Problems with polyphony and latency

[Storm Music]

I'm feeling quite some latency on my TI2 polar, when I - for instance -play 4 tones with a trancy hook sound with voices set to 4.
You know, it's just a little, but definitely some - and too much...
And if I want to stack different sounds for really fat leads in multi-mode, it's really bad...
The feeling is... like I'm unable to play all the tones at exactly the same time, but as a trained pianist I know I'm playing it at exactly the same time...

I thought the TI2 was supposed to have lots of DSP power...

Anyone knows if this can be fixed or anything?

BTW - we're talking stand-alone mode... No USB, only audio cables to my mixer (rehearsing for a spring tour starting in one week, so I have to get the problems solved pretty soon)...

[ShortBus]

Are we talking Ti to mixer to daw and then to speakers because that would make sense but ive never had latency from the Ti itself. If so I would have that checked out. Maybe check the amp and filter attack on the patches.

[Storm Music]

No daw involved... (well I'm monitoring through my RME soundcard right now, but still directly in an analogue mixer the same thing happens)...
Already checked the amp and filter attacks...

[Storm Music]

No daw involved... (well I'm monitoring through my RME soundcard right now, but still directly in an analogue mixer the same thing happens)...
Already checked the amp and filter attacks...

[Alakhai]

that's strange...I understand you could have some polyphony problems...but it would be like some cuts, or notes ending abruptly, not latency

[Storm Music]

I just came across another thing:
I was tweaking ROM-C 3 "BadCntaxHS" - and changed it from mono to poly, and suddenly it started making a whole lot of noise, when I play more than one note at a time...
Check it out:
http://www.stormclements.dk/mp3/Virus%20Noise.mp3

It's not directly related to the other problem I have experienced, but maybe they are both symptoms of a sick DSP unit or something...

And it's not the converters of my RME soundcard that are being overdriven - it was recorded at around -30db... And it sounds the same in the headphones...

This really makes nervous, 'cause something similar happened to my old virus indigo just before it crashed and totally died.... :-(

If somebody would please try to do the same thing with the same sound to see if it's not just my TI2, I would be really grateful!!

[feedingear]

that's strange...I understand you could have some polyphony problems...but it would be like some cuts, or notes ending abruptly, not latency


Tried doing some soft resets (arp edit held down) of the TI and seeing if that helps? As Alakhai said, it's more likely youd hear notes being stolen, dropping out then a latency issue. Try turning off various parts of the TI (fx etc) and see if you can get the latency reduced.

[ShortBus]

Try hooking up speakers or headphones directly to the headphone or analogouts of the TI.

[Storm Music]

Did that already.... Same thing happens...
Also tried resetting... No change...

[paulkeeley]

i may be interpreting this incorrectly, but if you're talking about notes not playing at exactly the same time as you are inputting them when running your Virus in standalone mode (no VC), i have the exact same problem. it seems that the DSP can't handle playing back voices in time when there is significant processor load. in VC mode, playback is better because it has time to buffer (i think). in standalone, it has to do it all in realtime, and shows its true (lack of) power.

as a test, try creating a very basic patch - just 2 standard sawtooth oscillators, no effects, no fancy filters, and play with the amp env release time. should be spot on. then, try changing one of the oscillators to any other oscillator type and do the same. now there should be some noticeable delay. of course, stacking voices will further throw things out of whack.

i've talked to support about this a number of times and it is apparently a known issue, but there is no fix forthcoming. i guess they're focusing all their effort on the VC integration.

so, if you're not using your unit in VC mode, probably best to stick to using it for pads. or keep your patches ridiculously simple.

hope that helps.
cheers, paul.

[Storm Music]

Thanks, Paul.

That's exactly the same thing I'm experiencing...

It really sucks, since I primarily use it for live concerts...

Did not have any problems of that kind with my old indigo... :-(

Pretty lame that this problem has not been adressed by access...

[marc]

Thanks, Paul.

That's exactly the same thing I'm experiencing...

It really sucks, since I primarily use it for live concerts...

Did not have any problems of that kind with my old indigo... :-(

Pretty lame that this problem has not been adressed by access...

have you been in touch with access support?

best, marc

[Storm Music]

...
i've talked to support about this a number of times and it is apparently a known issue, but there is no fix forthcoming. i guess they're focusing all their effort on the VC integration...

No, I haven't talked to support about this exact matter, but apparently Paul has... And with no luck...

I did create a ticket regarding the polyphony problem, but they told me to try less dsp power demanding waveforms and less fx and less release and stuff like that.
But the thing is, I did not buy the TI2 to downgrade the power of my sounds from my indigo. On the contrary I was expecting it to kick ass on my indigo - which is certainly not the case...
Also the converters in my old indigo seems more clear and distinct... IMO

[paulkeeley]

But the thing is, I did not buy the TI2 to downgrade the power of my sounds from my indigoit's funny you mention that - i've been toying with the idea recently about picking up a used B (or C), just to have a solid Virus in the studio. but then i wouldn't have all the wonderful oscillator types i've been spoiled with using the TI. it's just such a shame so little attention has been given to the standalone side of things.

btw, sent another helpful reminder to support, we'll see if they've made any headway (not holding my breath). OS4 seems like it offers a lot of really cool features, but it's useless to us if we can't play back notes in time.

cheers,
paul.

