• Registered: 2007-05-03
  • Posts: 18

Topic: MADI and AES no god mix.

Hi
Finnaly a forum for us rme fans!

But its not all well in the rme world.
After struggling for month wid mysterious clicks and popps, unstabillity and WERY high buffer settings, it all came down to that the following sentence is not correct!

"Up to 3 HDSP MADI and HDSP AES-32 can be used simultaneously. With this up to 192 inputs and 192 outputs are available at 48 kHz, 96 inputs and 96 outputs at 96 kHz".

There is a weird sync begavor when using one MADI and one AES-32 card, it simply not working. Audio sync suddenly dissapear on the AES card, there are clicks, unstability. Sometimes it works for an houer or so, but then suddenly....no sync.....and this no matter how much free prossesor power you have.

As far as I know rme has rekognised this as a bugg.

Stay away from mixing MADI and AES on the same machine until rme confirm that this is adressed.

  • From: Oslo
  • Registered: 2007-05-01
  • Posts: 114

Re: MADI and AES no god mix.

I too have the same problem. But I do not know if this is confirmed as a bug yet.
Anyone from RME like to comment?

psvennevig's Website

  • From: Washington DC
  • Registered: 2007-05-02
  • Posts: 17

Re: MADI and AES no god mix.

Hello

I run one PCI MADI and one PCI AES. MADI in slot one (1), AES in Slot two (2) (wished I had swapped them but too many project have gone to change now). Audio record flow is 32 Mic pre's to 32 AD's > to ADI648 > to MADI In > to Nuendo. Nuendo 32 tracks Out > to MADI out > to ADI648 into DM2000 for 32 track playback. The AES card also connected to the DM20000 via XLR patchbay for monitoring, mixing, mastering and also Aux sends and returns. Set up for stereo or/and 5.1 monitoring.

Clocking: Big Ben feeds the ADI648, the DM20000, and AD's. MADI card slaves from the ADI648, AES card slaves from the WC out of the MADI card.

This setup is rock solid and incredibly great for 48k or 96k operation (one button switchable at the Big Ben).

Hope this info helps
Fox

Win7x64.Nuendo.Samplitude.AES32 PCIe.MADI PCI.ADI 648.Apogee Big Ben.DM2000.Apogee 16XAD.RME ADI 8s

Jim Fox's Website

  • Registered: 2007-05-03
  • Posts: 18

Re: MADI and AES no god mix.

Hi.
Our setup is Dell Precision 690, Dual Xeon CPU
MADI, AES32, UAD 1, Powercore cards, MADI as clockmaster. I have 3 setups like this.

This setup is unstable and loses sync from time to time. I have wasted more or less 3 month trying to work with this setup. I get click when VST performace is 30% (Cubase) and buffer is 256 samples. There where a lot of small "peaks" in the VST performans meter.
Programs like Reaktor couldent be used as standalone as Reaktor had difficoulty finding all the 80 ASIO busses (dident help seting outputs to not connected).
It worked from time to time but then.....:-(

The solution was removing the AES card and set the PCI card to be master clock. Its now rock solid!

I have heard that Madi and aes card works together, as jim says, but not on our PC, and not when using a lot of softsynths!
We dident have many problems recording audio but a lot of problems when using VST pluggs

I can now run several Reaktor VST pluggs and a VST performance (of 70-75% in Cubase) and a buffer size of 32 and almost no clicks. 64 and its good! 128, perfect!

So all the softsynth guys, be warned.

P.S
psvennevig its me mats :-)

  • From: Germany
  • Registered: 2007-04-10
  • Posts: 9,806

Re: MADI and AES no god mix.

Hi,

stepside wrote:

Hi.
Our setup is Dell Precision 690, Dual Xeon CPU
MADI, AES32, UAD 1, Powercore cards, MADI as clockmaster.

I'm not aware of a specific problem between the MADI and AES cards (but that doesn't mean that there isn't such a possibility), but this particular setup has a lot of PCI devices and traffic... What happens when you remove the PC or UAD cards for a test?

Regards,
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

RME Support's Website

  • Registered: 2007-05-02
  • Posts: 62

Re: MADI and AES no god mix.

stepside wrote:

Hi.
Our setup is Dell Precision 690, Dual Xeon CPU
MADI, AES32, UAD 1, Powercore cards, MADI as clockmaster. I have 3 setups like this.

This setup is unstable and loses sync from time to time. I have wasted more or less 3 month trying to work with this setup. I get click when VST performace is 30% (Cubase) and buffer is 256 samples. There where a lot of small "peaks" in the VST performans meter.
Programs like Reaktor couldent be used as standalone as Reaktor had difficoulty finding all the 80 ASIO busses (dident help seting outputs to not connected).
It worked from time to time but then.....:-(

The solution was removing the AES card and set the PCI card to be master clock. Its now rock solid!

