Switch to Print All | Page 1 of 1

Does Zebra's timing sound sloppy here to you?

PostPosted: Thu Feb 14, 2013 1:10 pm
by Aiynzahev
I'm just waiting for SC to process the file. But has anyone else found Zebra a bit sloppy? I am programming chords and it sounds like each note is hitting slightly off compared to the other, especially in comparison to Sylenth1.

PostPosted: Thu Feb 14, 2013 1:40 pm
by pdxindy
try turning OSC Reset on in Zebra... see if that is it...

PostPosted: Thu Feb 14, 2013 1:48 pm
by Aiynzahev
pdxindy wrote:try turning OSC Reset on in Zebra... see if that is it...


Thats not it. I'm running an arp presets now that I made, very short attack, reset on, drift off, smooth attack of, 16x envelope. The timing goes funny when the bar loops. The loop points are correct.

I'm hearing it against a kick, kick is bounced down and the timing is checked to the grid.

It's Studio One v 2.5 and this not the firs time I've had timing issues, virtually No arp is solid and with synths it varies. Corona for example was all over the place, and now Zebra is. Sylenth is fine though.

PostPosted: Thu Feb 14, 2013 1:58 pm
by pdxindy
I've heard that Studio One has timing issues...

What OS do you use?

PostPosted: Thu Feb 14, 2013 2:10 pm
by Urs
Yeah, we highly discourage using the Arp or MSEGs in a looped DAW region. Even more so if Swing was activated.

It works well when retriggered after a few bars, but for some reason there's something odd happening when a loop wraps around.

That is most likely related to the way VST (and all other plugin standards) propagates timing information and the way DAWs handle buffers. If a DAW uses only fixed buffer sizes, they either render shortly *past* the loop point, or the loop may happen right within the middle of a buffer. In both cases we're f**ked for timing.

PostPosted: Thu Feb 14, 2013 3:03 pm
by Aiynzahev
Thanks that at least tells me the problem is a known one. It drives me batty and sometimes I wondered if it's something to do with my computer build. Most people seem to be able to use Arps just fine but I always have issues, not always at loop points either.

I am using OS 7 64-bit.

PostPosted: Thu Feb 14, 2013 3:12 pm
by Urs
Well, my global Swing thing in Zebra was a big let down - the way it works just makes good timing impossible in the long run. Which is why Arps, MSEGs and even LFOs run out of sync after a few bars. It has been improved vastly since 2.5.2, but it's still not optimal.

Zebra3 will work without global Swing, we'll add swing to individual components instead.

PostPosted: Thu Feb 14, 2013 3:24 pm
by Aiynzahev
Any ideas why timing is a little off when playing chords without the ARP?

I programmed a chord riff and took a close look at the bounced result, the start time did vary a little, hard to tell if it's enough to hear (of if my mind is playing tricks.)

In others words the visual finding deep seem to back up what I was hearing.

I found it's possibly affected by the Release and the polyphony Zebra is set to.

PostPosted: Thu Feb 14, 2013 3:29 pm
by Aiynzahev
pdxindy wrote:I've heard that Studio One has timing issues...

What OS do you use?


The again whenever I've had timing issues in S1 I've replicated then in Cubase 6. hence the thought that it's my pc or something.

PostPosted: Thu Feb 14, 2013 4:13 pm
by Urs
Aiynzahev wrote:Any ideas why timing is a little off when playing chords without the ARP?

I programmed a chord riff and took a close look at the bounced result, the start time did vary a little, hard to tell if it's enough to hear (of if my mind is playing tricks.)

In others words the visual finding deep seem to back up what I was hearing.

I found it's possibly affected by the Release and the polyphony Zebra is set to.

Gosh, yes, that's another issue.

When Zebra's voices are maxed out then it needs to "steal" existing voices. In order to prevent nasty clicks, it quickly fades a voice out that is scheduled for stealing, and postpones the new note by a few milliseconds, typically within the boundaries of live play.

Our later synths don't do that, they just restart the voice from where it is and use other means to prevent clicks. I haven't transferred this principle to Zebra yet, but Zebra3 will probably adopt the newer style note stealing.

PostPosted: Thu Feb 14, 2013 4:33 pm
by pdxindy
Urs wrote:
Aiynzahev wrote:Any ideas why timing is a little off when playing chords without the ARP?

I programmed a chord riff and took a close look at the bounced result, the start time did vary a little, hard to tell if it's enough to hear (of if my mind is playing tricks.)

In others words the visual finding deep seem to back up what I was hearing.

I found it's possibly affected by the Release and the polyphony Zebra is set to.

Gosh, yes, that's another issue.

When Zebra's voices are maxed out then it needs to "steal" existing voices. In order to prevent nasty clicks, it quickly fades a voice out that is scheduled for stealing, and postpones the new note by a few milliseconds, typically within the boundaries of live play.

Our later synths don't do that, they just restart the voice from where it is and use other means to prevent clicks. I haven't transferred this principle to Zebra yet, but Zebra3 will probably adopt the newer style note stealing.



24 and 32 note polyphony would be nice too... across the u-he synths...

PostPosted: Thu Feb 14, 2013 4:58 pm
by hollo
I also use to restart the notes after a few bar to solve the timing.

I'm happy to hear that you are fighting with Zebra arps Sami,
that's mean we'll have a soundset from you soon! ? :wink:

PostPosted: Thu Feb 14, 2013 7:04 pm
by Aiynzahev
Urs wrote:
Aiynzahev wrote:Any ideas why timing is a little off when playing chords without the ARP?

I programmed a chord riff and took a close look at the bounced result, the start time did vary a little, hard to tell if it's enough to hear (of if my mind is playing tricks.)

In others words the visual finding deep seem to back up what I was hearing.

I found it's possibly affected by the Release and the polyphony Zebra is set to.

Gosh, yes, that's another issue.

When Zebra's voices are maxed out then it needs to "steal" existing voices. In order to prevent nasty clicks, it quickly fades a voice out that is scheduled for stealing, and postpones the new note by a few milliseconds, typically within the boundaries of live play.

Our later synths don't do that, they just restart the voice from where it is and use other means to prevent clicks. I haven't transferred this principle to Zebra yet, but Zebra3 will probably adopt the newer style note stealing.


So I read in the manual that the polyphony is not straight forward, it's optimized, so would enabling reverb for example lower the polyphony? and how about the number of oscs and using superwaves on them?

Thanks, at the very least I know It's not my mind playing tricks on me.

PostPosted: Thu Feb 21, 2013 9:10 pm
by Aiynzahev
So can anyone tell me if the voice allocation is dynamic, so that the more intense the patch is the fewer voices you actually have?

Also something was driving me batty, clicks, I found out if you back of the volume in the lane (which turns everything down I think) then the clicks stop appearing. I mean these are annoying clicks or pops, not the click from the filter (awesome features, saves me an envelope)

PostPosted: Thu Feb 21, 2013 11:11 pm
by pdxindy
Aiynzahev wrote:So can anyone tell me if the voice allocation is dynamic, so that the more intense the patch is the fewer voices you actually have?


I have not noticed that... I think polyphony is polyphony... but then there is a natural tendency for high cpu patches to be played with fewer notes...