U-he ACE!

[rob_lee]

Thanks for trying to answer my question, guys; what I'm asking for is to be able to 'put a precise fence' around ACE's run-time environment, so that I can run it HOT in a live situation.

-goldenanalog

Just tossing a random thought, another option might be a preset folder(s) with only CPU-safe patches. A further variation on that might be putting those presets in a MIDI Programs folder.

nice idea..but

most people work on different machines therefore if i have a preset set to accurate mode on my machine and works perfect and you open it up on yours then the cpu may be fried
so when i design a sound im aiming for the best quality i can get with the tools i have because i tried last night to reduce quality on some of my own new presets which only resulted in my having to adjust a whole bunch of knobs..more so the resonance controls due to hiss.
the best thing to do is if we all buy the same computer spec hehe
good idea though.

BTW

my music computer is an old Q6600 no internet/porno files/phone homes
so i run perfect with no glitches.

cheers

Rob

[PaulSC]

Im trying to sync one oscillator to the main oscillator and have the synced oscillator's pitch modulated by an envelope is this possible????????????????????????????????? with ACE that is, and how

The "Template - Synced Lead" preset fits this description exactly.

Thankyou Thats good but i want it to sound like Underworld's Cowgirl

UnderCow.mp3

UnderCow.h2p

[xh3rv]

most people work on different machines

Yeah, I just meant something in user-space, not something that would make sense in the factory library at all.

[sophtrazor]

Im trying to sync one oscillator to the main oscillator and have the synced oscillator's pitch modulated by an envelope is this possible????????????????????????????????? with ACE that is, and how

The "Template - Synced Lead" preset fits this description exactly.

Thankyou Thats good but i want it to sound like Underworld's Cowgirl

UnderCow.mp3

UnderCow.h2p

U R AWESOME

I making some stuff based on that with ACE and its really good

Tanks again

[PaulSC]

You're welcome!

[eXode]

I just realise eXode maybe the same eXode who did some of my favourite reason refills back when I use to use reason.. If so, glad to have you onboard.....
rsp

Maybe you're right...

J/K ;)

Yes I've done a bunch of free refills for Reason 3 and upwards.
Thank you, I'm glad to be onboard!

[goldenanalog]

Just tossing a random thought, another option might be a preset folder(s) with only CPU-safe patches. A further variation on that might be putting those presets in a MIDI Programs folder.

What you say makes perfect sense, intellectually; but the CRAZY thing about (at least) my computer is that the 2 dozen or so patches that I played with were set for 'medium' (8-voice) and 'accurate' (HQ real-time rendering) and used almost EXACTLY the same amount of CPU resources...~58.7%!

I guess that's the impetus for requesting a 'fixed' # of voice(s)...it looks like at runtime, ACE uses a specific amount of CPU resources based on those settings...polyphony and RT render quality....and stays squarely in that 'operational zone'.

I saw very little %cpu movement from that figure when I played 8-voiced chords. It was impressive, actually.

I think that Urs built us a real jewel. Then again: What I'm requesting has been in implementation for years...in Timewarp 2600, for example, you can set the precise # of voices that 'play' at runtime...TW26K is also a rampant resource hog; higher polyphony settings will 'stall' my quad, no prob. So reducing the # of voices at times yields a reasonable comprimise...real-time use with no probable danger of stall, in exchange for reduced polyphony.

-goldenanalog

[xh3rv]

I guess that's the impetus for requesting a 'fixed' # of voice(s)...it looks like at runtime, ACE uses a specific amount of CPU resources based on those settings...polyphony and RT render quality....and stays squarely in that 'operational zone'.

Ya, Urs mentioned somewhere in this thread I think - if I understand correctly, any patches with equivalent stacking and quality setting will have equivalent CPU usage if all voices are being used, it does not vary with use or signal flow between modules. Number of voices is dependent on MIDI input but has an upper bound defined by the polyphony control. These parameters are all conveniently grouped, and grouped away from the experimental area of the synth, very nice design - so long as one is comfortable with the CPU performance, there's no chance of a CPU hit from cabling or tweaking modules at any point.

[Urs]

What I'm requesting has been in implementation for years...in Timewarp 2600, for example, you can set the precise # of voices that 'play' at runtime...

Hehe, of course it seems easy at first sight to implement precise control over the number of voices. I have however deliberately decided against that, right from the early days (2003 or so). There are situations and functions that clash with fixed voice settings.

For instance, setting number of voices to 5 and using voice stacking of 2 would lead to confusion. Would that be 10 voices, 4 voices or would the third stacked voice be played with only 1 voise? The more abstract form of Few-Many-More circumvents this dilemma and simply assign a number that works.

