AAS Chromaphone 3 : ???

[Zo]

Hey guyz , late night here and don't ask me why lol decided to puagrde from chromaphone 2 to 3 (as well;string studio 3 and that modular thing , all for 34 $ witch was hard to resist) 

Cromaphone always been a superb synth but the V3 is barely usable due to the cpu hit , and don't even think about using pitch wheel ...

 

Am i missing something ? any infos from AAS on that subject ? what's going ? 2 instance of C2 are less cpu intensive than C3 !!

[Starship Krupa]

Sigh.

Yeah, I love me some Chromaphone, my favorite A|A|S Player soundpacks are Chromaphone sounds. I really like what they can do with the arpeggiator(s), especially once they start stacking layers.

Most of the time I can ease the resource hit by knocking the number of simultaneous voices down to 8. Many of the sounds have long decays, and once you add in arpeggiation, they can really stack up.

I want to purchase at least one full A|A|S synth while they're half off, so first I tried the Chromaphone demo. Even with cutting the number of voices and shutting off reverb (I prefer to use my own), I still couldn't jack my buffers up high enough to stop the dropouts, and that was with only one instance. My system is no rocket sled, but I don't feel it's quite ready for the boneyard, and I don't have good feelings about dropping a hundy on software that's going to make me yearn for a more powerful (expensive) system. Maybe after the next computer upgrade.

I know that I can freeze synth tracks and all that, but my composing workflow for my electronic stuff doesn't lend itself to that. I do a lot of slight nudging of note start and stop times and I'd be constantly freezing and thawing.

String Studio and Ultra VA are somewhat better on resources, but they suffer in varying degrees from FrankenUI, where there are new, Chromaphone 3-style elements mixed with the kinda outdated clunky and hard to navigate "Session" UI. As if they were released in mid-update.

If you're disappointed in your upgraded versions, please take a moment and send them a polite email telling them what you've said here. I've known them to be responsive in the past, and they should know that earlier customers are struggling to be able to use the newer versions.

[Promidi]

Yes, Chromophone 3 can be a bit of a  CPU hog

One thing you might like to do with Chromophone 3 is to to reduce the Mode Density on the resonators.  The caveat there, however, is you are going to lose some of the high end crispness on some of the sounds....

[Zo]

6 hours ago, Starship Krupa said:

Sigh.

Yeah, I love me some Chromaphone, my favorite A|A|S Player soundpacks are Chromaphone sounds. I really like what they can do with the arpeggiator(s), especially once they start stacking layers.

Most of the time I can ease the resource hit by knocking the number of simultaneous voices down to 8. Many of the sounds have long decays, and once you add in arpeggiation, they can really stack up.

I want to purchase at least one full A|A|S synth while they're half off, so first I tried the Chromaphone demo. Even with cutting the number of voices and shutting off reverb (I prefer to use my own), I still couldn't jack my buffers up high enough to stop the dropouts, and that was with only one instance. My system is no rocket sled, but I don't feel it's quite ready for the boneyard, and I don't have good feelings about dropping a hundy on software that's going to make me yearn for a more powerful (expensive) system. Maybe after the next computer upgrade.

I know that I can freeze synth tracks and all that, but my composing workflow for my electronic stuff doesn't lend itself to that. I do a lot of slight nudging of note start and stop times and I'd be constantly freezing and thawing.

String Studio and Ultra VA are somewhat better on resources, but they suffer in varying degrees from FrankenUI, where there are new, Chromaphone 3-style elements mixed with the kinda outdated clunky and hard to navigate "Session" UI. As if they were released in mid-update.

If you're disappointed in your upgraded versions, please take a moment and send them a polite email telling them what you've said here. I've known them to be responsive in the past, and they should know that earlier customers are struggling to be able to use the newer versions.

To be honest , i was so pistoff that i wanted to ask for a refund , yeah for 30 euros , because i just installed stuff that i know i ain t gonna use , simple as that ... sessions during production are loaded with other stuff ...

