Highlighted
Contributor - Level 2

My current Poco Virus weirdness

For what it's worth...

I've set up a very simple Cubase 4.5.1 project:

145BPM, 4/4 kick played on 1 virus instance, 16th bass played on another, snare played on a Poco 01 (to use all Poco plug-ins).

If I save and load the project, the Virus instances play back with massive latency. i.e. take a long time to actually start playing when I hit play and stop when I hit stop.

Now here's the interesting bit:

If I change the buffer depth on my Hammerfall (to anything), they both fall back into sync.

If I then close that project and load it again, same thing. Huge latency for the Viri, until I change the Hammerfall's buffer size.

I'll attempt to video it and post here.

I'll also try similar on Ableton Live.
Tags (1)
Purusha Contributor - Level 2 2008-09-11

2008-09-11

My current Poco Virus weirdness

For what it's worth...

I've set up a very simple Cubase 4.5.1 project:

145BPM, 4/4 kick played on 1 virus instance, 16th bass played on another, snare played on a Poco 01 (to use all Poco plug-ins).

If I save and load the project, the Virus instances play back with massive latency. i.e. take a long time to actually start playing when I hit play and stop when I hit stop.

Now here's the interesting bit:

If I change the buffer depth on my Hammerfall (to anything), they both fall back into sync.

If I then close that project and load it again, same thing. Huge latency for the Viri, until I change the Hammerfall's buffer size.

I'll attempt to video it and post here.

I'll also try similar on Ableton Live.

  • 0 Kudos
  • 31 Replies
  • Reply
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

Know what? Ableton 7.0.10 seems to be OK!?

Now - this is starting to look possibly more like a Cubase Bug.

It's almost like Cubase isn't informing its plug-ins correctly of the audio buffer size until the buffer size actually changes.

Is anyone else able to see something similar?

Edit: tried to video it, but it's too hard to see and hear what's happening I'm afraid.

Can anyone who's having timing problems give this a go? It might be a work-around!
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

Hi,

Does it in Logic 8.0.2 on my mac pro, Leopard 10.5.4, so probably not Cubase at all. I use Cubase 4.5.1 on PC as well, but the virus runs on my mac. If I set the buffer range to medium, I get an incredible latency with the poco virus plugin, however, if I set it to small, it works just fine. Interestingly enough, I see you're on RME. I use the FF800 though, wether or not thats got anything to do with it I don't know, although I doubt it. Virus pleasure eh? Access have done wonders for my TI recently though, which makes me wonder why the poco virus is so shoddy.... Cheers.
Mac Pro Late 2013, 64 GB RAM, Mac OS Sierra, Logic X, UAD Apollo Duo&Quad Thunderbolt 2, Powercore FW.
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

Thanks for the response.

Can we be sure it's the same problem in Logic?

Have you tried changing the buffer sizes in the same way (presuming you can do that on Mac)?

The effect I'm seeing appears to be very controllable in the way I describe.

It might also account for some of the "vanishing problem" tests I've run where it suddenly seemed to fix itself. I might very well have changed the audio buffer sizes to see if different sizes worked any different.

Think I might be on to something here...

Edit: I've now got 8 instances of Virus running (2 Powercore FW units), stressing them with looped 16th notes. All now staying in sync if I change the audio buffer as described.

Wonder if this will affect the arp too?

Edit: Arp seems OK either way. This seems to be purely latency related.
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

Ive made deep tests in various host (live 6, 7.01, and cubase 3 and 4) and 3 different computers setup (MPB, alienware, selfmadePC) and I can affirm the virus has timing issues , it's not sample accurate, checked with an oscilloscope, there are random timing, and overal host sync can be lost in a short time.

In live, if I got the virus playing at more than 5 min later the start of my liveset, the arp is nearlly always out of sync, have to press stop play again (what a shame during live session!)

changing latency setting seems to rezet timing information the virus get and then it goes sync..... but for a short while, and if you look in a precise way (ie oscilloscope to check phase relating to time), it's clearly not precise, whereas other synth, ie Massive or operator are strong sync locked, so I think it really comes from the virus, and not from the host...
MBP 2.3 W7 touch--TC FW X8--TC fried conga X4--TC konnekt24D korg Zero 8, abl live, sensomusic usine.
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

Hmmm. I've not lost sync since I re-installed my Via based 1394 card some months ago. Before then, I was losing sync all the time.

This is sounding pretty timing accurate so far if I use my work-around.
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

Hi Purusha,

This is how it is on the mac in logic. I can set my I/O buffer size in samples to whatever I like, and obviously other than the naturally incurred latency going up, its good. (32 - 1024 samples). If I change the process buffer range, an additional setting in logic, the virus poco goes to s***. It only works on the small setting. Don't ask me why, cause I don't know! Maybe someone else can tell me why. All I know is, there's an issue. Cheers.
Mac Pro Late 2013, 64 GB RAM, Mac OS Sierra, Logic X, UAD Apollo Duo&Quad Thunderbolt 2, Powercore FW.
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

I think the likelihood (in my case) is that Steinberg and Access's interpretation of the VST spec is different.

Obviously other plug-ins (Poco01 included) seem to work OK, but then Ableton seems to be OK.

The other thing I'll try (maybe this evening) is: one of my other sound outputs. ASIO4ALL through the on-board sound, or the Virus TI as sound card.

My testing from last night seems to show that I've actually now got things stable if I do the buffer size change after loading a Poco Virus instance (including newly added instances).
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

I don't know how the VST API works, but there's a problem I've seen quite often in my day-job that could explain this:

If the plug-in registers a call-back with the host for getting sample depth info, maybe Poco Virus is registering it later than most others?

In which case, perhaps Ableton actually makes a callback on registration with relevant info, but Cubase doesn't (unless and until something changes).

This scenario fits the problem quite well.
Highlighted
Contributor - Level 2

Re: My current Poco Virus weirdness

Ive noticed that I cannot run a buffer higher than 1024 on Mac OSX 10.4.11
the previous version would run at higher buffers if needed know it reports
a device error.
OSX 10.6.4 / Mac Pro Dual Quad 2.66 /6 Gig Ram Steinberg MR816X Tc Powercore PCIe / UAD2 / SSL Duende PCIe Cubase 4.5.2/Cubase 5.5