Questions about total integration

[piVVi]

I have some questions about Ti! :D
I'll think to buy a Ti keyboard and use it with my computer, without audio card. I'll use ableton live to control the 16 parts of the virus (by usb midi).
I want also use some vst effects on the virus parts. I want to monitoring all by headphone out of the virus.
Is this possible?
I'll have some problems with latency or is like to use a normal hardware synth (zero latency)?.

thanks to all

[nutekk]

in theory you should be able to accomplish what you want to.
in actuallity it may come down to your computer components/drivers.

hardware synths are not zero latency...however they are a very usable latency of say under 15ms.
you should be able to keep your system latency around this.

my TI is working nicely with a DFI lanparty nf4 motherboard
and a creamware scope system.

ill caution you though by monitoring all by headphone...
this will make it an uphill battle to get a usable mix(as if it wasnt hard enouph)
and to much headphone use will damage your hearing. :wink:

[piVVi]

hardware synths are not zero latency...however they are a very usable latency of say under 15ms.
you should be able to keep your system latency around this

mmm... actually I play with 2 ms with my setup

my TI is working nicely with a DFI lanparty nf4 motherboard
and a creamware scope system.

I have also a nf4 motherboard (asus)

caution you though by monitoring all by headphone...
this will make it an uphill battle to get a usable mix(as if it wasnt hard enouph)
and to much headphone use will damage your hearing.

eheh no other options actually :D

thanks for your reply.
a last question: if you use the output of your Ti for monitoring (so without your creamware) what is the latency that you can obtain?

[nutekk]

sure 2ms...for your computer sound card.
but many hardaware synths do not have this tight of timing!

as for the TI Latency i had it around 6ms...
dont have it set up that way now,using my creamware
but it seemed stable @ that setting.\

[grs]

I've got the TI in Cubase SX 3.02 in sync using a combination of USB outs and hardware outs. But I'm not using the Virus Asio, not enough channels realy. The trick is to route the virus TI analog outs to your seqencer and then using a delay plugin to sync back up. The analog outputs are acutally ahead of time due to the Virus Control vsti plugin having to be delay compinsated by the host. Make audio inputs and audio tracks for you hardware outs and press the preview speaker on the audio track.

In all the delay plugin is set to 24ms on each track that is a virus hardware out. If you need realtime monitoring just undo the delay plugin for recording a part.

It's also better to use the analog if your sound card is good because of the extra bit depth. Try putting a harsh pumping 50:1 compressor on the USB outs and listen to the awfull tails.

[piVVi]

mmm... I understand nothing! ;-)
But, If I monitor all the part from virus headphone out, why I have to send audio through usb port? usb must trasmit only midi, right?
So, If I play a key on the virus and listen it by his headphone out I have latency?

Maybe in the last moment, when I need to render the complete track I have to use usb for audio, right? why usb audio sound not good as analog audio? In theory if you use analog you have to do dual conversion, with the usb all remain in digital domain, right?

thanks for your time

[Timo]

hardware synths are not zero latency...however they are a very usable latency of say under 15ms.

Standalone hardware synths (going straight into speakers, without going into your computer) have latencies of 15ms?! I seriously don't think so!

True, by the very nature of DSP processing/chips being used in hardware synths, there will be a finite latency (1ms, maybe?), but absolutely nowhere near 15ms!

For example, at 10ms latency, even, you can already feel stuff lacking that critical, immediate response when using soft-synths. Hardware synths just don't suffer from that, as their signal flow is much leaner, cleaner, and much more heavily optimised (machine code, dedicated DSPs, etc.).

Unless you're talking about triggering the hardware synths via MIDI through an external sequencer (from your computer), in which case, midi latencies triggered and interfaced via Firewire would be about 5ms, USB probably the same or slightly more, but PCI MIDI interfaces suffering much less @ around 1ms.

For monitoring and routing audio from an externally played hardware synth through effects on a computer, the audio latency will be limited by your computer resources (CPU) and also your soundcard interface and its respective software drivers. All being well, there are options, I think, allowing you go down as far as 1.5ms, but your CPU usage will rocket, depending on load. Most people compromise their latencies slightly, to allow them to process more channels at any one time, such as 5ms or greater. 10ms or more and you start to notice the lag. It also depends on what you want to do with the processed audio... Due to the lag induced by computers, you can't often mix the processed audio back with the original, as it creates comb-filtering due to the altered delays/phases.

PS > Slightly OT, nutekk, but how are you finding your NForce4 based computer? I heard NForce4 was initially a no-no with regards to critical realtime music programs and processing, causing unduly high CPU usage and high latencies?

