1 Topic by ramses (edited by ramses 2021-05-26 11:17:20)

  • ramses
  • DAW Mastermind
  • Offline
  • From: UTC+1
  • Registered: 2012-10-14
  • Posts: 13,231

Topic: XTC and 12Mic on a MADI bus

Because of this thread I question myself, what a "best practise setup" from RME perspecitive.
https://forum.rme-audio.de/viewtopic.ph … 00#p171000

[ optimized for yes/no replies as a time saver for you ]

Is in a serial MADI setup the order of devices important, when having a mix of MADI devices
- with and without extended MADI features (Auto-ID, Auto-CA) and
- which follow or break the 8-port scheme ?

Given example setup: 3x XTC, 1x12Mic in a serial setup.

What is the best setup / order of devices for a "best practise" approach with amen from RME ?

Is it correctly assumed that

1. delay compensation still would be useful for the three XTC so that they stay sample synchron to each other ?
2. the three XTC need to be in one block cabled (not with a 12Mic between them) for delay compensation to work:
    "12Mic-XTC-XTC-XTC" or "XTC-XTC-XTC-12Mic", but NOT XTC-12Mic-XTC-XTC ?
3. channel assignmet/mapping on a MADI device is independend from it's position in the serial MADI setup ?

Would this be a valid setup and what would happen: 12Mic-XTC-XTC-XTC, Auto-CA enabled on all XTCs.
I assume that:
4. XTC1 - XTC3 get MADI ID 1-3 for Delay Compensation
5. Auto Channel Assignment would map them to MADI channels 1-8, 9-16, 17-24
6. 12Mic - although being the 1st in the serial MADI setup - can be configured to use MADI channels 25-36

And in terms of delay compensation for the XTC ...
7. The XTCs are sample synchron to each other no matter whether 12Mic is at the front or the end.
8. Delay Synchronisation is not achieveable between XTCs and 12Mic

Although Delay Synchonisation is not achieveable between XTC and 12Mic..
9. Is it possible to minimize audio runtime differences between 12Mic and the "block of 3 synchronized XTCs"
    by putting 12Mic either as 1st or last device on this serial setup ?
10. Or is it the same to either have "12Mic-3xXTC" or "3xXTC-12Mic" ?

Especially on the last questions I am struggeling in terms of the timing, what has most advantages interms of sample synchony.

Would be very kind and useful to get a guidance from RME as more and more customers will have such a mixed setup, many thanks.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub14
  • Registered: 2008-05-07
  • Posts: 131

Re: XTC and 12Mic on a MADI bus

I agree that this would be useful.

My problem, described in the link that ramses posted, was solved. However, I am not completely clear on how it was solved.

  • Max
  • Max
  • Moderator
  • Offline
  • Registered: 2007-05-09
  • Posts: 316

Re: XTC and 12Mic on a MADI bus

ramses wrote:

1. delay compensation still would be useful for the three XTC so that they stay sample synchron to each other ?

yes

ramses wrote:

2. the three XTC need to be in one block cabled (not with a 12Mic between them) for delay compensation to work:
    "12Mic-XTC-XTC-XTC" or "XTC-XTC-XTC-12Mic", but NOT XTC-12Mic-XTC-XTC ?

yes

ramses wrote:

3. channel assignment/mapping on a MADI device is independent from its position in the serial MADI setup?

yes

ramses wrote:

Would this be a valid setup […]: 12Mic-XTC-XTC-XTC, Auto-CA enabled on all XTCs.
I assume that:
4. XTC1 - XTC3 get MADI ID 1-3 for Delay Compensation

It would be a valid set-up. It is correct that they get MADI ID 1-3.

ramses wrote:

5. Auto Channel Assignment would map them to MADI channels 1-8, 9-16, 17-24

yes

ramses wrote:

6. 12Mic - although being the 1st in the serial MADI setup - can be configured to use MADI channels 25-36

yes

ramses wrote:

And in terms of delay compensation for the XTC ...
7. The XTCs are sample synchron to each other no matter whether 12Mic is at the front or the end.

yes

ramses wrote:

8. Delay Synchronisation is not achieveable between XTCs and 12Mic

Delay compensation within a daisy-chain of different types of products is not achievable unless manually configured with input delays in the DAW.
It is possible when using AVB (see below).

ramses wrote:

9. Is it possible to minimize audio runtime differences between 12Mic and the "block of 3 synchronized XTCs"
    by putting 12Mic either as 1st or last device on this serial setup ?

yes, as the first device. A simultaneous impulse on all microphone input channels would show a delay of nine samples on the XTC inputs when compared to the 12Mic inputs.

ramses wrote:

10. Or is it the same to either have "12Mic-3xXTC" or "3xXTC-12Mic" ?

if the setup is turned around, the impulse would be recorded 16 samples later from the XTC inputs than the 12Mic inputs, so the other order is better.

When using AVB, one would connect the three XTCs directly to the 12Mic MADI input with Auto CA, and route those three blocks of eight channels to three outgoing 8 Channel streams with a presentation time offset of (for example) 1 ms. The 12Mic inputs would be sent to a separate stream with a presentation time offset of 1 ms + (6 + number of XTCs in chain * 3)/samplerate. As a result, in an AVB network, all receivers of the four streams would receive the analog input signals from all four devices perfectly delay compensated and sample accurate. Nice feature of AVB Presentation Time Offset.

4 Reply by ramses (edited by ramses 2021-06-02 06:22:57)

  • ramses
  • DAW Mastermind
  • Offline
  • From: UTC+1
  • Registered: 2012-10-14
  • Posts: 13,231

Re: XTC and 12Mic on a MADI bus

Many thanks Max for answering, tbh .. esp the last two points I have still to "digest". It's not fully clear to me yet, why the different delays in regards to the order of devices. Maybe I need something like a drawing / calculation.

Is it right to assume, that audio signals from MADI / such a serial chain of devices arrive later at the DAW compared to the rest of audio signals like e.g. an analog port on UFX+ ?

So the RTL reported by ASIO driver to DAW can only cover those ports on an UFX+ where you have a known converter directly attached to the device with a known converter latency (analog port = mic/line/instr inputs), correct ?

I question myself, whether enabling delay compensation is the way to go so that you have at least all devices on MADI more or less sample synchron to even allow plugging of stereo Mics to delay compensated units in a larger setup.

Or does delay compensation introduces even more delay for some of the devices on the serial MADI chain, so that it would be "faster" = less delay, when turning delay compensation off in the examples above ?

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