2 Cards = Delays: RME-REAPER-ME? or?

[tweed]

point: RME has NEW Hdsp drivers AND firmware. [3.02]


TOP LINE:
I get brilliant 'lack of latency' response. with EITHER RME card's inputs and outputs set as Audio device range...

with BOTH RME cards [even just a pair of outputs on second card] I get 200 ms plus POOR latency.

NOTE: haven't tried much in the way of experimentation. Slot changes/reversing cards/downgrading drivers/etc.

NOTE 2: Can't say WHETHER RME firmware and driver update recently done to both cards has anything to do with issue of 'either' card seperately getting NO latency, YET, using both TOGETHER get's huge latency. [I did the updating Before trying both cards together] This is BIG RME update. [HDSP 3.02] VERY different drivers they say.

I use an RME multiface 2 mostly, and then I said...hey why not plug my RME 9632 stereo card in as well. That way I can move my PC to another room or place and not have to take the multiface II box and cable and stuff. [while still having RME drivers at work! ]

So: in the past whenever I've used TWO RME cards I absolutely HAD to workclock sync them as no manner of spDif or internal card to card 'jumper cabling' would work.. I'd get the weirdest stuff....like after 5 minutes there'd be a creeeping distortion begin that would eventually overtake all audio...
[This was back in my NUENDUH 3 days.... hehehe]

Anyways... I mention that as maybe it's a similar thing... ALSO: I should mention that both cards [2 days ago!] are now firmware updated to work with the NEW version 3.02 of RME's HDSP drivers...


So So SO: I can select the ENTIRE Multiface input and output list ][it shows up first in REAPER's Audio device list... YET. if I include even the first pair of the 9632 card I get about 200 ms of delay in resultant realtime played audio...


I have NOT tested just the 9632 [leaving the multiface II outa the list]

Could this be that they're NOT wordclocked? [don't have the WC add-on for the 9632--- yet]

Or, could it be RME's new drivers?

Or could it be that Reaper is doing something ...like compensating or something...

or a combination of stuff???

Gonna test JUST the 9632 now..


[BACK from testing for delays with JUSt the 9632 input-output range selected]
Sweet... no delays...

The latency is amazing...[READ: VERY low] I'm set at 6 ms. In Nuendo 6 ms with the RME drivers used to feel a wee bit sluggish [delay was just noticable]

ALL good with REAPER...


expect the fact the at the moment I cannot use both cards, without a delay ... [IE. >>> when setting REAPER's Audio devices range to include BOTH cards I/O range]


[FACT IS, I'm NOT even sure 2 cards is the most logic use of ..um, a lotta stuff]

I will eventually find out what's up.

1) I missed something?

2) The new RME drives are causing it? [suppose I could downgrade to 2.94 though fireware is not upgraded as well.. so dunno as obvervations wouldn't be all that comprehensive]

3) Reaper is aware that the two cards are NOT truly synced and is ...erm.. letting me know.. ??? cool!

[MlewisDE]

I don't know of any app that can simultaneously address 2 ASIO audio cards... I'm surprised that reaper even lets you see both of them. RME makes a point of saying that even 2 HDSP cards are separate entities (because totalmix is part of the box, not in the driver). The latency you are experiencing may be because reaper is physically shifting audio streams from one card to the other.

In cubendo, you have to select the ASIO driver you want and then it restricts you to those IO. Doesn't reaper do that? Is that design or omission?

[pipelineaudio]

normally RME cards of the same type are seen as one big card

It works fine in reaper with the older drivers.

[MlewisDE]

Thanks, I wasn't aware of that, but then I've never put 2 RME cards in 1 comp (except a pci card and a fireface, where I get asked to select the ASIO driver in nuendo).

Is this specific to reaper? RME says their drivers moved from MME to WDM. Do you think that is the reason for the delay?

[tweed]

just found this thread again... actually the delay is TOTALLY gone now... using BOTH cards.... and latest updated from before RME NOEW WDM drivers [03.07-ish]


Anyways, who knows what really happened... though i'd probably get the wordclock module for the 9632 as I'm pretty sure I'd have a better latency before crackles that way.... right now I'm syncing with 2 ADAT cables though there wasnb't a delay without the ADAT cable though I think there's less crackles WITH the adat cable sync..

So... yeah, in cubase UNLESS I used workclock with 2 RIME cards I'd get the weird increasing distorting thing...