Check out these threads:-

[Nuendo forum thread - "NForce4 tests" (http://forum.nuendo.com/phpbb2/viewtopic.php?t=706)]

[Sound-On-Sound forum thread - "Warning NForce/PCI-E" (http://www.soundonsound.com/forum/showflat.php?Cat=&Board=PCMus&Number=145032&Searchpage=1&Main=143845&Words=nforce4&topic=&Search=true#Post145032)]

[RME Audio NForce4 warning (http://www.rme-audio.com/english/techinfo/nforce4_tests.htm)]

I guess/hope it's been ironed out, now?

[nutekk]

do some research dude!
yes through synth history..there have been models with bad keyboard latencies.
memorymoog was noted for having a 12ms latency.
i believe an emulator2 also was a problem
even a roland jv1080 has a 5ms latency!
and many others.
i have a paper somewhere with lists of synths and there respective latencies.(i used to manage a used synth store in NYC -CROCODILE)
NONE WERE ZERO LATENCY!


im not going to just talk bullshit on this forum.

anyway the nf4 is fooking great for audio.
no problems what so ever...
as for the pci-e problemsi have put it in an 8x slot instead of the 16x
so the the bandwidth availible is less:)
but i think it was hit or miss anyway
i upgraded from a nf2 barton3000 and experianced huge gains
so im happy!
the ultra-d's bios is very programmable so it lets you really tune the
system nicely..plus the overclocking on this board kills most other models.

[techno_7]

12ms? Thats not bad. I run my ASIO drivers on 11 ms, and it all works(and sounds) perfect.

[piVVi]

mmm... I understand nothing! ;-)
But, If I monitor all the part from virus headphone out, why I have to send audio through usb port? usb must trasmit only midi, right?
So, If I play a key on the virus and listen it by his headphone out I have latency?

Maybe in the last moment, when I need to render the complete track I have to use usb for audio, right? why usb audio sound not good as analog audio? In theory if you use analog you have to do dual conversion, with the usb all remain in digital domain, right?

thanks for your time

please, can someone reply to this?
thanks!

[marc]

do some research dude!
[...]
memorymoog was noted for having a 12ms latency.

i wonder if there's a place where i can read more about it - simply for the fact that i cannot understand how it would - but at the same time, i've never owned one

marc

[grs]

mmm... I understand nothing! ;-)
But, If I monitor all the part from virus headphone out, why I have to send audio through usb port? usb must trasmit only midi, right?
So, If I play a key on the virus and listen it by his headphone out I have latency?

Maybe in the last moment, when I need to render the complete track I have to use usb for audio, right? why usb audio sound not good as analog audio? In theory if you use analog you have to do dual conversion, with the usb all remain in digital domain, right?

thanks for your time

please, can someone reply to this?
thanks!

If you understand nothing it's hard for someone here to teach you everything. Basically the TI can send sounds to the computer via the USB aswell as midi data, this mode is found when you run a vsti plugin from Access called the Virus Control. If you don't get into this mode your basically running the TI like a normal synth.

At the moment, as I understand it the headphone output is just a mirror of the OUTPUT 1 of the TI also the spdif out is the same. If you choose the TI as a sound card (ie audio playback in your software) you will hear your software outs through OUTPUT 1 along with any TI synth sounds.

You will not hear any sound that has it's output set to output2 in the head phones unless you route it through your studio / computer and back via Access Virus TI sound driver (ASIO or MME).

That will get you started.

[ben crosland]

But, If I monitor all the part from virus headphone out, why I have to send audio through usb port? usb must trasmit only midi, right?
So, If I play a key on the virus and listen it by his headphone out I have latency?

If you want to play a part live with no perceptible latency, then you can set the output of that part to the main outs. When the part is being played back by the sequencer, you set it back to a usb out, so that the delay compensation works properly.

[Hollowcell]

do some research dude!
[...]
memorymoog was noted for having a 12ms latency.

i wonder if there's a place where i can read more about it - simply for the fact that i cannot understand how it would - but at the same time, i've never owned one

marc

Me too Nutekk. Could you set us up with a link please?

[piVVi]

But, If I monitor all the part from virus headphone out, why I have to send audio through usb port? usb must trasmit only midi, right?
So, If I play a key on the virus and listen it by his headphone out I have latency?

If you want to play a part live with no perceptible latency, then you can set the output of that part to the main outs. When the part is being played back by the sequencer, you set it back to a usb out, so that the delay compensation works properly.

Thanks Ben for your reply.
but, when " I set it back to a usb out" the sound is played back from virus main output or from an additional sound card?

