Ulli-Settings vs. Delaytime

A place for developers to share ideas and assist each other in solving problems.
Post Reply
User avatar
roy thinnes
Posts: 586
Joined: Thu Jan 09, 2003 4:00 pm
Location: Graz
Contact:

Ulli-Settings vs. Delaytime

Post by roy thinnes » Mon Mar 31, 2014 5:40 am

Hi,
lately I switched my Ulli settings and saw all delaytimes being out of sync.
Is there someone who can tell me the correlation between Ulli & DT (or even got a formula? :) )

User avatar
sharc
Posts: 637
Joined: Mon Oct 29, 2001 4:00 pm
Location: No idea. All looks the same down here

Re: Ulli-Settings vs. Delaytime

Post by sharc » Mon Mar 31, 2014 6:55 am

Nice timing Roy :)

I'm working on ULLI and sample rate dependent ranges for delays at the moment. If I'm not wrong only delays using host ram should be offset by ULLI setting. Logically this should be a round-trip latency, but I've yet to confirm this.

User avatar
roy thinnes
Posts: 586
Joined: Thu Jan 09, 2003 4:00 pm
Location: Graz
Contact:

Re: Ulli-Settings vs. Delaytime

Post by roy thinnes » Mon Mar 31, 2014 8:27 am

Logically this should be a round-trip latency
Thanks sharc, I think you placed me on the right track.

1000ms->48000
1ms->48

the Ulli setting w/ exact delay times was 25ms (44.1kHz)
my new Ulli setting is 3ms
25-3=22
22x48=1056
1056x2(round-trip :wink: )=2112

I added 2112 to delaytime value
>voilà it's in sync again! :)

so it's just that easy for all Ulli settings:
make a textfader string with 5 values, for each Ulli setting.
the values are:
2112 for 3ms
2016 for 4ms
1728 for 7ms
1152 for 13ms
0 for 25ms
and add the textfader value to the delay value.

User avatar
sharc
Posts: 637
Joined: Mon Oct 29, 2001 4:00 pm
Location: No idea. All looks the same down here

Re: Ulli-Settings vs. Delaytime

Post by sharc » Mon Mar 31, 2014 12:54 pm

Thanks for sharing Roy. That explains why my test results were getting skewed between the SDK and SFP - Different ULLI settings :)

happycritter
Posts: 52
Joined: Wed Oct 23, 2013 8:51 pm

Re: Ulli-Settings vs. Delaytime

Post by happycritter » Mon Mar 31, 2014 9:16 pm

roy thinnes wrote:
Logically this should be a round-trip latency
Thanks sharc, I think you placed me on the right track.

1000ms->48000
1ms->48

the Ulli setting w/ exact delay times was 25ms (44.1kHz)
my new Ulli setting is 3ms
25-3=22
22x48=1056
1056x2(round-trip :wink: )=2112

I added 2112 to delaytime value
>voilà it's in sync again! :)

so it's just that easy for all Ulli settings:
make a textfader string with 5 values, for each Ulli setting.
the values are:
2112 for 3ms
2016 for 4ms
1728 for 7ms
1152 for 13ms
0 for 25ms
and add the textfader value to the delay value.
Thanks Roy!

This is very useful information and another wrinkle to development that I hadn't given full consideration. I knew sample delays can introduce significant issues, but thanks to your post I realize the hardware end is by no-means 'hands-free'. :)
xw8400
2.33 GHz Xeon Quad core (E5345 Socket 771) x2
32Gb ECC buffered memory
AMD HD7700
Samsung 850 EVO 250Gb root drive
Windows 7 Pro (64bit)
Sonar Platinum
XITE-1 (flipping awesome!)

Post Reply