hello folks,
i need some advice from the community here, if i need a new pc or how i find out why audio is so terrible crackling.
i am on ubuntu 24.04 and usually audio in vcvrack works fine.
But now my current patch seems to be too big.
after start the buildup on screen of the patch is slow and stuttering. audio is terrible stuttering too.
for further investigation i upload the patch, perhaps someone could see or give an advice what is going wrong.
i dont see any possibility to decide, if my graphic is too slow or the cpu.
currently i use AMD Ryzen 7 4700U with Radeon Graphics with 16MB Ram.
Laptop HP Laptop 15s-eq1xxx
please be considerate with my knowledge of patching. I`m not a developer.
this patch is WIP. currently my focus is to patch the Ciani Performance Patch inspired by @pyer and slowly integrate features from the Marf, as far as i understand . Because i dont own one i watch videos from YT-channel Sourceofuncertainty - the best source for Buchla modules.
usually i use one core. if more cores, the crackle increase.
the performance meter shows nothing special. i replace plateau, but no difference.
i will drop the frame rate. perhaps it helps.
will come back here and report.
thank you for the quick answer.
Holy cow - that patch is massive. I have a Win 10 Dell desktop with Intel 97-9700 CPU with 8 cores @3.00GHz with 16 GB memory. Certainly not particularly special but better than some. I can barely run the patch at 22.05 kHz - even at that sample rate the max CPU occasionally strays past 100%.
Amazingly, my bare bones original M1 MacBook Air with 8GB memory comfortably runs the patch at 48kHz with max CPU < 90%. You can pick up one of these at Walmart for $650. It is an amazing machine with excellent battery life and it runs silently!
You can spend more and get an M2 or M3 machine with more power, but much of that power is for better graphics processing which is totally not needed to run VCV.
did you turn on the CPU meters in VCV? Usually a small number of very inefficient modules are dominating. If thatâs the case, and you are lucky you can find a similar module that behaves itself.
While I can run it without crackling at 48kHz under Windows (using 6 cores and a dedicated sound card), Max. CPU tends to go through the roof and FPSs rarely go past 15 (with a nice video card).
Along with what Squinky and VirtualModular said above, I noticed some modules are sitting there without doing any actual work or connections (like the wire color, patch credits and such). They may take little CPU slivers; but they still count both for CPU and rendering purposes; also, maybe, not every Scope is needed (or could be substituted by a single, larger scope with more IO, as stated below).
Multiple copies of certain modules could be replaced by a single one that is either polyphonic; offers more IO ports, or is flexible in that regard (for example Venomâs multi merge and split can probably take over a few Merge and Split modules).
so big I suggest developers to make an utility module to highlight/find the AUDIO module too big for me (but it could be a reason for you to buy a new computer )
On my rizen 9 5900 cpu pc I can run the patch
with occasional crackles when I use 6 threads, more treads = more crackles,
I think it is a mixture of graphics and parallel processing problem, I use a dedicated graphicscard rtx4600 and it can only run the patch at 10hz.
So, yes the patch is pretty big, anyway I dont understand what for are most of the modules,
imho it should be possible to optimize the patch quiet a bit.
short explanation:
the AFG1 +2 Knobs (Arbitary Function Generator) produces ramps (stepped voltages).
with the help of @Yeager i managed it to produce forward and backward ramps. Moving knob CW or CCW.
this ramp is used to syncronize multiple modulation sources to every selected step of the marf clock/ramp.
As far as i understand the different âMarf-Filesâ from Sourceofuncertainty
the wonderful phasor modules helps a lot. Very useful!
ok, i think i see now. you analyzed that a better graphiccard will help, because the onboard gpu is overwhelmed. correct?
samplerate, blocksize does bring a bit, but not enough.
For a standard Windows laptop, builtin graphics is too weak for more than modest patches. As the requirements say:
Graphics: Dedicated Nvidia/AMD graphics card from 2013 or later with recent driver update
Integrated (non-dedicated) graphics such as Intel HD are not recommended and may cause significantly increased CPU usage.
So yes, a dedicated graphics card would help, but how much? No one can tell you how big a patch you can run on a given computer. You find out, try all the performance tips, and then you know. But in general:
A âgamingâ computer is always better than a âregularâ computer, for Rack.
A desktop computer is always better than a (equivalent) laptop.
An Apple Silicon laptop is always better than a (equivalent) Windows laptop.
Itâs simply a question of âin the end you canât beat physicsâ. The enemy is almost always heat.
Yes, you have multiple VCV Scope modules in the patch. I wonder, for example, if all of them are strictly necessary, if so, Count Modula offers one free module called Oscilloscope: bigger and with 4 signal inputs, for example.
this patch is WIP. currently my focus is to patch the Ciani Performance Patch inspired by @pyer and slowly integrate features from the Marf, as far as i understand 
. Because i dont own one i watch videos from YT-channel Sourceofuncertainty - the best source for Buchla modules.
too big for me (but it could be a reason for you to buy a new computer 
)