You are not logged in. Please login or register.
RME User Forum → FireWire & USB series → Motu 828mk3 Firewire better latency than FF800?
I just switched from a Motu 828mk3 to a FF800 since I always heard so good things about them.
To my surprise the Motu wins the latency stats?! I always been told that RME were superior regarding this?
64 samples, 48khz Ableton Live 11, iMac Pro 2017:
Motu 828mk3 (FW400 Connection)
Input: 2.23 ms
Output: 2.17 ms
Overall Latency: 4.40 ms
Fireface 800 (FW800 connection)
Input: 2.94 ms
Output: 3.33 ms
Overall Latency: 6.27 ms
Fireface 800 from 2004
MOTU 828mkIII from 2011
You need to honor / compare converter latencies.
FF800 is an older design as waedi already told you.
Here some additional information about converter latencies from manual ch 30.2 latency and monitoring at 44.1 kHz:
AD: 43.2 samples = 0.98 ms
DA: 28 samples = 0.63 ms TOTAL=1.61ms
Newer products like the UCX II have the following sample rates:
AD: 5 samples = 0.11 ms
DA: 6 samples = 0.12 ms TOTAL=0.23ms
This alone makes a difference for converter latency of 1.61 - 0.23 = 1.38ms.
Then next point is that both manufacturer might have a different idea about stability.
As you can see in the RME manual, RME adds a safety buffer for Apple's macOS where audio has to pass the sound libraries of Apple. You do not have direct access to the recording hardware like with Windows and ASIO drivers.
At the end of the day it can't be faster than "fast and stable" ...
But there are many other manufacturers / products where the latency is much higher compared to RME (and Motu in this case), up to double the amount of time.
Another advantage of RME compared to Motu, the internal FPGA is performing the I/O and can be re-flashed in case of an error. Motu uses 3rd party communication chips where a such a re-flashing is not possible at all.
RME has the better overall design and is in many details better compared to Motu.
Fireface 800 from 2004
MOTU 828mkIII from 2011
Nope.
The Motu 828 mk3 Firewire is from late 2007.
The one you googled is the "828 MK3 Hybrid".
Anyway... used the 828 since 2009 without any issues. Still works even in Ventura (as the FF800).
The one thing that made me nuts was that the 828 takes 35 secs to boot up!
The FF800 takes 3 secs! Win!
That latency you have measured was a software routing thru the computer, right ?
How about latency for direct monitoring during a recording session ?
The numbers are coming straight from Ableton 11 Audio Preferences. Same setup for both cards.
I never use direct monitoring. Pointless for my workflow. (Tracking mostly external synths, FX, drum machines etc).
Motu 828mk3 have direct monitoring as well. Kind of pointless to compare DM anyway.
Direct monitoring is what it is. Kind of "almost zero latency".
I’m very happy with my UCXII, but nothing wrong with acknowledging that MOTU’s stuff is terrific for Mac users. They’ve been doing it forever and their know-how manifests in their devices.
But MOTU wasn’t that great for Windows users back in the day, don’t know if that has changed
FF800 is an older design as waedi already told you.
Well, Motu still beats "newer" cards with their older 828mk3 2007 card.
64 samples, 48khz
Motu 828mk3 (FW400 Connection)
Overall Latency: 4.40 ms
Fireface 802 (FW800 connection)
Overall Latency: 5.27 ms
Fireface 802 (USB connection)
Overall Latency: 4.5 ms
I think the question here is stability, with how many safety buffers they calculate.
Well, Motu still beats "newer" cards with their older 828mk3 2007 card.
64 samples, 48khz
Motu 828mk3 (FW400 Connection)
Overall Latency: 4.40 msFireface 802 (FW800 connection)
Overall Latency: 7.29 msFireface 802 (USB connection)
Overall Latency: 7.17 ms
That link shows these numbers at 128 samples. Did you read the table wrong?
Also coming back to your first post: I don't believe any numbers from any DAW, as they are typically not measured but just output from the driver's given numbers. These can be wrong. Use RTL Utility for the real numbers.
https://oblique-audio.com/rtl-utility.php
This tool also lets you see how big the Safety buffers are.
Ah. Right. I took the numbers from the wrong row (128). I have now corrected my post.
Will test out the RTL Utility 
Thanks for input.
Getting these numbers measuring with RTL:
64, 48 khz
Motu 828mk3: 5.083
RME FF800: 6.271
Converter chip latencies may play a part here to some degree.
...here's an overview that implies the stability of the interface under different buffer sizes...congrats, rme...:)...these are real world latency values btw, not copied from the manufacturer's website
https://dawbench.com/images/DAWbench%20 … r-2022.pdf
...here's an overview that implies the stability of the interface under different buffer sizes...congrats, rme...:)...these are real world latency values btw, not copied from the manufacturer's website
https://dawbench.com/images/DAWbench%20 … r-2022.pdf
Well, those are only measurements in Windows. My own results for MAC speaks different as shown above.
Could you explain further which figures that: "implies the stability of the interface"?
I have no clue on what to look for regarding this in those tables?
Scheffkoch wrote:...here's an overview that implies the stability of the interface under different buffer sizes...congrats, rme...:)...these are real world latency values btw, not copied from the manufacturer's website
https://dawbench.com/images/DAWbench%20 … r-2022.pdfWell, those are only measurements in Windows. My own results for MAC speaks different as shown above.
Could you explain further which figures that: "implies the stability of the interface"?
I have no clue on what to look for regarding this in those tables?
Here is the associated thread explaining that PDF. The ranking is done with PC (Windows) under load.
RME User Forum → FireWire & USB series → Motu 828mk3 Firewire better latency than FF800?
Powered by PunBB, supported by Informer Technologies, Inc.