V0.2.x-stable: Thank you!

i get wrong slice addressing when using 12tet with the er101, but only occasionally…
this happens with or without delaying the incoming triggers.

Re: missed slice triggers

I’m assuming that this problem is restricted to 12ET addressing mode so I’ll be taking a close look at the 12ET slice mode over the next few days. If anyone has had similar problems with the original index addressing mode, please let me know.

edit: Actually I see that @miminashi has previously reported the same issue on the index addressing mode also.

Most likely CPU usage is spiking over 80% but too briefly to show on the CPU load screen. The CPU load screen samples the load once per second so it can easily miss short spikes in CPU usage. I’ve tried to build all units so that their CPU usage is as constant as possible, except for one unit: Manual Grains. That unit’s CPU usage will depend on the number of simultaneous grains playing at any particular moment (from 0 to 16) which will depend on the interaction of grain duration and grain triggering.

Yes, I noticed high CPU usage (84%) when using just two manual grain units, looper, and a clocked delay. Killer work on the grains unit Brian, it has been super fun to explore!

I got rid of one of the mixer channels with a Manual Grains (and a couple of Tri Osc modulating it) and the crackling went away.
One thing I noticed is that the crackling was only happening in the mixer channels that I was using for external inputs (as I am using my 301 as my main mixer).
When I soloed the Manual Grains mixer channels there wasn’t any crackling.

When you got up to 84percent usage did you notice any crackling sounds at all?
Kind of the same as when you are pushing a DAW to its limits CPU wise.

The missing slice 0 triggers should be fixed now. I suspect this fixes @kilchhofer’s 12ET problem also. If not let me know.

I also had implemented the parameter for skewing the grain envelope in the Manual Grains unit but unfortunately the effect was just too subtle and not worth it in my opinion. So I took it back out.

v0.2.4 CHANGES

  • FIXED: Inconsistent triggering when slice 0 is at the exact beginning of the sample.
  • FIXED: Displayed slice number was off by one in the slicing view’s slice list (but was correct in the sub display).
4 Likes

One thing I have just noticed is that muted mixer channels are not saved as part of the quicksave.
Upon reloading the channels that were muted when saved are unmuted.

Yes, I had the crackling sound for sure.

@odevices, V0.2.4 is sounding good. I had a tense moment when the 301 failed to reboot after I switched from the 48kHz to 96kHz firmware, but it booted up fine after a power cycle and then a reversion to 48kHz (no power cycle), so all is good in the hood.

i still have wrong slices being triggered with v0.2.4, i don’t have a slice at the beginning of the wav file. i test some more, it seems to happen really randomly and rarely, twice in appr. 10 minutes playing.

do you mean you were going to implement different grain envelopes such as square/hard cut and such? i’d be surprised if that was difficult to discern, so maybe that’s not what you were doing… but i’d love to see the toption of going with a no-cross fade hard cut grain to any slower cross fade all the way up to right below the grain length. also max grain times of more than one second… 2, 4 or something… maybe that gets memory intensive…

I had implemented a skew parameter for the grain envelope which lets you shift the location of the envelope’s peak from front to back in a continuous fashion. It sounded great when the grains were isolated (non-overlapping) and durations were longer than 100ms but the effect quickly disappeared when grains started overlapping. So instead I will spin this skewed sine envelope off as its own unit as an alternative to the existing ADSR unit.

At the moment, I’m mainly interested in finding an envelope parameter that can be swept continuously rather than just an enumeration of different envelope types. Morphing was considered but that is rather expensive since it means you have to calculate at least 2 envelopes per grain for linear interpolation.

1 Like

Bug 0.2.4:

I sliced up a sample and used the 12TET slice addressing to choose the slices - so far so good, worked as expected I.e. index 0 on the ER101 equates to slice 0 and so on - most happy :grinning:

I then added a second sample player, assigned the same sample and assigned all the same modulations (trigger & CV from the ER101). Now the slice addressing this is all over the place I.e. the same sequence of slices is not played. For example playing a sequence of slices 1,1,1,6 now gives me something like 1,1,1,23. Really odd.

I saved the original sample player as a chain and reloaded & repeated adding the second sample player - same problem.

May I have some more details on what you mean by [quote=“vcoadsr, post:101, topic:386”]
I then added a second sample player…
[/quote]

Inside a mixer? In another chain? Did you set the addressing mode to 12ET?

Ahhh, I think you just isolated the user error!!

Apologies. :worried:

0.2.4

Newly Saved chains put in the chooser menu no longer populate the Unit selection list.

Worked in prior versions… or maybe I’m totally wrong… did this only ever work on a per unit basis??(ie: saving a mixer that had a ton of stuff within?)

Hrmm… now I’m feeling hazy.

Chain presets never worked in the chooser menu, only unit presets. However, that is a good idea.

1 Like

if it’s an easy tweak in a future update… yeah that would be sweet!

Loosely-related: is some sort of file browser coming back to the admin area? It would great to be able to move or delete presets throughout the 301 without relying on a computer.

1 Like

I just had an issue where I added a limiter to the end of a chain controlling the pitch of a sine wave and suddenly the machine made really awful loud clicks and then stopped. I deleted the limiter but the oscillators were not outputting any sound. I did a power cycle and added the limiter again and no problem. Just wanted to point out my weird experience.