Another thing that prevents me from adding exact voice counts is the mode setting itself. While it might not be obvious in ACE, it is obvious in FilterscapeVA and Zebra. There you have an arpeggiator with its own polyphony setting. This setting can exceed the normal limit of "Few" in poly mode. Hence, when the arpeggiator is turned on, the actual number of voices is adjusted internally to accommodate the possible arpeggiator settings.

Between you and me, you may believe me or not, possible areas of confusion like these two points typically lead to *numerous* requests, suggestions and even bug postings, hence these issues are rather avoided. No one has ever complained that the arpeggiator in Zebra uses only 4 voices even if set to 6. Because it doesn't.

Cheers,

Urs

[goldenanalog]

Ya, Urs mentioned somewhere in this thread I think - if I understand correctly, any patches with equivalent stacking and quality setting will have equivalent CPU usage if all voices are being used, it does not vary with use or signal flow between modules. Number of voices is dependent on MIDI input but has an upper bound defined by the polyphony control. These parameters are all conveniently grouped, and grouped away from the experimental area of the synth, very nice design - so long as one is comfortable with the CPU performance, there's no chance of a CPU hit from cabling or tweaking modules at any point.

Beautiful! Excellent UI/infrastructural design!

And well-put, xh3rv! Got it!

OK: So if what you are saying is indeed true (fingers crossed), it stands to reason that if the P&R (polyphony and render quality) settings had precision, you could set your P&R pre-show, and then go so far as doing knob-twisting and cable-pulling ON THE FLY (WOW! ) in a performance situation with little/no chance of 'stall!'

You just made my case, AND my evening, xh3rv!

-goldenanalog

[pdxindy]

Just tossing a random thought, another option might be a preset folder(s) with only CPU-safe patches. A further variation on that might be putting those presets in a MIDI Programs folder.

What you say makes perfect sense, intellectually; but the CRAZY thing about (at least) my computer is that the 2 dozen or so patches that I played with were set for 'medium' (8-voice) and 'accurate' (HQ real-time rendering) and used almost EXACTLY the same amount of CPU resources...~58.7%!

Use different stack settings will change that though...

[goldenanalog]

Hehe, of course it seems easy at first sight to implement precise control over the number of voices.

Oh, no, Urs, I'm not thinking that it's 'easy'....far from it. My request is based on logic, circumstances, & history; I am not the developer, and make no assumptions.

It's up to you...your product, your implementation, Urs.

I have however deliberately decided against that, right from the early days (2003 or so). There are situations and functions that clash with fixed voice settings.

For instance, setting number of voices to 5 and using voice stacking of 2 would lead to confusion. Would that be 10 voices, 4 voices or would the third stacked voice be played with only 1 voise? The more abstract form of Few-Many-More circumvents this dilemma and simply assign a number that works.

Use even numbers! Of course I'd be willing to settle for a reasonable comprimise, as if it really mattered what I wanted.

But I'll ask anyway: How about 1/4/8/12/16?

Another thing that prevents me from adding exact voice counts is the mode setting itself. While it might not be obvious in ACE, it is obvious in FilterscapeVA and Zebra. There you have an arpeggiator with its own polyphony setting. This setting can exceed the normal limit of "Few" in poly mode. Hence, when the arpeggiator is turned on, the actual number of voices is adjusted internally to accommodate the possible arpeggiator settings.

Understood.

Between you and me, you may believe me or not, possible areas of confusion like these two points typically lead to *numerous* requests, suggestions and even bug postings, hence these issues are rather avoided. No one has ever complained that the arpeggiator in Zebra uses only 4 voices even if set to 6. Because it doesn't.

Cheers,

Urs

The power of absolutes! But you built Frankesteins' Monster, the genie is out of the bottle, blah blah blah...Urs: ACE is not a lightweight synth that runs comfortably with multiple polyphony on ye olde laptop; ACE demands a lot (~58.7% ), and is right up there with the best software synths developed to date.

I won't bore you, or waste everyone's time with my (pretentious) list of softs & other stuff (what does that mean, anyway?); suffice it to say that I've been into pianos/organs/keyboards/computers etc etc for literally decades, so I'm telling you from what amounts to vast exposure, that ACE is a H*LL of an achievement.

So really: it's about doing the right thing....I'm pushing you, so that you will (hopefully) push ACE as far as it possibly can go.

Urs: Thanks for giving a d*mn about what I say! I'm a fan!

-goldenanalog

[goldenanalog]

Use different stack settings will change that though...

Yes, but in my experience thus far, your 'voice' resources stay the same; ie (1) stack, 8-note poly; (2) stacked, 4-note poly, etc.

-goldenanalog

[PaulSC]

A half-dozen presets -- might as well post them since I've stayed up way too late making them...

[Hlis93]

Thanks I'll give them a try in the morning