But i think i ll give up , i googled and people said that the aas response was :this is the way it is ...

even on their faq page about chromaphone , damn i should have tested it before , i m prettybsure i did n t upgraded before because of that. I don t trust myself enougth lol ....

Shame , really , the worst thing for a dev is making plugins people gonna skip more and more and being ok with that is a suicide ...

[Zo]

Oh by thebway fx have little imapct , i even reduced voiçes to 4 ... gonna check resonnator , but if those changes could be locked it would be great but having to tweak each patche loaded is a pain .

I started to load patches and everytime i had 2parts , disabled one or the other that is not usefull ... but to be hinest , it entered straight iinto dustware .... i dont even bother getting my xmas gift

[Starship Krupa]

A|A|S licenses may be transferred, so if you wanted to flip your full licenses, you might be able to find a buyer and recoup your cost. You'd need to surrender the full upgraded license(s), though, as I'm pretty sure the upgrades themselves aren't transferable. So if you still want to be able to use your version 2 stuff, that wouldn't be possible.

I am very glad that I ran the trial before paying for a license. I started out very enthused and ready, and then became less so. I still want it, just not now.

I think there's a tradeoff that media software companies (I used to do software QA) have to keep making: hold back functionality from the latest version in the hope that they'll sell some licenses to people with older hardware, or put in the features and hope to more licenses to people with newer hardware. And the answer is often to do the second thing, because people like me who squeeze every last bit of use out of older hardware tend to spend less money in general, which includes for software (especially upgrade licenses).

[Zo]

https://www.applied-acoustics.com/chromaphone-3/faq/

 

What’s up with Chromaphone’s CPU usage? 

Chromaphone 3 can indeed be quite CPU intensive. It’s important to understand that a Chromaphone 3 instance can only run on one core of a multi-core system, so that particular core can be overloaded.

There are internal and external variables that can affect the performance of Chromaphone 3. The internal variables are: unison, number of voices, arpeggiator, resonator quality, decay, and pitch modulation. These vary per preset/program. The external variables include: sample rate, buffer size, CPU model. The external variables can be tweaked but don’t vary per preset/program.

Strategies for better performance

Unison—It literary doubles or quadruples the CPU usage per note, so it’s a good idea to toggle it to see if its contribution is worth the load for your particular situation. For some presets, another possible workaround would be to use a chorus effect instead of the unison.

Number of voices, decay—With the number of voices set to 16, you can trig up to 16 notes before one is cut out; if that particular preset has a long decay, you end up with a lot of calculation going on in real-time—very high CPU usage.

Mode Density—The Mode Density control is located just below the resonator selector and it allows one to adjust the number of modes taken into account in the synthesis and therefore the richness and complexity of the sound. This control has four positions, Low, Medium, High, Full, corresponding to 4, 16, 30 and 70 modes respectively. CPU usage is directly proportional to the number of modes that needs to be calculated.

Pitch modulation via LFO and enveloppe—All modulations add extra calculations, and so increase the CPU load. If the modulation’s contribution doesn’t add anything to what you want to achieve you should remove it.

Sample rate—Running at 88.2 kHz vs. 44.1 kHz effectively doubles the CPU usage.

Buffer size—Here lies the compromise of CPU load vs playability. You would set it as high as it’s bearable. In our experience, at a sample rate of 44.1 kHz there is no point in increasing the buffer size above 512 samples—there is no increase in performance above that point.

 

But sure it sounds superb ....

[abacab]

13 hours ago, Zo said:

Hey guyz , late night here and don't ask me why lol decided to puagrde from chromaphone 2 to 3 (as well;string studio 3 and that modular thing , all for 34 $ witch was hard to resist) 

Cromaphone always been a superb synth but the V3 is barely usable due to the cpu hit , and don't even think about using pitch wheel ...