I have heard that Madi and aes card works together, as jim says, but not on our PC, and not when using a lot of softsynths!
We dident have many problems recording audio but a lot of problems when using VST pluggs

I can now run several Reaktor VST pluggs and a VST performance (of 70-75% in Cubase) and a buffer size of 32 and almost no clicks. 64 and its good! 128, perfect!

So all the softsynth guys, be warned.

P.S

svennevig its me mats :-)

I would suggest using a PCI expansion chassis for your UAD-1 and POCO cards and possibly another one for the RME cards in order to be able to aviod any semblance of IRQ sharing with the DSP cards, the RME cards (and the AGP if that's what you're using). I am currently running 2 x HDSP 9652's and a Multiface PCI in a Magma 13 slot chassis interfacing with one mobo PCI slot and 4 x UAD-1 cards in a second 13 slot Magma interfacing with a second mobo PCI slot. This provides plenty of bandwidth and much more stability and IRQ configuration options than using the motherboard slots. I have tried many permutations of PCI cards-to PCI slots on many different mobos over the years. Using two 13 slot Magmas to gang certain cards on a common bus that share common IRQ's has been, by far, the friendliest solution I've found.

HDSPe MADI-FX, M-32 AD, M-32 DA, ADI-8 QS, ADI 192 DD, ADI 648, Mytek Stereo 96 AD/DA
  • From: Oslo
  • Registered: 2007-05-01
  • Posts: 114

Re: MADI and AES no god mix.

Any news from RME here?

P

psvennevig's Website

  • From: Oslo
  • Registered: 2007-05-01
  • Posts: 114

Re: MADI and AES no god mix.

RME Support wrote:

Hi,

stepside wrote:

Hi.
Our setup is Dell Precision 690, Dual Xeon CPU
MADI, AES32, UAD 1, Powercore cards, MADI as clockmaster.

I'm not aware of a specific problem between the MADI and AES cards (but that doesn't mean that there isn't such a possibility), but this particular setup has a lot of PCI devices and traffic... What happens when you remove the PC or UAD cards for a test?

Regards,
Daniel Fuchs
RME

Daniel,
MC knows a lot about this case. Talk to him!
We?ve tried removing the PoCo and UAD cards, no difference at all.
The only cure (at least in this computer) is to remove the AES32 card.
MC are looking at a solution, but maybe there is none!

Anyways! I think it is very important the combo of several RME cards are tested by RME in new HOT computers. This Dell P690 for instance have NO problem running a ProTools HD3 system, and a HD3 is loads of PCI load.

If there is a small quirk with the combo MADI and AES32 card, it will be a great thing to find it now before the PCIe market sky rockets and people start to combo many PCIe cards.
Hopefully this quirk (bug) do not exist at all with the PCIe versions.

As the other poster said; We?ve been testing for months with this problem (MC knows all about it), and it isn?t working like it should. What happens is that the ASIO system looses internal sync and start to break up. This even if we are running only 6 channels output and NO inputs.
The problems is easy to provoke with softhsynths, very hard to provoke with just loading ASIO channels and record.
The problems occurs with Cubase4, Nuendo3, Reaktor5 and Live 6, so it isn?t only a Steinberg bug :-)

Please look more into this problem.

P?l Svennevig

psvennevig's Website

  • Registered: 2007-07-28
  • Posts: 15

Re: MADI and AES no god mix.

Hi,

as all seem to agree it is most likely to be a HW compatibility issue. I have had such a thing with a single MADI card two years ago in a Shuttle PC. Same phenomenons: Sync lost once I produced some load.

One question: is it possible to activate/deactivate devices in Nuendo, i.e. all MADI card devices, all AES devices to see if the issue is related to the size of the ASIO loop? In Sequoia, I knew how to do that, but not in Cubendo.

For a solution, you also might want to try installing the newest Intel chipset drivers.

Greeting from Vienna,

Sebastian

  • From: Germany
  • Registered: 2007-04-10
  • Posts: 9,806

Re: MADI and AES no god mix.

Sebastian.Gabler wrote:

One question: is it possible to activate/deactivate devices in Nuendo, i.e. all MADI card devices, all AES devices to see if the issue is related to the size of the ASIO loop? In Sequoia, I knew how to do that, but not in Cubendo.

You can do so here (picture nicked from the internet...):
http://eulipia.net/Projects/Audio/VST_Inputs.JPG

Regards
Daniel Fuchs
RME

RME Support's Website

  • Registered: 2007-05-02
  • Posts: 62

Re: MADI and AES no god mix.

Just wondering if there was any resolution to this issue. I am getting ready to install a MADI and AES 32 in my computer

HDSPe MADI-FX, M-32 AD, M-32 DA, ADI-8 QS, ADI 192 DD, ADI 648, Mytek Stereo 96 AD/DA