Expert Sleepers disting EX

[Blond 12]

So here is some conclusion of all my afternoon test.
First of all im sure now its a SD card issue + Buffer Disting management issue reading Wave files

I tryed 4 SD card with différents caracteristics in term of reading spead.
( I put a pics with the 4 SD card used from left the worst to the right the best.)

Samsung Evo 32G provided with the disting Ex:
- SD speed test : Min: 2.92/ Max 2,96/ Avg 2.93.
- Sound report : the worst one , like i said before the module just play few note and stop , then come back with latency then crash totally with obligation to reboot after main freeze.

Kingston 16G :
- SD speed test : Min: 2.89/ Max 3/ Avg 2.90
- Sound report : way better, the module now crash one time after 3 min play. At the différence with the samsung one the module now play all notes , he only make some "clicks" at some point when you play too much notes at same time. Looks like a buffer or "cpu full "noise alert.

PNY 32G :
- SD speed test : Min 2.38/ Max 3.03/ Avg 2.54
- Sound report : Again a bit better, module never crashed, but i have still some "clicks if i play or stress with too much notes at the same time depending the preset. You can really hear that with piano or string waves. If you reduce the enveloppe time you can win a bit of tolerance of that buffer glitch.

SANDISK ULTRA 128G :
- SD speed test : Min 1.86/ Max 1.91 / Avg 1.89
- Sound report : The best one, No crash or freeze at all, way more tolerance and " headroom" with lot off notes pressed at same time. Still some clicks if you play too much quick with too much notes.

Conclusion : my conclusion is a bit weird cause opposit of what the manual said about SD card recommended speed.
On my side more the SD card his low speed and far of the 3Mb/s target and better it is .

I hope it will help OS to find the problem.

[os]

Thanks. I'm off over Easter but I have some ideas for what to try when I'm back at work.

[recliq]

Adding on this topic..

It seems I mostly have problems when changing sample folders too fast, the disting tends to freeze on various folders while loading the samples- This doesn't seem to be related to specific folders...
It would be much easier to scroll through the folder list when there was a larger pause before loading the actual folder... maybe even 2 to 3 seconds or even another action required to actually load the current folder.

[metajon]

Hey
Yes I did that with DiskUtil. Also tried it with my old PC but there was no different in Performance

reformattet the Stock SD (kingston) with FAT32 and 32KB cluster

When you reformatted the cards did you also partition them using "Master Boot Record"?

[chart]

For those still having issues, please try the test firmware I posted above:

https://github.com/expertsleepersltd/is ... st.hex.zip

Sorry these issues are dragging on. The EX is a complex thing, especially when you factor in the SD cards and the potential variability in their performance. At some point I may just have to gather the problematic units here.

Hello Os,
((I jut read your off over Easter no worries ))

Thank you for your quick reply.
I totally understand this must be one of the most complex modules out there and it must be a pain trying to find a balance between troubleshooting, working on new algorithms, and keeping up with all of our requests and complaints.

I installed that firmware (v1.7.2inst as you can see on the vid) before my first post.
I uploaded a YT video to demonstrate the problem. I tried 4 and 8 voices (both with the same problem although 4 voices was much worse) and changing the gate offset. The audio was directly recorded from the disting ex into my phone. After this test I tried to find a way to go back to normal so I tried loading a different multisample folder and when I did that the module crashed and I had to power off the system.

Here is the YT link to demonstrate the problem
(this is an unlisted video so only people with the link will be able to see it, I don't want to spread bad publicity about this amazing module)


Thank you for taking the time to respond to all of us

[metajon]

Okay i Tested now that 1.7.2inst FW. Still have drop outs and freezing
Also once the module stop playing bc of the glitches i cant make an SD Card Test in misc menu. After some hangups the module freeze completly
I tested the Samsung EVO select 64GB and the Stock Kingston 32GB and there was no differnce.

hope that helps a bit

[BlackBearDiner]

Adding on this topic..

It seems I mostly have problems when changing sample folders too fast, the disting tends to freeze on various folders while loading the samples- This doesn't seem to be related to specific folders...
It would be much easier to scroll through the folder list when there was a larger pause before loading the actual folder... maybe even 2 to 3 seconds or even another action required to actually load the current folder.

I have the same experience too, and been thinking that it would be good to be able to scroll through folders and choose with a push rather than that it loads automatically. Unless there already is a setting for that and I just should rtfm.

[banjodan]

On a somewhat related tangent, it would be helpful if, for modes that have the Folder parameter, the folder could be selected via MIDI program change (from the first 128 folders in the list).

[Smapti]

On a somewhat related tangent, it would be helpful if, for modes that have the Folder parameter, the folder could be selected via MIDI program change (from the first 128 folders in the list).

