October 14, 2008 at 21:56 #861
I’m seeing (or rather hearing) a rather noticeable drift in the Octopus’s midi clock signal. I’ve only noticed it after updating to OS 1.60. Is it possible to downgrade to an older OS so I can do some testing and rule out the new software?
To test the midi clock I first created a 44100hz audio file with a click every 22050 samples (1 click a beat at 120 bpm). I then recorded my machinedrum slaved to the Octopus also playing a click every beat, when compared to the audio file I created the Octopus’s clock drifted a bit slower and after about 1 minute was 159ms behind the click track.
Using the machinedrum as master, or slaved to Ableton Live (via the midi out of a Motu 828MKII) played almost perfectly in time to the click track (both ways deviating by about 8 samples or so, which is less than a milisecond), so it looks like it’s the Octopus that’s drifting. I’d like to check if the older OS gives a more stable midi clock signal but I don’t think it’s linked on the website any more.
Has anybody else noticed any midi clock drift?
Post edited by: admin, at: 2009/03/14 04:19October 15, 2008 at 01:07 #1935
I should probably shut up as I actually don’t know for sure – but I have a vague memory of me reading a discussion somewhere ages ago (might have been on Analog Industries) about how the clock in Machinedrum and Monomachine were drifting. But I really can’t remember for sure.October 15, 2008 at 02:28 #1937
Hartley, if you want to test it in your setup, the way to go back to a 1.4 version is via the USB upload. The document on how to do it is linked in the OS software section of the service area of the web site. I hope it is not too technical to give it a try. The OS file you should use is linked here: http://www.genoqs.com/downloads/OCT_OS_v14308.elf
Let us know how it goes. Thanks and cheers, Gabriel.October 17, 2008 at 12:56 #1938
Just to let everyone know – Hartley and I are working on this offline, and have already made some good progress so far, and are still exploring some more ideas.
If this is an issue that causes major headache, contact me directly. Thanks, GabrielOctober 18, 2008 at 18:51 #1946
i found the drift was there in 1.42 but i have found very little gear that doesnt drift midi clock wise…..some more than others , the main issue for me is that i record into cubase sx alot and on ling jams the midi data is well out of sync to the cubase clock and needs quantize to get everything back in – and sometimes when the octopus is grooving this kills the octopuses swing.Some drift is expected maybe but less is betterJune 4, 2009 at 13:28 #1949
thstd the big issue , over say a 10 minute track it drifts so much that when recording to cubase etc it requires qunatisation and that then loses the swing , you have to swing quantize in cubase.Its common on laods of midi gear i found , it all seems to lose time over time.Maybe cubase doesnt help !October 6, 2009 at 07:32 #2360
I think for sure Cubase doesn’t help.If you notice on 1.60 incoming midi clock
….(cubase does not accept midi clock as a stable enough input signal)….
makes the outer circle BPM lights flicker 2 or 3 bpm either side of the set
tempo being sent from Cubase,through my edirol FA101 anyway.
You must be logged in to reply to this topic.