Admin Knut wrote:If seems strange to me that 32 samples with the 96 kHz rate won't work since other companies, for yeas now, has used that configuration to achieve the low roundtrip latency the advertise with.
Which companies achieve this for years?
best regards
Knut
Apogee, for one.. they launched the Symphony card in 2007 and posted that it was usable with 96kHz + 32 buffer. That system also gets reduced latency due to the fact that it's using PCIe and not FireWire. However, since FW seems to add some latency PCIe doesn't, it's even more important to be able to use 96 kHz + 32 buffer on a FW based system.
MC wrote:There is no right and wrong. It always depends on your personal computer and the way you work.
The Mac I use is a 8-core Mac, 2.8 gHz.
If at all I would agree that PCIe still has the edge over FW and USB, but that is a well-known fact.
Sure, and I don't expect a FW based system to offer the same latency as a PCIe based system since FW as such may introduce a millisecond, a half millisecond or whatever to the monitoring roundtrip. The question is if there's any way to make the FF800 work with a Mac at the 96 kHz/32 buffer setting at all. The local distributor recently suggested that I may have a faulty model, so the validity of the statement above ("At 96 kHz a setting of 32 samples equals 16 samples in single speed - which won't work"). If I need something else, that's somehow fine (except of course that the information I've got has been wrong - and that the dealer keeps *not* calling back about my issue in spite of a few promises about doing that, but that's another story which isn't relevant in this thread).
I just need to move on to something that works as expected. If FF800 can't handle this, but other FW or USB interfaces can, I need to look at other possible options, or move on to PCIe or ExpressCard solutions. If 96/32 won't even work on a UFX model and an 8-core Mac), I guess RME have no FW/USB options at all that can be used with my 8-core Mac and 32/96 (?).
Comparisons to others are mostly flawed except you measure the real round-trip latency yourself.
Sure, I've measured a number of interfaces - the problem right now is that we all know that higher sample rates give lower latencies, and that I get pops/clicks with the FF800 at 96 kHz and the 32 buffer. Going up to 64 sample buffers or 44.1 sample rate increases the latency up a level where the things sound a less lot tight. Higher sample rates than 96 are unusable with this buffer setting - but working @ 192 kHz isn't worth it anyway (IMO). I hope you understand that I don't feel a need to send the unit to repair for a week or so if the configuration I use simply isn't one I can expect to work.
And no, I can't 'hear' eg. 1 millisecond, it's just that lower latency provides a 'tighter' performance which is much more inspiring to work with.