• Registered: 2011-11-02
  • Posts: 55

Topic: Firefac UFX ADAT Latency

Hi,

I've got a Digiface (sync/clock master) connected to a Fireface UFX (sync/clock  slave) with the following routing:

(1)  Digiface ADAT Out 1 going to UFX ADAT In 1
(2)  UFX Analog Out 3 looped back to UFX Analog In 3
(3)  UFX ADAT Out 1 to Digiface ADAT In 1.

Logic's Ping is reporting an latency offset 69 samples.  I know from researching the converters for the UFX that the ADC+DAC latency is around 22 samples.

This means that the internal routing of the UFX is adding around 47 samples of latency.

I have been told on this forum that the internal ADAT routing of the Fireface UCX is under 7 samples.

Does 69 samples of additional latency for the UFX running off a Digiface via ADAT sound right.  It sounds a bit high to me considering the fast ADAT latency of 7 samples for the UCX.

Logic 9.1.8 (1700.67) (32-bit) - OS X Lion 10.7.5
Digiface <-> [Fireface UCX & Prism Orpheus - ADAT]

2 Reply by Randyman... (edited by Randyman... 2012-10-31 02:08:51)

  • From: Houston, TX
  • Registered: 2007-05-11
  • Posts: 1,191

Re: Firefac UFX ADAT Latency

I see specs of 12 Samples for ADC (@ x1fs) and 28 Samples for DAC according to Page 99 UFX's online Manual (lower than I thought!).  That's 40 samples of AD/DA latency, plus 7 samples in TM (x2 for two instances of ADAT routing in TM on the Digiface) is a total of 57 Samples.  I'm not sure if TM has the same offset for a Digiface and for the UFX. 

You also have the Analog-to-ADAT routing happening TWICE in the UFX (ADAT to Analog, then Analog to ADAT).  So, another 14 samples or so.  That might account for the "extra" latency you are measuring (seems to be within 2-3 samples of your reported measurement).

Beyond that, Does Logic report a zero offset for a Digiface ADAT I/O Loop?  It could be possible Logic is the oddball, but I'm thinking the incorrect AD/DA offsets plus the extra TM routings that were un-accounted for should explain it...

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

3 Reply by DanRand (edited by DanRand 2012-10-31 06:51:30)

  • Registered: 2011-11-02
  • Posts: 55

Re: Firefac UFX ADAT Latency

DanRand wrote:

... ...

I bought the Digiface to improve the latency of the Orpheus, but only saw a 2 ms improvement (EDIT:  due to a problem with large internal ADAT buffering in the Orpheus; the Digiface works perfectly by the way) ... ...

They (Prism) have also confirmed that the added round-trip latency (non AD-DA) of this ADAT path is around 4 ms!  I was shocked to hear this.

... ...

With the UCX in stand-alone mode, can someone tell me how many samples of round-trip latency I can expect (not including AD-DA conversion) to incur inside the UCX ... ...

... ...

Danno

MC wrote:

The delay is below 7 samples.

So it looks like the internal (non-converter) round-trip latency of the UCX is below 7 samples.  This is for ADAT I/O and Total Mix.

I've read the manual and read the latency details a number of RME's interfaces except the UFX!  Didn't know the DAC latency was 28 samples.  Thought I knew the Group Delays of the UFX's converter chips.  Maybe I know nothing.  Oops HeadScratch

For the UCX:  That's (12 + 28) + 7 = 47 samples, which represents 40 samples of ADC+DAC converter latency plus a maximum of 7 samples of internal latency (ADAT + Total Mix).

69 - 47 = 22 samples unaccounted for.

Haven't got round to doing an ADAT loopback yet.  Just driven 300 miles with my gear to get everything ready for some studio time.

Of course the 7 sample internal latency figure is for the UCX, not the UFXxxXXX...

Logic 9.1.8 (1700.67) (32-bit) - OS X Lion 10.7.5
Digiface <-> [Fireface UCX & Prism Orpheus - ADAT]
  • Registered: 2011-11-02
  • Posts: 55

Re: Firefac UFX ADAT Latency

I assume there should be no problems running a Digiface and UFX on the same Mac?

Logic 9.1.8 (1700.67) (32-bit) - OS X Lion 10.7.5
Digiface <-> [Fireface UCX & Prism Orpheus - ADAT]