Reading the DSP meter

[auricle]

Hi,

Just wondering how to interpret the DSP meter properly. What's the Sync and ASync telling us? And what's the upper limit of these readings?

Thanks

[dawman]

Left column is where the # 7 -12 numbers ahould be of concern.
The right column only starts to load up because of Windows from what I have been told.
6500 is the number to stay under that has worked for me.
The first 6 DSP's are for comaptability and are the older Sharcs. VDAT. VRC128 & VRC-S are used there. They seem happy under 1200.
I have great success by trying to keep as many devices under the max 6500 as possible.
For example, I use a mixer on 8, then routing window effects, including Modular effects on DSP 9, 2 x B2003's on 10, 11, 12, 13, 14, 15, and 16 Modular synths.
This allows me to then use DSP 7 as the way to load really hungry synths that need to span across the DSP's. Prowave w/ 2 voices default, assign it to DSP 7, then clear the assignment and start packing on the poly.
When I save the project I make sure PWave has 2 voices.
These are workarounds until unoptimized devices become re made or better optimized.
This is also why I am using Modular more. I can really get large patches w/ high polyphony. On my cards I couldn't anywhere near the depth and power I am getting w/ Modular now.
Unfortunately, I cannot achieve the infamous JBowen PolyGlide analog feature in Modular. So QWave and Prowave still have their places depending on the gig.
I also use the FP28C and MB4-II SpaceF mixers. I have grown so fond of the AUX mixing way of Scope I cannot use other mixers. Wolf's mixers require spanning but I will check them out with the new version next week.
Because of SpaceF and Sharcs great BCF, and BCR templates I have to get a pair of those.
I see me using the hardware Solaris synth for my authentic Analog needs w/ the SE-1's, and Modular, B2003 and SpaceF for everything else. To not take advantage of the 14bit MIDI and control surfaces would be a waste.
I have already made some powerful 8 I/O Modular synths where the surface controllers would be a cinch to use as mute/bypass switches on Modular and also for genereal mixing needs.
When Solaris is ready I will lose a KS88 which means many CC controllers. Hence the desire to use the Behringer stuff. Plus it's dirt cheap and perfect disposable controllers for stage.

[Mr Arkadin]

The first 6 DSP's are for comaptability and are the older Sharcs.

Not everyone has XITE Jimmy

[auricle]

Left column is where the # 7 -12 numbers ahould be of concern.
The right column only starts to load up because of Windows from what I have been told.
6500 is the number to stay under that has worked for me.
The first 6 DSP's are for comaptability and are the older Sharcs. VDAT. VRC128 & VRC-S are used there. They seem happy under 1200.
I have great success by trying to keep as many devices under the max 6500 as possible.
For example, I use a mixer on 8, then routing window effects, including Modular effects on DSP 9, 2 x B2003's on 10, 11, 12, 13, 14, 15, and 16 Modular synths.
This allows me to then use DSP 7 as the way to load really hungry synths that need to span across the DSP's. Prowave w/ 2 voices default, assign it to DSP 7, then clear the assignment and start packing on the poly.
When I save the project I make sure PWave has 2 voices.
These are workarounds until unoptimized devices become re made or better optimized.
This is also why I am using Modular more. I can really get large patches w/ high polyphony. On my cards I couldn't anywhere near the depth and power I am getting w/ Modular now.
Unfortunately, I cannot achieve the infamous JBowen PolyGlide analog feature in Modular. So QWave and Prowave still have their places depending on the gig.
I also use the FP28C and MB4-II SpaceF mixers. I have grown so fond of the AUX mixing way of Scope I cannot use other mixers. Wolf's mixers require spanning but I will check them out with the new version next week.
Because of SpaceF and Sharcs great BCF, and BCR templates I have to get a pair of those.
I see me using the hardware Solaris synth for my authentic Analog needs w/ the SE-1's, and Modular, B2003 and SpaceF for everything else. To not take advantage of the 14bit MIDI and control surfaces would be a waste.
I have already made some powerful 8 I/O Modular synths where the surface controllers would be a cinch to use as mute/bypass switches on Modular and also for genereal mixing needs.
When Solaris is ready I will lose a KS88 which means many CC controllers. Hence the desire to use the Behringer stuff. Plus it's dirt cheap and perfect disposable controllers for stage.

Thanks for that great explanation, Jim. It's great to have a figure to work to - now that I know to go up to 6500 max will make my life a bit easier. BTW, I've heard Wolf's stuff mentioned a lot but where do I get it from?

You also mentioned DSPs up to 16 but what about 17 and 18?

[dawman]

http://wolf-audio.com/

I got everything he made and then some.
His mixers were working fine until the last upgrade.
So once the final release is done I'll see if his stuff works, if not perhaps he will do another fix.
He has a 32 Channel mixer meant to be used w/ Mackie control. It's a work of art but I can't turn all 32 channels on.
I could until August.
But he always jumps in for the big win if needed.
Until then if you need a good step sequemcer outside of Modular he has a couple in his MIDI Tool Box which is sweet.
But if you have Mehdi's MIDI stuff you might not need it.
But the Matrix Delay is a fierce bastard. There's nothing it won't do..
Totally unique in the Scope world too.

[voidar]

Well, in with Scope there are synchronous and asynchronous variables.

Synchronous variables update in real time and is typically used to represent the audio signal, VCA, LFO etc..

Asynchronous variables are updated "when needed". Typically a couple of times during a second. Used for parametere modulation, like when you turn a knob.
Effects based on FIR and IIR will also need some asynchronous variables due to the design nature of the unit delay in Scope.

So what the DSP meter tells you is how much of what is loaded.

[auricle]

Well, in with Scope there are synchronous and asynchronous variables.

Synchronous variables update in real time and is typically used to represent the audio signal, VCA, LFO etc..

Asynchronous variables are updated "when needed". Typically a couple of times during a second. Used for parametere modulation, like when you turn a knob.
Effects based on FIR and IIR will also need some asynchronous variables due to the design nature of the unit delay in Scope.

So what the DSP meter tells you is how much of what is loaded.

Are there upper limits to these variables? For example, Jimmy says that 6500 for the synchronous seems to be the limit. However, in my current project I have a reverb running that takes the asynchronous variables up to the hundreds of thousands!!

[auricle]

In scope cards the limit of each DSP is around 1100 sync.
In Xite a newer DSP type (actually 2 types) is used that should be much higher.

The verbs usually need system memory and therefor it makes sense that you get high async values since that kind of Xite to system memory communication is significant.

Ah, that makes sense. Thank you. I need to learn how the whole Xite system works so I can better utilise it more effectively.