[ben crosland]

but, when " I set it back to a usb out" the sound is played back from virus main output or from an additional sound card?

From the Virus' main out, so you'll still hear it through the 'phones. If you don't set it back to usb, the sequencer still treats it the same as though it was, so it get's delay compensation when it doesn't need to. This results in it playing well ahead of the beat.

[nutekk]

who said i have a link?
if you dont believe that some synths were designed poorly and had latency issues...then dont believe it.
i have extensive experiance with most of the classic synths
And through this experiance i have gained the knowledge.
as said the memorymoog(not sure if it was the original)
emu emulator2,roland d-70 was notorious,some obreheim synths suffered
there are plenty others..
why is this so hard to believe?
i doubt there is a magical link that has all this laid out
and has been stamped by a judge..for you to trust in.

As i said i have a sheet somewhere that has a bunch of popular synths with higher latencies...and there was a full page.
this info belonge to the synth tech that we would use to refurbish/and check out all the synths that we bought and sold.

i dont really care to prove it to you.
do your own research.

[piVVi]

but, when " I set it back to a usb out" the sound is played back from virus main output or from an additional sound card?

From the Virus' main out, so you'll still hear it through the 'phones. If you don't set it back to usb, the sequencer still treats it the same as though it was, so it get's delay compensation when it doesn't need to. This results in it playing well ahead of the beat.

ahhh now I understand!!!!!!!
delay compensation anticipate USB tracks to cancel the latency so it's all in sync and I can play the keyboard (assegned directly to audio out) without latency!!!
If I understand right is fantastic!!!!

thanks Ben! :D

[Hollowcell]

who said i have a link?
if you dont believe that some synths were designed poorly and had latency issues...then dont believe it.

It's not that I don't beleive, it's just that I've never encountered anything like it.

i have extensive experiance with most of the classic synths
And through this experiance i have gained the knowledge.
as said the memorymoog(not sure if it was the original)
emu emulator2,roland d-70 was notorious,some obreheim synths suffered
there are plenty others..
why is this so hard to believe?
i doubt there is a magical link that has all this laid out
and has been stamped by a judge..for you to trust in.

As i said i have a sheet somewhere that has a bunch of popular synths with higher latencies...and there was a full page.
this info belonge to the synth tech that we would use to refurbish/and check out all the synths that we bought and sold.

I've played with a few boxes in my time as well. I get 5ms of latency with my software synths and I can definitely feel that latency when playing hard 'n' fast leads. I have never felt this on any hardware synths.

i dont really care to prove it to you.
do your own research.

Fair enough....

As you sound like you have first hand knowledge, I have some questions for you to help my research:

1: When you were getting latency, was it occuring when playing from the onboard keyboard, or triggered from something else?
2: Were you triggering from a thru-daisy-chained config?
3: Were you using a PC/Mac sequencer running a USB midi interface?
4: Was the latency occuring when you were running many midi channels at the same time?

The research continues...... :D

[nutekk]

as i said i dont really care to prove it to you!
believe it or not.

[DIGITAL SCREAMS]

Nutekk is talking about sluggish software EG's on some 80's synths. He's a bit confused....

DS

[marc]

who said i have a link?
if you dont believe that some synths were designed poorly and had latency issues...then dont believe it.

i dont really care to prove it to you.
do your own research.


relax - i wasn't saying that you have a link, i was asking you for one. it was you saying that the memorymoog has a latency of 12ms ...

i was just curious since the memorymoog (which i never owned) seems to have analog filters and oscillators and therefore i wouldn't be able to understand where the latency would come from. you are certainly right for many synths build in the earlier days of the digital age, especially the D-70 etc.

marc

[nutekk]

nah im not confused...
latency is latency...for whatever reason.

here google memorymoog latency...
3rd hit
http://www.macintouch.com/makemusic.html
(somewhere in this page)
actually it was the memorymoog +
seems it was takin care of in some update...but it was there :wink:

there are plenty of others.

even a jv1080 had between a 4.5ms-5 ms latency due to 64note poly design..and thats as you know a rompler.

[nutekk]

man you guys had to get me goin.... :evil:

http://www.sequencer.de/oberheim/oberheim_ob12.html

even mentions it being worse! then the d70 lol..

shall i continue?

[nutekk]

http://www.soundonsound.com/sos/Sep02/articles/pcmusician0902.asp
jv1080
ok so they say 4.4 ...but i bet they are just being nice!

[nutekk]

there are more...
you find emm if your so interested..
im not.

[Hollowcell]

Cool, thanks for the links.

Always interesting to read about this sort of thing when I'm meant to be working. :D