Home | ER-101 | ER-102 | ER-301 | Wiki | Contact

Clicks...cpu 66%

support

#21

so this sadly continues to be a problem, now with the delay out the equation i still have some very loud clicks:

is this expectable at cpu 67%? i now have an entire chain bypassed so i can split the performance into two main parts but the clicks do worry me and i’m thinking of changing the whole set which is sad since the er301 is so well integrated now. i do use quite a few samplers as wavetable oscillators and i think this eats up the cpu heavily. anyway, what i would like to know is if this is something to be expected with this cpu usage or is it weird and should not occur? i once drove it up to over 90% and the noises and behavior then was clearly different and it affected everything, sound, clocks, speed etc.


#22

It’s definitely weird at just 67%. Is it possible to get access to your quicksave for further analysis?


#23

yes sure, here is the zip:

slot7.zip (26.1 KB)

i don’t know if that’s enough or do you need the samples too?

thanks brian


#24

Best wishes sorting this out - sounds like you have put a lot of work into it @kilchhofer and I would like to hear what you have been up to :wink:


#25

Thank you! I’ll see what I can understand without the samples first.

Q1. The cpu67% clicks.wav was recorded from OUT1+2?
Q2. The cpu67% clicks.wav was recorded with the Manual Grains unit (in chain OUT3+4) bypassed and disabled?
Q3. If no, at what approximate rate is the Manual Grains unit being triggered (G1)?


#26

yes, recorded from out1+2 with chain 3+4 eg the grain unit disabled. g1 which drives the whole patch is the speed of the marimba 1/16th pulses…don’t know, maybe at 125 bpm or so.


#27

thanks!


#28

Could you try this firmware when you have a moment?

er-301-v0.2.14p3-unstable-48kHz.zip (1.3 MB)


#29

i try it now, thanks! what is new in this version?


#30

ok, it seems to work a bit better but i still have a few clicks, not as many as before though, this of course with grain and looper deactivated.

however when i enable the looper and grain unit on channel 3+4 the clicks get more severe again.
cpu with everything active (which should only last for the transition, maybe one minute) is at around 78% and the clicks sound like an old record again, really loud and somehow looped?!
(i made sure to zero the buffer and no signal came into the looper or grain, so it definitely wasn’t a recording of crackles playing):


#31

Are you clipping out the looper? Sounds like digital overage.


#32

+1 @NeilParfitt

Agreed… my sound card does things like that when I am driving it too hard at some point in the chain!

I just go through the entire patch from sources through to the sound card and check all the gain staging… I’ve invariably cranked something up at some point!

Sounds good otherwise (and I kinda like the crackles too tbh) :smile_cat:


#33

no, nothing in the looper, no signal at all.


#34

If possible I would like the sample files that go with the quicksave.

Also, could you do the following experiment to check your hardware, please?

  • Configure your channels as 4 mono chains, all empty.
  • In the OUT1 chain insert a Sine Osc (f0=220Hz) and a Stress Unit.
  • Listen to OUT1 and start slowly turning up the load on the Stress unit.
  • Note the load parameters value when you hear your first pop.

For me this is 0.91 and the CPU load is at 94%.
(I’m assuming we are using the 48kHz firmware.)


#35

ok, i send you the samples tonight and do the stress test then. thanks


#36

hey brian,
here are the samples of my quicksave nr seven, two single cycle waveforms and a marimba sample:

i made the stress test and i heard the first pop at 0.894 but just one single pop, at 0.9 the pops were consistent, cpu 94%, and yes, always with the 48kHz firmware.


#37

Thank you for the samples and the stress test.

Here is another firmware for you to try:

er-301-v0.2.14p4-unstable-48kHz.zip (1.3 MB)

I am optimistic about this one!


#38

thank you very much!
i can try on wednesday and will report back.
i have another short question: when i load a quicksave some samples start playing without incoming trigger, actually when i think about it only the samples without slicepoints start playing, could it be that they receive some sort of ghosttrigger when the quicksave loads, or is this a problem because they don’t have slices? so for example i made a new patch with the marimba bounced down to one single file, just one marimba hit, so i didn’t need to slice the audiofile, this sampleplayer however now loops the marimba when the patch is loaded, as soon as i start the clock, eg the triggering and stop again the looping stopped too, but on startup it just loops?
thanks brian


#39

If there’s no slices, the sample player is free running and just plays the moment a file loaded by design.

I think you’d have to have at least a single slice set for it to not auto-play.

@odevices can confirm or correct me on this one.


#40

hm, yes i thought so too, but when i start the trigger and stop it again it doesn’t loop anymore. i guess triggering disables the looping. anyway, i just put a slice at the beginning. thanks @NeilParfitt!