Topic: Downsampling ADAT clock rate only

Hello, a question/feature request regarding ADAT clock rates:

I prefer to record at 96khz, but I also want to have 8x channels of ADAT to convert in eurorack modular domain (using eg optx v2). Is there a way on devices like UCX II, etc. to downsample the ADAT output to 48khz to maintain the 8x channel count, while running everything else at 96k?

If not, would love to have a phase locked downsampling feature.

Re: Downsampling ADAT clock rate only

Sorry for no better answer than no.

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

Re: Downsampling ADAT clock rate only

objectgroup wrote:

Is there a way on devices like UCX II, etc. to downsample the ADAT output to 48khz to maintain the 8x channel count, while running everything else at 96k?

If not, would love to have a phase locked downsampling feature.

1) UCX II has no SRC;
2) Even if it had, there is only one clock domain, so it's not possible to have different sampling rates for different inputs/outputs.

Fireface UCX II + ARC USB > ADI-2 Pro FS R BE > Neumann KH 750 DSP + MA 1 > KH 120 A

4 (edited by ramses 2024-11-17 12:58:53)

Re: Downsampling ADAT clock rate only

objectgroup wrote:

Hello, a question/feature request regarding ADAT clock rates:

I prefer to record at 96khz, but I also want to have 8x channels of ADAT to convert in eurorack modular domain (using eg optx v2). Is there a way on devices like UCX II, etc. to downsample the ADAT output to 48khz to maintain the 8x channel count, while running everything else at 96k?

If not, would love to have a phase locked downsampling feature.

This is not possible in several ways

1. in an environment only one sample rate is allowed, the only way out would be an SRC (sample rate converter)

2. I never heard of multi-port SRC for 8 channels, there are only 3 RME products with SRC but only for 2 channels
    and if I remember right only in one direction: ADI-2 Pro FS R BE, ADI-2/4 Pro SE, Digiface AES

3. The Optx v2 is missing a 2nd ADAT port and doesn't support S/MUX, so you can't even use 4ch@double speed.
    In the specs they tell 44.1/48 kHz, 24bit.

To solve this issues several investments would be needed, easier and less expensive is to stay at single speed.

Recording Interface

To get a recording interface with similar feature set you would need to upgrade to an UFX II or even UFX III to have some more headroom in terms of channels (-> MADI).

Eurorack Modules

You need to check if there are 8-port Eurorack modules available which support 2x ADAT and double speed. I didn't find some.

A solution might be "system 500", but I have not much experience with it, sometimes we had questions in this forum related to expert sleepers modules and then this Rack from Cranebourne was the only one with 2x ADAT I/O for €1758
https://www.thomann.de/de/cranborne_audio_500adat.htm

Additionally you would need to check whether there is an Expert sleeper module for system 500 is available for you.

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

Re: Downsampling ADAT clock rate only

Maybe it would be possible somehow using RME Digiface USB that would run 44.1/48 kHz. You would have to do downsampling in computer and then send it to Digiface. But I guess it would be hardly possible to have everything in sync....

Re: Downsampling ADAT clock rate only

An the easiest solution is........
Use 48k. You should really do a double blind test.
Only if you sample synths and play them 2 or more octaves lower, you might have a case. But I would really check if there is anything usefull above 24khz and not just noise and garbage.

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
BFpro fs, 2X HDSP9652 ADI-8AE, 2X HDSP9632

Re: Downsampling ADAT clock rate only

+1

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

Re: Downsampling ADAT clock rate only

Or use an analog connection....

Regards
Daniel Fuchs
RME

9 (edited by ramses 2024-11-17 13:18:22)

Re: Downsampling ADAT clock rate only

+1
I had already thought about this, but rejected it because I was afraid that there wouldn't be enough analog I/O available.
It would be ideal if the number of analog ports on the UCX II was already sufficient, or if a bit of "channel loss" was OK.

Also, for such things, UFX II would have been the better choice, giving a few more analog ports of the same format
and with 2x ADAT I/O.

Or even UFX III to have MADI to have some headroom for even more Analog I/O. UFX III + M-1620 Pro (or if budget is limited with Ferrofish converters) via MADI.

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

Re: Downsampling ADAT clock rate only

@ Daniel. The way I understand it he is using the adat out for CV. The analog outs of the the rme devices are not suitable.

Vincent, Amsterdam
https://soundcloud.com/thesecretworld
BFpro fs, 2X HDSP9652 ADI-8AE, 2X HDSP9632

Re: Downsampling ADAT clock rate only

Oh, ok...

