Arpeggio playing faster than project (timing issue)

Hello.

I'm not sure how many other people are having the same problem with Arpeggios, but I have seen a few posts over the internet about timing issues. This one seems different than most and has been happening since installation of OS4 on the Virus.

I have noticed that the Virus TI is playing arpeggios at a slightly faster BPM than in my Cubase project. My project is running at 129BPM and the arpeggio sounds as though it is running at 130BPM, yet the MIDI signal being sent through USB to the Virus seems to be keeping it roughly in line as the project carries on playing. It's almost like listening to two tracks playing out of time but being forced back in time every so often.

For each MIDI note that I'm using to trigger the arpeggio (notes are 1 bar in length), the Virus will play every single arpeggio note perfectly with an additional note on the end due to it playing the arpeggio faster than the project BPM.

Kickdrum timing: |___|___|___|___|___|___|___|
Arpeggio timing: |___|___|__.|___|___|__.|___|.| (a 9th note plays for a short duration before the next bar comes in.)
With timing adj*: |___|___|__.|___|___|__.|___| (9th note doesn't play when Cubase's note length adjustment is used at -12 ticks.)

*I have partially resolved this 'note crossing over' issue by using the global note length adjustment setting in Cubase to shorten all notes by 12 ticks, but it is still playing the arpeggio faster, making it go in and out of time as the project attempts to drag it back in time with itself.

It's almost as if the arpeggio is being triggered by the MIDI notes from my project, yet running off a slightly faster internal clock until it gets another MIDI note trigger to push it back in time with the project again.

I am running the Virus TI through the USB cable (3 audio outputs / No input) and am also using the Virus TI plugin in Cubase 4 (the non-VST3 version of it.)

Hope I've provided enough information to explain the problem I'm having.
Regards, Jim.
-----------------
Hardware/Software info:

Motherboard: Gigabyte GA-X58A-UD7
CPU: Intel Core i7 980X Extreme Edition (6-core)
Memory: 12GB Corsair XMS3 1333MHz DDR3
Audio Interface: MOTU 24i/o (with PCI-e 424 card)

Operating System: Windows XP Professional (32-bit) ( Yet to test for the bug on Windows 7 64-bit )
Sequencer: Cubase 4 (32-bit) (build 4.5.1)

Zitat von Marc

hi,

can you send a project illustrating the problem to access support so we can compare results?

thanks, marc

I will send the project file as soon as possible, but here is an audio sample first to demonstrate the issue.

http://www.audiofascist.co.uk/…psync__bugged_vs_orig.mp3

The first half of the mp3 is a recent export of only the drums and the synth arpeggio (using OS 4.1.0.16). The synth sounds like it's having difficulty staying in time and there's an extra note at the end of each bar that overlaps onto the next bars.
The second half of the mp3 is an older export using one of the first releases of OS4 (version 4.0.5.01), but you can hear it is in perfect time with the track.

{EDIT: I have just rolled back to OS 4.0.5.01 and the problem has disappeared. This rules out anything in my project aswell as the recent computer upgrade I forgot to mention. I'm guessing some changes were made to the arpeggio section between versions 4.0.5.01 and 4.1.0.16 that need to be addressed.}

Regards, Jim.

Zitat von Marc

there were no changes and without a project we can't help you or fix anything. in fact the arp works fine here, at least i got it running for half an hour now and it is not out of sync.

best, marc

Hi Marc. Thanks for the response.

I have sent a link to the project download to Access support along with the audio sample and a link to this forum thread.

The only thing I can think of is that some sort of change in OS 4.1.0.16 will have altered the way the Virus functions that none of the arpeggios work correctly with my computer setup. It's probably a local issue if you can't recreate the conditions there.

I tested as many arpeggio patches as possible with OS 4.1.0.16 and they all suffered from the timing issue. It's only very slight, but is very noticable to my ears. After rolling back to OS 4.0.5.01, all of the patches I tested worked perfectly in time with no glitches or clicks.

The only other thing I can possibly think of is that the installation may have gone wrong somewhere and that reinstalling OS 4.1.0.16 again may fix the issue. I will try reinstalling it one more time and get back to you within the hour. An 'EDIT' will be put on this post with the progress update.

Thanks, Jim.

{EDIT 1: I have reinstalled OS 4.1.0.16 beta and the timing issue has returned on all arpeggio patches. It sounds like the arpeggio is time-shifted backwards by a few milliseconds compared to everything else. It jumps in early and misses off the first note of the arpeggio then continues to play time-shifted back a few ms. I am currently reinstalling OS 4.0.5.01 and doing another mixdown of the exact same audio section as a demonstration}

Zitat von Marc

there were no changes and without a project we can't help you or fix anything. in fact the arp works fine here, at least i got it running for half an hour now and it is not out of sync.

best, marc

Hi again.

I have uploaded a rar containing the following:
• Simple Cubase 4 project with an arpeggio playing in time with the Cubase click.
• Two audio samples demonstrating the exact same Cubase project when using OS 4.0.5.01 and OS 4.1.0.16 respectively.
• Three screenshots. The first two pictures are the spectral view on Adobe Audition, showing how with OS 4.1.0.16, the arpeggio notes are pre-delayed vs the Cubase click. The third screenshot is the waveform view of one of my projects, showing the measured 9ms pre-delay of the arpeggio notes.

PROJECT DOWNLOAD

I hope this information will show you what you may not have been able to hear in your own tests. 9 milliseconds difference isn't very audible unless you've already listened to a project hundreds of times, only to have it suddenly change after an OS update.

Thanks, Jim.