 

Am i missing something ? any infos from AAS on that subject ? what's going ? 2 instance of C2 are less cpu intensive than C3 !!

Sorry to hear it's not playing well for you!

I'll just provide my observations here.

Chromaphone 3 is now a 2 layer synth, so there is inevitably going to be a bigger CPU hit when a patch uses both synth engines.

I noticed that when opening the Chromaphone 2 factory library patches, that are provided in Chromaphone 3, there is only one synth layer enabled in each by default. I noticed that I get roughly the same CPU utilization as in Chromaphone 2 when using any of these legacy C2 factory presets in C3.

So when checking Chromaphone 3 for some performance data on my system, a 9th gen Intel Core i5-9600K CPU running at 4.5GHz, Chromaphone 3 does appear to use a single CPU core/thread so single threaded CPU clock is the most important factor here.

When auditioning  Chromaphone 3 factory patches using both layers, selected at random, it averages about 7-8% CPU hit. If I load a single layer patch from the legacy Chromaphone 2 library, the CPU runs about 5%. That's about the same thing that I get when I just load Chromaphone 2.

[chris.r]

This reminds me why I'm still holding off with upgrading to Chromaphone 3. I really like the upgraded sound, it seems to me that there's huge quality jump in them from v2 but all the different issues like memory leak or CPU hit made me stop for now. Testing the trial first is a good advice though.

[Zo]

1 minute ago, chris.r said:

This reminds me why I'm still holding off with upgrading to Chromaphone 3. I really like the upgraded sound, it seems to me that there's huge quality jump in them from v2 but all the different issues like memory leak or CPU hit made me stop for now. Testing the trial first is a good advice though.

Exact i don't know why i skipped trying , stupid as i am , but the gui is so much better to work with so i take it as a consolation , i'm gonna use it as a way to sound design in a better place than the V2 as simple as that ,going one layer instead ....

[chris.r]

What is your single core CPU clock, Zo? It might be useful information for future reference. Mine is 8x @4.0 GHz (6th gen).

The classy upgraded sound is very tempting.  

I agree with the voices that it kills production workflow. I have similar issue with a few other CPU hogs, like Amplitube for instance.

[Zo]

3.2 i disabled turbo boost 8 threads

[chris.r]

ugh... that's close enough, I'll trial it when the right time comes for upgrading. thanks

Edited December 26, 2021 by chris.r

[locrian]

Zo,

I suspect that you already know this but, another option -- and this works for all VIs -- is to load a low-CPU usuage patch close to what you want timbre-wise and use that during the writing process.  Then once your track is done, apply the finished MIDI data to the (higher CPU) patch you want, render that track to audio, and then disable the VI track with the high CPU patch (in case you want to go back and make changes later).

Is this an ideal/optimum workflow?  Hell no!  But it does work.

More and more I'm finding that the latest plugins (both VIs and FXs) are CPU hogs.  And it's not even just one company (which in the past for me was Acustica).  I'm now experiencing this with the latest plugins from IKM, NI, AAS, Waves, even FabFilter.  I'm not sure what's going on, but this seems to be the general trend (and one that I hope plugin designers are aware of and address).  It's really taking the joy out of making music.

[Zo]

3 minutes ago, locrian said:

Zo,

I suspect that you already know this but, another option -- and this works for all VIs -- is to load a low-CPU usuage patch close to what you want timbre-wise and use that during the writing process.  Then once your track is done, apply the finished MIDI data to the (higher CPU) patch you want, render that track to audio, and then disable the VI track with the high CPU patch (in case you want to go back and make changes later).

Is this an ideal/optimum workflow?  Hell no!  But it does work.

