CPU).
This also seem to depend on how complex the waveform and the resulting filter response is. A simple Sawtooth waveform usually needs only very few CPU.

Like with most other synths of curse also the Amp envelope release time plays an important rule for the CPU use.


About RAM use:

Rayblaster needs quite a bit of RAM for each instance which is not a big problem with the 64-bit version and a lot of installed RAM (8 GB+) but at 32-bit there is a limit of 3,3 GB usable RAM and a 2 GB limit for a single program.

I am using Windows 7 32-bit and here are the maximum of instances i could currently load without problems:

in Ableton Live 8.3.4: 7-8 instances (sometimes also 9 instances)

in Reaper 4.22 32-bit: 8-9 instances

The first instance of Rayblaster needs a bit over 200 MB of RAM and each additional one around 150 MB.

As alraedy mentioned with 64-bit and enough RAM you could do a lot (30 or more) instances.


UPDATE:

I have checked on my Sony VAIO notebook now which uses a Core 2 Duo CPU (2 x 2.53 GHz), 4 GB installed RAM and Windows 7 32-bit.

In Ableton Live 8 (32-bit) the CPU use for my first audio demo (6 instances of Rayblaster, 1 instance of Drumazon) is around 13-32%.

Maximum used voices including Unison voices should have been 19 voices (+ a few voices for the drum in Drumazon).

Maximum instances in Ableton Live 8 are 7-8 which currently seems to be the maximum for a 32-bit OS (maybe one instance more with Reaper that needs less RAM itself).


Ingo