Topic: Nothing to report but satisfaction

Re-installed XP recently and took the opportunity to fit a used 9632 side to side with my current 9652, all fine
The 9632 runs for Windows WDM main audio interface - listening for my favorite radio streaming - and also my K7 tapes / vinyls digitazing with Wavelab
while the 9652 along with my ADI8-QS runs for the overall multichannel layout playback (ASIO) and DAW various activities

I have some questions though:
- Why are the two cards clocks and buffer sizes identically set and correlated when a change is made from one side or another ?
I use the highest sampling rate on the 9652 but don't need that much on the 9632

- What is the limitation for increasing the 9652 buffer size to 8,16K (ASIO) as I do not care of latency but focus on system playback stability with high sampling rate, high channel counts, high CPU load from the VST plugins engaged and so on...
Is that a matter of the FPGA or a preset limitation in the driver, or ?

All the best

HDSP9632, HDSP9652, ADI-8QS, FFUC, Quadmic II

Re: Nothing to report but satisfaction

I've just realized a big english mistake in my title  fryingpan
Understand "Nothing to report BUT satisfaction"

For sure I still get daily satisfaction with all my RME gears cool

That said, a bit of answer to my questions would be nice, because when crackling occurs despite using the highest buffer size makes me wonder if the bottleneck for huge audio transfer rate is on the RME side or at the motherboard, operating system tweak, chipset...
Since my post, I have cleared the problem using a convolution engine approach for emulating my favorites EQs plugin, hence reducing the CPU load from 82% to 35%
I am wondering if the Raydat would allow me more throughput as I have a free PCIe slot

HDSP9632, HDSP9652, ADI-8QS, FFUC, Quadmic II

3

Re: Nothing to report but satisfaction

You know that you can edit your posts? In that mask you will see that you can also change the thead title itself...

Regards
Matthias Carstens
RME

Re: Nothing to report but satisfaction

I knew that already, but cannot access to the title for change, only the text
Anyway, it was not the technical content help expected from you so far

Since my post, I used to look deeper in the HDSP manual and found the reason for the crackles when addressed by AudioMulch my VST host
Was due to the fact that the two cards where clocked separatly (two masters) while mapped into AM, causing this one to loose perfect ASIO syncing after some time (15 ~20 mn), no matter of the CPU load or channel count
Now that the 9632 is slaved to the 9652 by external SPDIF link, the overall setup works impressively good at the maximum sampling rate of 96 kHz / 6 playback channels, and I can reach down to 128 samples buffer size to the two cards or even more  32 samples for the 9652 alone (the additional 9632 stereo channels starts crackling at 64 samples)
So I came up to set a robust and safe 256 samples block allowing unsensitive latency

HDSP9632, HDSP9652, ADI-8QS, FFUC, Quadmic II

Re: Nothing to report but satisfaction

Schubertriad wrote:

I knew that already, but cannot access to the title for change, only the text

Done.

Regards
Daniel Fuchs
RME

Regards
Daniel Fuchs
RME

6

Re: Nothing to report but satisfaction

Hmm, as you started this thread I thought you could do so, when clicking edit on your first (!) post. But maybe only mods can do that...

Regards
Matthias Carstens
RME