sporadic corruption of midi data by scope

An area for people to discuss Scope related problems, issues, etc.

Moderators: valis, garyb

Post Reply
nebelfuerst
Posts: 479
Joined: Tue Jun 23, 2009 10:55 am

sporadic corruption of midi data by scope

Post by nebelfuerst »

I experienced some hanging notes since years. Now I took the time to trace it down.
My scope software is V7. I used IO-Linux and MidiOX to record the midistreams.

Result:
It is not my midisequencer, it is not my timepiece, it is the scope, which causes corruption.

I recorded the midi-through of scope (14+6 DSP) and found no corruption.
Then I made a direct link from midiinput to midioutput within scope and recorded this, too.

The value of the note to play sometimes changes to 127. This can be found on scope midi-out aswell on the sequencer-output, which I recorded with midiox.

The project used 90% of the DSPs, but this shouldn't be an explanation for data corruption.
Is there a known issue ?
\\\ *** l 0 v e | X I T E *** ///
User avatar
Marco
Posts: 1260
Joined: Thu Oct 10, 2013 4:27 pm
Location: Ödenwald

Re: sporadic corruption of midi data by scope

Post by Marco »

I do not use the Midi IO of scope, because I always have hangig notes, this disqualifies scope as a midi processor. But Midi from any sequencer, uff what a lucky man I am works. So a bug castrates scopes Midi Hardware IO s. Only an Update could solve this Problem.
:wink: out and about for music production. Are you still configguring your Studio :lol: music first!
User avatar
valis
Posts: 7306
Joined: Sun Sep 23, 2001 4:00 pm
Location: West Coast USA
Contact:

Re: sporadic corruption of midi data by scope

Post by valis »

I have clean midi from scope after filtering out active sensing and other extraneous Sussex from external devices.
nebelfuerst
Posts: 479
Joined: Tue Jun 23, 2009 10:55 am

Re: sporadic corruption of midi data by scope

Post by nebelfuerst »

I don't remember such problems with scope in the old days, when using 3.1 or 4.x. But in the old days, my PC hardware was different and WinXP was alive.
@Valid: Do you filter inside scope ( .. like I do ) or before the Midi goes into scope ?

I'll do further tests, to get this fixed. There is nothing more annoying than a hanging note.
\\\ *** l 0 v e | X I T E *** ///
dawman
Posts: 14368
Joined: Sun Jul 24, 2005 4:00 pm
Location: PROJECT WINDOW

Re: sporadic corruption of midi data by scope

Post by dawman »

Id check the sustain CC# 64 on a MIDI Monitor by having every application using its own Monitor.
I do this by having Scopes next to Bidules and as soon as you hear it stop and check.
The main reason I use Scope is for its rock solid MIDI and Mixers precise automation.
User avatar
valis
Posts: 7306
Joined: Sun Sep 23, 2001 4:00 pm
Location: West Coast USA
Contact:

Re: sporadic corruption of midi data by scope

Post by valis »

I filter outside of Scope. MOTU Micro-Express & MidiTimePiece AV both have the ability to filter any messages on any channel I choose, and I have one of the Blokas usb hubs on the way which will do so as well.

What DAWman suggests is on the correct path, you need to walk your chain of devices and see which one is sending spurious information. Filtering alone may solve it, but you won't know the source of the issue for future enlightenment when it occurs again.
David
Posts: 733
Joined: Wed Jan 09, 2002 4:00 pm
Location: Modular IV

Re: sporadic corruption of midi data by scope

Post by David »

Jopo had this to say about midi
JoPo wrote: Sat May 21, 2016 4:56 pm I just wanted to say how it's good and pleasant to not have any midi jitter ! I can control modular exactly as I want by drawing midi CC in the daw, it plays back always the same ! I can go up and down with control inside a single 1/16 note ! And that is thank to usamo and silentway : 2 tools I'm stupid to not have purchase before ! Silentway for modular or Scope devices with modulation audio input, like Pro-wave, for intance and usamo when accurate midi automation is needed on any Scope device.
Midi jitter was breaking my balls since so many time ! That's simply a-we-so-me !!

