Unexpected behaviour when changing programs

osflaa · « **on:** June 30, 2018, 02:01:34 AM »

I have saved "Redn Blue Sweed" in U3P99, "Ultra leader" in U3P100 and "Blood Vain" in U3P101. But "Ultra Leader" sounds completely different when I am going from "Redn Blue Sweed" compared to when I am going from "Blood Vain". How is this possible? Am I doing something wrong?

jg666 · « **Reply #1 on:** June 30, 2018, 02:54:39 AM »

What version of firmware are you on? I think I remember that one version did have problems with effects carrying on when you change to another patch.

osflaa · « **Reply #2 on:** June 30, 2018, 03:35:24 AM »

Quote from: jg666 on June 30, 2018, 02:54:39 AM

What version of firmware are you on? I think I remember that one version did have problems with effects carrying on when you change to another patch.

I was on 1.1.0.3 but updated to 1.1.4.4 today, but it did not change this behavior. And the differences are more than just the effects.

jg666 · « **Reply #3 on:** June 30, 2018, 03:49:22 AM »

I think you would be best to do a video of this if you can as it’s hard to understand what problems exactly that you’re experiencing.

Another common solution to a lot of problems is to do a recalibration of the oscillators, although I’m not sure if that would fix a problem like the one you are having.

Gerry Havinga · « **Reply #4 on:** June 30, 2018, 04:15:21 AM »

Quote from: osflaa on June 30, 2018, 02:01:34 AM

I have saved "Redn Blue Sweed" in U3P99, "Ultra leader" in U3P100 and "Blood Vain" in U3P101. But "Ultra Leader" sounds completely different when I am going from "Redn Blue Sweed" compared to when I am going from "Blood Vain". How is this possible? Am I doing something wrong?

This is not behavior I recognize.

If you are using the Rev2 bi-timbral, with separate A + B audio routed to your mixer, and driving the Rev2 in multi mode over 2 MIDI channels you might be hearing layer A or B unexpectedly. This can be confusing especially for layer B when you are working on layer A and B is still set to init patch or it's factory settings. I hope this helps, quite far fetched but I thought I post it in anyway

mdts · « **Reply #5 on:** June 30, 2018, 05:28:44 AM »

If it sounds like heavily distorted it's a bug that DSI has completely ignored and/or failed to fix so far.

I've had it since I bought the synth last year and frankly I've given up hope it will ever be fixed.

If you read my thread https://forum.davesmithinstruments.com/index.php/topic,2073.msg23411.html#msg23411 you will notice some presets that have distortion effect enabled will "corrupt" previous or next patches when browsing, heavy distortion will be applied to them as well.

I found the Blood Vain patch (U3 or F3-29) to be the biggest offender sometimes messing up all patches and forcing me to power cycle the synth.

I assume it is fixable most likely being a software bug since the effect is digital, usually turning off the effect section then browsing to the previous/next preset will not cause the corruption. Also an easy way to find patches that manifest like this is to turn on the sequencer and just browse through.

osflaa · « **Reply #6 on:** June 30, 2018, 05:37:07 AM »

Audio output B and midi are not connected and the programs are not changed, just saved in a user bank. Here is a video that hopefully explains what I mean.

mdts · « **Reply #7 on:** June 30, 2018, 05:41:57 AM »

Yeah, it's the bug I mentioned.

Here is another thread where a DSI employee acknowledges the bug and even recommends staying clear of patch U3-29 https://forum.davesmithinstruments.com/index.php/topic,1959.0.html.

He mentions firmware 1.1.0.1 fixed it but obviously it didn't. You could try emailing dsi support and direct them to this thread.

osflaa · « **Reply #8 on:** June 30, 2018, 05:45:34 AM »

Quote from: mdts on June 30, 2018, 05:28:44 AM

...usually turning off the effect section then browsing to the previous/next preset will not cause the corruption.

Yes, you are right, this works! If this is a bug I hope it will be fixed. It would be nice to be 100% sure that every time I load a program it will sound exactly like it did the previous time.

osflaa · « **Reply #9 on:** June 30, 2018, 06:01:47 AM »

Quote from: mdts on June 30, 2018, 05:41:57 AM

Here is another thread where a DSI employee acknowledges the bug and even recommends staying clear of patch U3-29 https://forum.davesmithinstruments.com/index.php/topic,1959.0.html.
...You could try emailing dsi support and direct them to this thread.

Thank you!
If they recommend staying clear of patch U3-29 I hope they will remove it from the presets when shipping new synths.
I have now written a support ticket where I direct them to this thread as you suggested.

Gerry Havinga · « **Reply #10 on:** June 30, 2018, 06:57:43 AM »

Quote from: osflaa on June 30, 2018, 06:01:47 AM

Quote from: mdts on June 30, 2018, 05:41:57 AM
Here is another thread where a DSI employee acknowledges the bug and even recommends staying clear of patch U3-29 https://forum.davesmithinstruments.com/index.php/topic,1959.0.html.
...You could try emailing dsi support and direct them to this thread.

Thank you!
If they recommend staying clear of patch U3-29 I hope they will remove it from the presets when shipping new synths.
I have now written a support ticket where I direct them to this thread as you suggested.