More and more I'm finding that the latest plugins (both VIs and FXs) are CPU hogs.  And it's not even just one company (which in the past for me was Acustica).  I'm now experiencing this with the latest plugins from IKM, NI, AAS, Waves, even FabFilter.  I'm not sure what's going on, but this seems to be the general trend (and one that I hope plugin designers are aware of and address).  It's really taking the joy out of making music.

yep i usually freeze , problem being when AAS are inserted time wise , if the project is already loaded , even just playing the part will force me to render other VSTi's , i don't want to kill creativituy because of lack of optimisation skills !!! usually bounce in audio once i validate the whole production 

[locrian]

3 minutes ago, Zo said:

... problem being when AAS are inserted time wise , if the project is already loaded , even just playing the part will force me to render other VSTi's ...

In a situation like this I'll actually use a low-CPU VI like Vacuum, Hybrid, or Halion Sonic and then swap these out for the high CPU VI before I start mixing.

Also not ideal, but ...

[Zo]

Just now, locrian said:

In a situation like this I'll actually use a low-CPU VI like Vacuum, Hybrid, or Halion Sonic and then swap these out for the high CPU VI before I start mixing.

Also not ideal, but ...

Yep , i can also don t use it , if aas was the only dev , ok , but man , even zenology and roland stuff are less cpu intensive ...

[Zo]

Just to be clear , people that are thinking of upgrading , do it , at least to have Chromaphone 2 remastered content and use it as chromaphone 2 but with a superb gui , if you need to go deeper you can but used in a ONE Layer configuartion cpu hit is OK !!

I'm actually creating patches now to conduct a real test and in one layers config it's smooth ...i'm gonna test one resonator only  in each layer now and report ...

Edited December 26, 2021 by Zo

[abacab]

I have seen similar feedback elsewhere regarding the latest generation of synths typically as being CPU hogs. And this history has repeated itself for the past few generations.

Progress!

But I will suggest that in the particular case of AAS, that they might consider making their synths multi-core aware...

Edited December 26, 2021 by abacab

[Starship Krupa]

1 hour ago, Zo said:

I'm actually creating patches now to conduct a real test and in one layers config it's smooth ...i'm gonna test one resonator only  in each layer now and report ...

Blew the dust off!

I did some experimenting with turning off layer 2. Since you can "solo" them, I found that in the incredibly dense sounds Chromaphone is capable of, the second layer wasn't contributing that much. Their sound designers create these sculptures using everything available, and sometimes in the context of a piece, they're too dense.

5 hours ago, Zo said:

3.2 i disabled turbo boost 8 threads

I was following the conventional wisdom on that for a long time, then tried enabling it and my processor clock shot up to 3.8 GHz (stock is 3.4) and stays there. This is on a Dell, which are notoriously clock blocked. So I'd encourage folks to experiment with that setting, monitor it with HWInfo, see what you get. The trick was to turn on the turbo boost, but keep the power saving one off.

4 hours ago, locrian said:

I'll actually use a low-CPU VI like Vacuum, Hybrid, or Halion Sonic and then swap these out for the high CPU VI before I start mixing.

Must confess at this point: I'm not (at this point anyway) much of a "synthesist." I call myself a "preset jockey." Mostly I turn off synths' internal reverb so I can replace it with R4 or Nimbus. My compositional process often involves browsing patches, finding an inspiring sound, then building the piece around that sound. It used to cause me a bit of shame, like "real" synthesists dive in and create their own unique sounds, blah blah. But then it struck me: some of my favorite composers' keyboards only had one "preset," and it couldn't be tweaked at all. ?

If A|A|S Player would only allow me to turn off reverb, I'd have much less interest in Chromaphone, which is probably why they don't allow it.

Hybrid 3 (great arps and sequences) is actually my most-used synth, and Vacuum Pro is up there too (some nice basses in there). XPand!2 is great for the kind of workflow you describe, and of course has sounds that can stand up to being used in a final mix (basses, arps, and pads).

As Cakewalk itself has gotten more nimble and resource-friendly over the last few years, I guess plug-ins have gone the other direction. AIR will probably never develop my favorite workhorses into the resource hog zone.?