That's all I wanted to express... When it's as pleasant, it's gOOd to say it to the world ! :D

Le PA-NARD ! (means the same but in good french).
In this topic - https://forums.scopeusers.com/viewtopic ... mo#p322561

Don't know if Usamo and SilentWay will help but they're interesting tools
dawman
Posts: 14368
Joined: Sun Jul 24, 2005 4:00 pm
Location: PROJECT WINDOW

Re: sporadic corruption of midi data by scope

Post by dawman »

Silentwat basically gives PBend resolution on controle previously using 0-127.
Half of the time MIDI Hangs from from latching onto CC#64.
nebelfuerst
Posts: 479
Joined: Tue Jun 23, 2009 10:55 am

Re: sporadic corruption of midi data by scope

Post by nebelfuerst »

I just did more measurements.
I replay one normal song several times, while recording the MIDI-thru and the Midi-out of scope.
Inside scope, there is a direct connection between scope-midi-a-source and scope-midi-a-dest, no more wires attached to these two.
So my expectation is a complete match.

I'm recording with kmidimon using linux on two separate PCs. ( These PCs don't make any errors, I tested them very well.)


A sysex message sporadically was 8 of 10 bytes in the midi-out. One program-change-message was missing, too.
Then I get no errors for 20min.

I really need to know, if this behaviour is scopeV7 related or I might have some hardware problems.
\\\ *** l 0 v e | X I T E *** ///
User avatar
valis
Posts: 7306
Joined: Sun Sep 23, 2001 4:00 pm
Location: West Coast USA
Contact:

Re: sporadic corruption of midi data by scope

Post by valis »

What was the CPU load while this was occurring?
nebelfuerst
Posts: 479
Joined: Tue Jun 23, 2009 10:55 am

Re: sporadic corruption of midi data by scope

Post by nebelfuerst »

the host cpu ( intel i5) was at 0%. It was just scope7 running.
The dsp load was at 80%.
\\\ *** l 0 v e | X I T E *** ///
User avatar
valis
Posts: 7306
Joined: Sun Sep 23, 2001 4:00 pm
Location: West Coast USA
Contact:

Re: sporadic corruption of midi data by scope

Post by valis »

Do you have all sleep states and TDP controls disabled on your CPU?

It might seem surprising, but a 50% CPU load can improve your audio performance for realtime devices like Scope over say 10% cpu usage.

So an artificial load created in your DAW/host (a few synths playing to an output that isn't connected to anything) is a great way to test this.
nebelfuerst
Posts: 479
Joined: Tue Jun 23, 2009 10:55 am

Re: sporadic corruption of midi data by scope

Post by nebelfuerst »

Time for an update:
I programmed some tools, which send a lots of notes into the system of 14 +6 dsp. In the project, I just connected the Midi-source to Midi-dest für the scope and the pulsar2-card.
Results so far:

There is absolute no error/missmatch if I compare the Midi-in to Midi-thru, so my interfaces/software and even the optocoupler in the cards seem to be fine.

When the project runs as slave to an A16ultra, the scopecard shows missmatches, but also does the pulsar2.
When switching to master, I still found missmatches on scope, but none on pulsar2.

Now I reduced the load in the project by deleting some synths. Suddenly, the errors on both cards disappeared.

So for the moment I suspect this to be a software issue of V7.
As all errors a statistical effect, I'll continue measuring ;-)
\\\ *** l 0 v e | X I T E *** ///
User avatar
Bud Weiser
Posts: 2684
Joined: Tue Sep 14, 2010 5:29 am
Location: nowhere land

Re: sporadic corruption of midi data by scope

Post by Bud Weiser »

nebelfuerst wrote: Thu Oct 10, 2019 10:18 am Time for an update:
I programmed some tools, which send a lots of notes into the system of 14 +6 dsp. In the project, I just connected the Midi-source to Midi-dest für the scope and the pulsar2-card.
Results so far:

There is absolute no error/missmatch if I compare the Midi-in to Midi-thru, so my interfaces/software and even the optocoupler in the cards seem to be fine.

When the project runs as slave to an A16ultra, the scopecard shows missmatches, but also does the pulsar2.
When switching to master, I still found missmatches on scope, but none on pulsar2.

Now I reduced the load in the project by deleting some synths. Suddenly, the errors on both cards disappeared.