Yes indeed I can re-produce this behavior at will now, with patch U3-29:

U3-29 --> U3-30 wrong
U3-31 --> U3-30 correct

Interesting.

I've owned the Rev2 exactly one year, I have not had any issues except for the multi mode poly sequencer bug. But I am not really interested in the factory presets, just making my own patches.

I've added the Blood Vain patch sheet. It does sound as if it copies more settings than just the effects. But the effects, when switched off, eliminate the problem when switching to U3-30.

Gomjab · « **Reply #11 on:** June 30, 2018, 07:08:02 AM »

I wonder if this was fixed in 1.1.0.1 then inadvertently added back as a regression in later firmware.

If so it should be an easy fix. If someone is still on 1.1.0.1 they could test. My Rev2 is now at latest and I don’t want to roll back so I’ll avoid those patches and wait until a future fix to update.

Gerry Havinga · « **Reply #12 on:** June 30, 2018, 07:26:26 AM »

Quote from: Gomjab on June 30, 2018, 07:08:02 AM

I wonder if this was fixed 1.1.0.1 then inadvertently added back in as a regression in latter firmware.

I am on 1.1.0.3, you might be correct. I will upgrade some time during this weekend, the encoder upgrade definitely sounds worth it. Got to make a backup first.

I am happy just knowing about this. First time I audited bank F3/U3 ha ha

Thanks all for pointing it out.

osflaa · « **Reply #13 on:** June 30, 2018, 09:15:37 AM »

Quote from: Gerry Havinga on June 30, 2018, 06:57:43 AM

But I am not really interested in the factory presets, just making my own patches.

Thank you! But how can we be sure this issue only applies to one specific factory preset and not patches we make ourselves? I mean, exactly the same parameters are available to us when we make the patches. It would be nice to be able to narrow the issue down as much as possible, do you think it's only the distortion effect causing this?

Gerry Havinga · « **Reply #14 on:** June 30, 2018, 09:32:55 AM »

Quote from: osflaa on June 30, 2018, 09:15:37 AM

Quote from: Gerry Havinga on June 30, 2018, 06:57:43 AM
But I am not really interested in the factory presets, just making my own patches.

Thank you! But how can we be sure this issue only applies to one specific factory preset and not patches we make ourselves? I mean, exactly the same parameters are available to us when we make the patches. It would be nice to be able to narrow the issue down as much as possible, do you think it's only the distortion effect causing this?

That is a good point of course. I have gotten into the habit the last 12 months to always start with an init patch. Seems that has not gotten me into trouble yet

I don't think it is just the effects. The incorrect behavior does disappear when you switch off the effects on patch U3-29 before going up one. But when I listen to the wrong version of U3-30 after gone up from U3-29 it doesn't sound even remotely like the original U3-30. So there is something more going on.

I don't have much time to experiment this weekend. That's is why I posted the patch sheet (see also the pinned forum thread at the top of this forum subject). The patch sheet should document the settings for U3-29. Strangely enough going down from U3-29 to U3-28 does seem to work fine ....... Possibly it is a strange combination quirk between U3-29 and U3-30.

osflaa · « **Reply #15 on:** July 02, 2018, 10:48:05 AM »

Quote from: osflaa on June 30, 2018, 06:01:47 AM

I have now written a support ticket where I direct them to this thread as you suggested.

Here is their response:
"This is a fringe case specific to this patch. It's on our list of things to look further into, but for the time being if you disable the effects on this patch and resave it, the problem will no long occur."

So they say it is impossible to run into this issue with any other patches. Well, then it's not a big deal as long as we know about it.

Robot Heart · « **Reply #16 on:** July 02, 2018, 10:58:56 AM »

Quote from: mdts on June 30, 2018, 05:28:44 AM

If it sounds like heavily distorted it's a bug that DSI has completely ignored and/or failed to fix so far.

I've had it since I bought the synth last year and frankly I've given up hope it will ever be fixed.

Have you reported this to us directly, through our dedicated support channel? Please include concise steps on how to reproduce this so we can officially document it: support (at) davesmithinstruments.com

Thanks!

AlainHubert · « **Reply #17 on:** July 02, 2018, 07:36:49 PM »

It's definitely a bug where the distortion effect isn't properly shut down when switching patches. The same thing happens when going from preset patches F1 P67 "Can't Kill Me" to F1 P68 "Gated community", the high level distortion effect is still there when it should not. Coming from any other patch to F1 P68 results in the normal sound of that patch, without the heavy distortion effect.

Gerry Havinga · « **Reply #18 on:** July 02, 2018, 10:46:31 PM »

Did anybody report this bug with instructions on how to reproduce to DSI support? If not I would be willing to write up a quick mail tonight (CEST).

(Edit: past tense in report wrong ;-)

osflaa · « **Reply #19 on:** July 02, 2018, 11:23:03 PM »

Quote from: Gerry Havinga on July 02, 2018, 10:46:31 PM

Did anybody report this bug with instructions on how to reproduce to DSI support? If not I would be willing to write up a quick mail tonight (CEST).

(Edit: past tense in report wrong ;-)

Yes, I did with link to this forum thread. Robot Heart can not have read the whole thread. Anyway, they came to the wrong conclusion about the issue and I will follow up with info on the F1P67 to F1P68 issue.