Access Music User Forum

Quoted from "creap"

But without live mode, I get non synced arps, clicks and pops and a whole host of problems. I simply CANNOT record the virus in pro tools LE without live mode. Honestly the most stable solution for me is to just use midi cables and audio out. I get perfect sync and no audio distortion at all. I hate this approach because I like to have patches reload with a project and I like the easy to use interface. Obviously the virus uses some different clock features when running on midi vs running off usb.

Every day I question why I went with pro tools and spent all this money on a 003 factory setup when I could have just got a nice firewire interface and some universal controller and used whatever DAW I wanted. My frustration with PT aside, I just want to make music, and the virus (among other things) fights me every step of the way.

i guess that happens with the tutorial song as well? why don't you send me a project, i have an mbox micro in the office and i could try to get a close to your setup as possible and let you know about the outcome. one real question though. in protools HD you can tell RTAS to use less CPU cores than the maximum. if you reduce the cores to, say 2 for instance, does that make a difference?

also here's some background for you guys listening: the live mode we really did for hosts which do not support some specifications or feature implementations total integration needs to run smooth. at the point in time we did this, ableton live was one of the most prominent examples because otherwise playing the virus with virus control would simply result in notes during recording would be really late (playback was always fine). the live mode has one big disadvantage though: in difference to its being disabled, the notes would not be sample accurate and possibly even slightly less accurate than normal midi. it is a compromise with the benefit that you don't deal with latency during recording your performance. it was never intended to cure anything else.
click and pops usually happen for four reasons:
- the host is totally overloaded and cannot cope anymore
- the usb transmission doesn't work as intended and data packages get "lost" or really late in time (this is where distortion can come in)
- the virus' mixer overloads (red in digital means damaged data, not added saturation like in the analogue world)
- software setup problems
by the way, i'm happy that this conversation happens right now because most of the time i see those threads, it appears that they are there for venting off instead of trying to solve it.
have a good weekend

marc

hmm I hope I'm posting this in the right topic , but I'm having huge problems with this live mode feature. I told it in other topics, but if I don't enable it: the arps are completely desynced, play unstable and play in front of all other cubase events. If I enable live mode it's somehow more stable but it can sometimes happen that the first arp note plays slightly in front of the first cubase metronome click.

So enabling live mode for me is actually the way to make the arpegio's tight. Unless I need to cope with another problem in my DAW. The particular project i'm working on is so heavy that I have to put my soundcard latency to 232ms, just to avoid all cpu overload pops and sound cuttofs on some other plugins.

The virus doesn't seem to handle this high latency very well as the arp will play as bad without live mode than it would play if live mode is enabled. Even with the live button off, at 23ms the virus sounds very unstable and messy. I tried to set the virus driver's setting to both "normal" , changed back to hammerfall DSP, but still the virus continues to play the arps unstable. The problem is I can't freeze the other plugins because it' doesn"t like "non realtime" freezes and will result in cut off played sounds and pops.

The only thing I can think of now is to first bounce the heavy sample plugins like play to disk, freeze and unloadd them sepparately and then then trying to mix.

Quoted from "CosmicD"

The particular project i'm working on is so heavy that I have to put my soundcard latency to 232ms

how many samples does this compare to? sample buffers over 1024 will not work for accurate timing.

marc

Yeah it's something like that. 23ms is exactly 1024 samples. The big problem is that even when I set the latency to 6 or 12ms and try to bounce the virus in solo (to get it out of the equasion) i still have pops and crackles because of cpu overloads. Because i've so much compression and verb plugins loaded and not to mention the eastwest play plugins that are loaded : it's not possible for me to export this song anyway. The extra catch 22 trap is that when I freeze the PLAY instruments, their release trails start cutting off cause they can't handle non real time freezes.

Only thing I can still try is try to start by bouncing play plugins solo track per track, then remove them so they are out of the way. Then bounce the virus plugin and remove it. Then freeze all NI plugins. I think at that point the system should be stable enough to export the song without overloads, unsynced arpeggios, cracks & pops

The problem with that is that it causes so much workload and clutter. It's nicer to just be able to export a song in one go cause I know myself. if I hear one little error i'd have to unfreeze everything. So I end up with 4 versions of the same projects. unfrozen, partially frozen, frozen final ......

Quoted creap:
"But without live mode, I get non synced arps, clicks and pops and a whole host of problems."
- It seems this delay you talking about is in every case the same to me, the systems are not able to sync in a specific way. I still try to work on it.


Quoted Marc:
"in protools HD you can tell RTAS to use less CPU cores than the maximum. if you reduce the cores to, say 2 for instance, does that make a difference?"
- With this feature in HD & LE you can choose how many cores of your CPU will handle the processing power for RTAS Plugins. For example: 2 cores will primary receive RTAS informations, and the 2 other cores are assigned to all other applications. (OS, PT, and so on...)

I was tried to setup virus to 1 core, because I thought the sample/phase accuration could may be instable in LE. multithreading works 100% with virus and RTAS.
Its more a hard-RTAS bug in LE and Windows. Never heared a problem from MacOS in combination with the 003 Rack.


Marc, if you try to solve the problem with your M-Box, let us know whats going on.
The M-Box could work better, but remember it's not the same hardware or may not the same RTAS source like the 003 Rack.
My System: XP, 003 Rack+ LE 8.0.3 - TIOS 4+


Quoted Marc:

"click and pops usually happen for four reasons:
the host is totally overloaded and cannot cope anymore"
- PT LE is not overloaded, RTAS CPU load is at 1%

"the usb transmission doesn't work as intended and data packages get "lost" or really late in time (this is where distortion can come in)"
- that sounds more like the sync/latency problem

"the virus' mixer overloads (red in digital means damaged data, not added saturation like in the analogue world)"
- the signal from virus USB to PT is not overloaded

"software setup problems"
- PT is up to date and all RTAS plugins could be compensated manually


Thanx for tuning in.