Regards
Daniel Fuchs
RME

Re: Downsampling ADAT clock rate only

objectgroup wrote:

If not, would love to have a phase locked downsampling feature.

+1

It will take time to enjoy different or customized sample rates for analog and digital on a single interface, the OS's and DAW's mediocrity is forced transmitted to the hardware design but OP's request is excellent and very coherent … Bravo!

UCX II FW106/34/104 v1.253 TM1.97 - PC Win11 23H2 / Fedora WS 41 - Reaper 7.27

Re: Downsampling ADAT clock rate only

While it might be "easy" to downsample to half of "main" sampling rate. Just divide clock by 2 and throw away every second sample... Maybe that even such an "easy" task is not doable on current UCX II hardware.

And other combinations, that would soon require other users, would be way more complicated.

But it might be handy in next generation RME devices....

14 (edited by ramses 2024-11-17 16:42:33)

Re: Downsampling ADAT clock rate only

Of course, one can wish for many things, but the key question is whether it's technically feasible and whether the implementation will be affordable. After all, we're talking about precise clock synchronization between devices, which takes place at the hardware level, not in software. If it were that simple, someone would have already brought it to market. The reason studios rely on expensive master clocks and specialized concepts for star-topology clock signal distribution is no coincidence – the matter is more complex than it may seem at first glance.

Throwing away every second sample? I doubt that it would be that easy. Because the multiplexed information somehow belongs together, it is only split across two links to achieve the higher bandwidth that is needed to transport the double amount of data.

Finally, does it make sense at all? No, not really. If you want a consistent quality in the studio of whatever sample rate, then you should / need to use the same sample rate on all devices.
You only have to handle such things if certain devices have a fixed sample rate, and you do not want this sample rate and also not this device as clock master.

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

15 (edited by Kubrak 2024-11-17 17:23:28)

Re: Downsampling ADAT clock rate only

I meant that downsampling in this case does not require complex math, using every second sample would do (IMHO). That probably would be relatively simple to implement in DSP part.

I did not say it would be easy or even possible further on the way to ADAT/SPDIF. Probably it would be neither easy nor possible in current UCX II HW.

And RME is to decide, if it is "cool" feature to implement in future designs and if it is financialy feasible at the same time....

It would have use cases. For example older HW effects may run digitally only on 44/48 kHz. Even current state of art HW units like Eventide H9000 costing about 8 000 EUR can handle only 48 kHz.... (Better said were able till quite recently, now it does DS on all algorithms.). But one maybe does not want just because of that run everything on SS. OK, one could use analog input of those devices, but that would mean two A/D and two D/A for each effect. And while digital part of those devices may be still good A/D and D/A may be weaker part of design.... Mainly for devices few decades old.

IMHO, feature to downsample to half (quatrer) and upsample to double (quadruple) would ease integration of old HW using digital signal path.

Re: Downsampling ADAT clock rate only

I recently experienced some similar clocking issues when using an iPad connected to an UCX which is connected via ADAT to the main interface which gives the clock (UFX+).
No matter what I set in Total Mix on iPadOS, some apps are locked to 44.1 kHz which results in glitchy sound when using 48kHz on the host (and Total Mix on iPadOS is not showing the sample rate change by the way, so it took a while to figure it out).
Some integrated resampling would be helpful there as well.

17

Re: Downsampling ADAT clock rate only

The change in sample rate can be seen on the UFX+. And the CoreAudio in iOS can do resampling. IMHO here only the apps that only support 44.1 kHz are to blame.

@ ramses:

I never heard of multi-port SRC for 8 channels

Not sure the ADI-192 DD qualifies here, but it indeed had 8 channels input SRC to handle ADAT signals. One of my favourite gadgets that can be a life saver with all its options. Unfortunately demand was so low that we had to cancel it...which also answers the request in this thread - no way to waste time and resources on such an unpopular feature.

Regards
Matthias Carstens
RME

Re: Downsampling ADAT clock rate only

Thanks MC, I will have a look at ADI-192 DD. I have few older HW effects (Eventide and Lexicon) that run only 44/48. SRC on few channels would be handy.

19

Re: Downsampling ADAT clock rate only

MC wrote:

The change in sample rate can be seen on the UFX+

Thanks yes, indeed!
It is just a bit confusing, as it doesn’t display the actual used sample rate on the iPad in this scenario.
The Fireface settings on a Mac behave different in that regard and usually display any sample rate change caused by macOS immediately. I will have to rely on that.