You can do this via a midi mapping.

[os]

You can map a MIDI CC but not a program change. (A program change can currently be used to change preset or algorithm, so you could save a whole load of presets with different folders.)

[Lera]

Trying to use the wave scrubber algorithm. every time I try to scroll through the wav folders it seems to randomly stop on folders I am not trying to get into, then I have to wait for that folder to load sometimes large files of wavs before I can continue scrolling. also I fairly regularly get total freezups attempting this. firmware 1.7.2

[os]

That's one of the algorithms that has to prepare reversed versions of the WAVs on the SD card, which can take some time. Leave it running once it starts, or you might end up with a scrambled SD card.

[recliq]

Some more info about the multi sd problems...
i just played around with some arpeggios played via midi and got the scrambled sound followed by freeze problem.
i went back to the latest 1.5 fw ( the beta with the fixed filename bug) and played around for an hour without any problems. i' m gonna test with newer firmwares now starting with 1.6 to check where the trouble starts...

[_gasStationShrimp]

Does anyone know where I can get the original factory content for the SD? Thanks!

[recliq]

Another observation in SD multisample: if I use chordmode and octave up or transpose up I start to get scrambled sounds for the higher notes.
If i transpose the midi notes everything plays fine one or two octaves higher. IT seems to happen only in the combination (octave up + chord mode or transpose up more than 8 semitones + chordmode).

[os]

Here's a beta to try:
https://github.com/expertsleepersltd/is ... a1.hex.zip

I've added a new setting, 'Enable High Speed SD', which defaults to '0' (off). Previous versions always let the SD card use High Speed mode if it claimed to support it. This lets you turn that off. The theory we're testing is whether that improves reliability.

Even without High Speed mode a good card will still support full polyphony in SD Multisample and SD 6 Triggers with the stock content.

[chart]

Here's a beta to try:
https://github.com/expertsleepersltd/is ... a1.hex.zip

I've added a new setting, 'Enable High Speed SD', which defaults to '0' (off). Previous versions always let the SD card use High Speed mode if it claimed to support it. This lets you turn that off. The theory we're testing is whether that improves reliability.

Even without High Speed mode a good card will still support full polyphony in SD Multisample and SD 6 Triggers with the stock content.

This is working pretty good for me. I tried a bunch of different things and it just kept going for over an hour.

[πασκαλ]

Hey
Yes I did that with DiskUtil. Also tried it with my old PC but there was no different in Performance

reformattet the Stock SD (kingston) with FAT32 and 32KB cluster

When you reformatted the cards did you also partition them using "Master Boot Record"?

Formatting it with this program might be worth a try: https://www.sdcard.org/downloads/formatter/

[recliq]

Here's a beta to try:
https://github.com/expertsleepersltd/is ... a1.hex.zip

I've added a new setting, 'Enable High Speed SD', which defaults to '0' (off). Previous versions always let the SD card use High Speed mode if it claimed to support it. This lets you turn that off. The theory we're testing is whether that improves reliability.

Even without High Speed mode a good card will still support full polyphony in SD Multisample and SD 6 Triggers with the stock content.

Working good for me as well with High Speed Mode off, if I turn it on the SD Speed Tests reports 2x Speed but the problems start sooner or later with SD multisample.

But can you tell what is the difference betweent using the setting octave +1 and playing notes an octave higher? It still makes a diffenrence in chord mode...

[os]

Is this specifically when using MIDI? Not using CV/gate?

[recliq]

No it's not specific to midi, but way more intense via midi, problem exists also via cv/gate.
It's not even a big deal for me, because I can just transpose the midi... was just wondering whats going on...

[os]

There was definitely an issue with applying octave/transpose to MIDI (even without chord mode), which I believe is fixed in this firmware:
https://github.com/expertsleepersltd/is ... ta.hex.zip

I can't repro anything bad with transposing CV/gate (with or without chord mode). Could you give more specific details of how to repro the issue?

[recliq]

Wow, it's fixed!

I can't reproduce the cv/gate issue right now, most likely I did something wrong last time...

Thank you so much for your great support!

[Blond 12]

I'll do a few hours check tonight on 1.8 with the original SD card Samsung SD tho... Stand by

[paoling]

For me the multisample mode is unusable too as it is. However, this evening I converted all the samples to Mono and resampled them to 16khz. Obviously, since the files are reduced the issue disappeared. However there is some strange tuning issue between the different samples. So I'm going to try other sample rates (like 22050 or 24000) and check if I solve also this problem. Sure that's a sub par solution, but I think it's preferable than having a crackling output.
Make sure you backup the SD Card if you try this.