So for the moment I suspect this to be a software issue of V7.
As all errors a statistical effect, I'll continue measuring ;-)
MIDI issues are crap if true.
Your reports are welcome !
IMO they should go into the sonic core bug reporting system - user reporting area https://forums.scopeusers.com/viewtopic.php?f=1&t=35261.

I really wonder why there aren´t more bug reports to find ...
How shall Holger be aware of all the issues when all complain across the forum´s threads, but rarely in THAT thread ?
We cannot expect he´s reading ALL the posts here.

I´d like to know if similar or the same issues also happen w/ SCOPEv7 and XITE-1 and I really wonder how DAWMAN is able performing live w/ his complex and ambitious MIDI rig based on SCOPE v7 x64 and XITE-1, when it´s true SCOPEv7 has so many MIDI bugs.

any idea ?

:)

Bud
nebelfuerst
Posts: 479
Joined: Tue Jun 23, 2009 10:55 am

Re: sporadic corruption of midi data by scope

Post by nebelfuerst »

I'll contact S|C with my findings, when I finished my testing.
The loss of a Midi-message is a rare event, maybe 1/hour, when the DSPs are loaded quite high.
This doesn't keep me from using scope.

The reason why I'm tracking this down is, that I'm using my self-programmed sequencer software. Often the bugs were on my side :)
Also one of my cheap USB-Midi interfaces was found to have buggy firmware. ( now I only use edirol and behringer )
So I decided to write some tools to do integrity testing on Midi.
As I also own a Xite, I'll try that the next days.
I enjoy making music with scope, but I also enjoy messing around with the technical side of electronic music equipment :)
\\\ *** l 0 v e | X I T E *** ///
User avatar
Bud Weiser
Posts: 2684
Joined: Tue Sep 14, 2010 5:29 am
Location: nowhere land

Re: sporadic corruption of midi data by scope

Post by Bud Weiser »

nebelfuerst wrote: Thu Oct 10, 2019 12:27 pm I'll contact S|C with my findings, when I finished my testing.
perfect !
nebelfuerst wrote: Thu Oct 10, 2019 12:27 pm The loss of a Midi-message is a rare event, maybe 1/hour, when the DSPs are loaded quite high.
This doesn't keep me from using scope.
Well, I recognized occasionally hanging/stuck notes also in SCOPE 5.1 w/ XITE-1.
But I had a Sycologic M16 midi matrix switcher in between,- so I was never sure if it is SCOPE or not.
That happened in my old studio and I´m unable reproducing actually.
OTOH, I have several Sycologic boxes and used ´em for touring over decades.
No trouble at all,- and Rick Wakeman still uses these today for touring.
Michael Jackson, Tony Banks and Peter Gabriel used ´em too ...
These were extremely rugged devices, so I doubt they are the culprit.
nebelfuerst wrote: Thu Oct 10, 2019 12:27 pm The reason why I'm tracking this down is, that I'm using my self-programmed sequencer software. Often the bugs were on my side :)
Also one of my cheap USB-Midi interfaces was found to have buggy firmware. ( now I only use edirol and behringer )
I always used the best MIDI interfaces I was able to find, which means nothing though ...
Even the best available in your area might be crap.
I fear that rules most for the multi-port MIDI interfaces offering 8+ physical MIDI I/O ports, especially when connected via USB.
For WinXP I used MusicQuest/Opcode 8PortSE w/ Earthvegaconnections MIDI driver which was/is excellent for parallel port and Win XP.
But driver developement ended and there´s a no-go for Win7, 8 or 10 x64.

But 8x8 or larger is what I need !

Next purchase is iConnectivity mio10, then we´ll see ...
nebelfuerst wrote: Thu Oct 10, 2019 12:27 pm So I decided to write some tools to do integrity testing on Midi.
As I also own a Xite, I'll try that the next days.
cool ! 8)
nebelfuerst wrote: Thu Oct 10, 2019 12:27 pm I enjoy making music with scope, but I also enjoy messing around with the technical side of electronic music equipment :)
Me too !

:)

Bud
User avatar
garyb
Moderator
Posts: 23248
Joined: Sun Apr 15, 2001 4:00 pm
Location: ghetto by the sea

Re: sporadic corruption of midi data by scope

Post by garyb »

hanging notes are usually cause by active sensing or midi clock.
filtering active sensing cures most ills...

the bug forum isn't that helpful. send an email to SC.
Post Reply