Topic: Needing Converter Delay on Digiface USB

Hi there. I purchased a Digiface USB for doing some specific testing and measurements. I've owned a few RME devices in the past and I'm familiar with TotalMix. I need to know how many samples of latency there is from digital in, through TotalMix, and out the headphone DA on the Digiface USB at 48kHz, please.

If available, those same specs at 44.1kHz and 96kHz would be greatly appreciated.

I can find specs on the converter latency for lots of RME devices and interfaces, but not Digiface.

Thanks!

2 (edited by waedi 2024-02-06 05:20:36)

Re: Needing Converter Delay on Digiface USB

There is only one converter in the Digiface USB, that one for the headphone output, if other headphone outputs from other RME interfaces have specs those are most probably similar or the same.
From guessing, less than one sample, there are no FX in the Totalmix of the Digiface USB.

M1-Sequoia, Madiface Pro, Digiface USB, Babyface silver and blue

Re: Needing Converter Delay on Digiface USB

Thank you. I need the precise number of samples, though.

Re: Needing Converter Delay on Digiface USB

Matt Hepworth wrote:

Hi there. I purchased a Digiface USB for doing some specific testing and measurements. I've owned a few RME devices in the past and I'm familiar with TotalMix. I need to know how many samples of latency there is from digital in, through TotalMix, and out the headphone DA on the Digiface USB at 48kHz, please.

If available, those same specs at 44.1kHz and 96kHz would be greatly appreciated.

I can find specs on the converter latency for lots of RME devices and interfaces, but not Digiface.

Thanks!

I can't remember any product where RME documented the latency of headphone outputs…
Only for analogue inputs and outputs which are designed as analogue I/O providing different line/reference levels.

You are free to use a headphone output as unbalanced line level output if you want or need, but it has its limitations:
unbalanced and for the Digiface USB the line level is limited to +9.5 dBu.

Regarding latency of internal processing, look here:

https://forum.rme-audio.de/viewtopic.ph … 60#p213660
MC: all our FX (in TM FX and the ADI-2s) run in real-time. Processing is done within a sample.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub14

Re: Needing Converter Delay on Digiface USB

ramses wrote:

I can't remember any product where RME documented the latency of headphone outputs…

MC: all our FX (in TM FX and the ADI-2s) run in real-time. Processing is done within a sample.

Thanks, Ramses.

I know that the effects are only 1 sample, but I think there are additional samples used for routing, which I believe may be either two or three samples.

Hopefully MC may have the answer for the converter delay and the routing/TM delay?

6

Re: Needing Converter Delay on Digiface USB

Nope. I would have to measure this myself. Will do.

Regards
Matthias Carstens
RME

Re: Needing Converter Delay on Digiface USB

MC wrote:

Nope. I would have to measure this myself. Will do.

Thank you very much!

8

Re: Needing Converter Delay on Digiface USB

The Digiface USB uses two different DAC chips (for the usual reason, AKM was no longer available). The difference in latency is 2 samples. The unit itself needs 4 samples in Single Speed and 6 samples in Double Speed for SPDIF receiver and internal processing.

So your DF USB takes either 7.5 / 9.5 samples or 9.8 / 11.8 samples from SPDIF input to analog output.

Regards
Matthias Carstens
RME

9 (edited by ramses 2024-02-08 12:40:34)

Re: Needing Converter Delay on Digiface USB

MC wrote:

The Digiface USB uses two different DAC chips (for the usual reason, AKM was no longer available). The difference in latency is 2 samples. The unit itself needs 4 samples in Single Speed and 6 samples in Double Speed for SPDIF receiver and internal processing.

So your DF USB takes either 7.5 / 9.5 samples or 9.8 / 11.8 samples from SPDIF input to analog output.

EDIT: corrected according to information from MC (https://forum.rme-audio.de/viewtopic.ph … 49#p214149)

https://www.dropbox.com/scl/fi/phb7qp1itabbyuunlqyuy/2024-02-07-Digiface-USB-Latency-from-SPDIF-to-Phones.jpg?rlkey=azlnqobt01nowq6m6bqpp3c8r&dl=1

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub14

Re: Needing Converter Delay on Digiface USB

MC wrote:

The Digiface USB uses two different DAC chips (for the usual reason, AKM was no longer available). The difference in latency is 2 samples. The unit itself needs 4 samples in Single Speed and 6 samples in Double Speed for SPDIF receiver and internal processing.

So your DF USB takes either 7.5 / 9.5 samples or 9.8 / 11.8 samples from SPDIF input to analog output.

That is fantastic! Many thanks for the data!

Re: Needing Converter Delay on Digiface USB

ramses wrote:
MC wrote:

The Digiface USB uses two different DAC chips (for the usual reason, AKM was no longer available). The difference in latency is 2 samples. The unit itself needs 4 samples in Single Speed and 6 samples in Double Speed for SPDIF receiver and internal processing.

So your DF USB takes either 7.5 / 9.5 samples or 9.8 / 11.8 samples from SPDIF input to analog output.

Is this correctly understood? Assumed ESS chip is the later one with much likely lower converter latency?

https://www.dropbox.com/scl/fi/phb7qp1itabbyuunlqyuy/2024-02-07-Digiface-USB-Latency-from-SPDIF-to-Phones.jpg?rlkey=azlnqobt01nowq6m6bqpp3c8r&dl=1

I would suspect the AKM has the lower latency.

12

Re: Needing Converter Delay on Digiface USB

Nope, AK4452 (AKM) with 5.8 samples and PCM5102A (TI) with 3.5 samples.

Regards
Matthias Carstens
RME

13 (edited by ramses 2024-02-08 13:02:39)

Re: Needing Converter Delay on Digiface USB

MC wrote:

Nope, AK4452 (AKM) with 5.8 samples and PCM5102A (TI) with 3.5 samples.

Thanks, Matthias. Updated the Excel and the screenshot.

But I wonder whether the numbers above (underlined) are of the same sample rate.

Based on your information in post #8 I re-calculated the converter latency in the Excel.
Col B: SPDIF and processing was given/fix
Col E+F: Total Latency was given/fix
Col C+D: converter latency was calculated to be the delta Col. E/F - Col. B

If I didn't make any mistakes, it looks, as if the numbers belong to different sample rates:
5.8 samples (AKM) seem to match to double- and 3.5 samples (TI) to single speed.

https://www.dropbox.com/scl/fi/phb7qp1itabbyuunlqyuy/2024-02-07-Digiface-USB-Latency-from-SPDIF-to-Phones.jpg?rlkey=azlnqobt01nowq6m6bqpp3c8r&dl=1

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub14

Re: Needing Converter Delay on Digiface USB

MC, may I ask what is the latency from ADAT in to ADAT out? Is it 4/6 samples? Or more if conversion to ADAT out also takes some time?

15

Re: Needing Converter Delay on Digiface USB

Subtract the DAC and you get the number. ADAT out is not delayed.

Regards
Matthias Carstens
RME

Re: Needing Converter Delay on Digiface USB

Thanks, MC. So it is 4 (SS)/6 (DS) samples.

Is it roughly similar for other RME interfaces? ADAT in to ADAT out.

17

Re: Needing Converter Delay on Digiface USB

It's different over the range of our interfaces due to differences in internal bufferring.

Regards
Matthias Carstens
RME