[marc]

No, I haven't talked to support about this exact matter, but apparently Paul has... And with no luck...

you should consider that option.

marc

[Roby31]

It's "funny" to see people saying "Access didn't care about the standalone" when everybody else says that standalone mode works better. I have no timing issues whatsoever in standalone mode, using my TI2 desktop, cheapass m-audio ks 49e - and my patches sure as hell aren't DSP-easy. 6 mod slots, 3 lfos fully used, delay + reverb + distortion + character, 2 hypersaw + 4x unison and I still get to play with some polyphony

[Storm Music]

Thought I might mention that by stand-alone, I mean "live".
I'm gigging a lot with my virus TI2 - and I did a whole lot of touring all over the world with my old indigo (which passed away after years of duty).
And IMO there's no doubt:
The Ti2 has an issue with processing several notes at a time if you start using just a little more than 2 voices...
Never ever had that problem with my indigo...
And no, I'm not just untight... :-)

[paulkeeley]

It's "funny" to see people saying "Access didn't care about the standalone" when everybody else says that standalone mode works better. I have no timing issues whatsoever in standalone mode, using my TI2 desktop, cheapass m-audio ks 49e - and my patches sure as hell aren't DSP-easy. 6 mod slots, 3 lfos fully used, delay + reverb + distortion + character, 2 hypersaw + 4x unison and I still get to play with some polyphony

interesting, this is very different than my experience. MIDI timing is really quite bad with my TI, even when using the USB (ruling out my MIDI interface).

would you be perhaps interested in trying a test for me? i created a test project in ableton a while back to illustrate my point, that other users on this forum were able to reproduce (the thread has since been deleted) - i'd be very curious to know if you are experiencing the same issues. if this is a hardware issue specific to my unit, i'd very much like to know, although support was able to replicate the problem (almost a year ago now).

i'll compile the test shortly and post it here when ready.

[paulkeeley]

ok i've uploaded the test project if anyone is interested to try it out. the project is saved in Ableton Live7 format, but the recorded audio files from my TI are there (pay no mind to the start points). the MIDI is just a repetition of the same 8-note chord quantized to 8th notes. also included is the sysex of the single test patch i created to illustrate the issue:

http://www.mediafire.com/?sharekey=2f9dd4903e617f52ab1eab3e9fa335ca10c5322c2d2fe635

for each recorded audio file, i changed OSC1 to a different oscillator type, starting with a classic sawtooth. oscillator balance is at 50%, and OSC2 does not change at all. i turned the filter cutoff down and set the filter envelope to a quick decay (sustain at 0) and moderate release. same goes for the amp envelope. otherwise, there are no effects, no doubling, etc. just plain-jane synth sounds that should not be taxing the DSP at all.

i've sent the same project along to Access support, but i'd be curious to know if anyone else has the same issue.

[Roby31]

Will this work on a demo version? Live is not my daw of choice

edit: doesn't open on a demo version of live 8, windows 7 x64.

[Timo]

Would you like me to host those files for you Paul? We've a little more bandwidth here.

[paulkeeley]

Would you like me to host those files for you Paul? We've a little more bandwidth here.

yes please! i've sent you a PM with the file details.

and, in case anyone else doesn't use ableton, i have exported the midi clip for use in your DAW of choice. it can be found here for the time being:

http://www.mediafire.com/file/kjdmfft1tzq/VR_TI_Timing_Test.mid

[Timo]

Sorted. :)

Here are Paul's files (right-click, Save Target As)...

Ableton project:
VR_TI_Timing_Test_Project.zip (http://www.infekted.org/virus/files/paulkeeley/VR_TI_Timing_Test_Project.zip) (24.4MB)

TI patch sysex:
Virus_TI_TestP.syx (http://www.infekted.org/virus/files/paulkeeley/Virus_TI_TestP.syx) (0.5KB)

Exported MIDI:
VR_TI_Timing_Test.mid (http://www.infekted.org/virus/files/paulkeeley/VR_TI_Timing_Test.mid) (0.5KB)

[Storm Music]

Thanks for your post, Paul.

I just can't believe that Access actually let down live pro keyboard players this way.
I used to own a virus indigo and never had this problem.
Actually it's driving me so crazy that I think I'm just gonna part with this machine.

[MBTC]

A lot of this, both latency and polyphony comments sound similar to what I wrote about in the "Nothing but trouble" post. That was all USB+VC+Host of course, but many of the symptoms sure seem the same.