Topic: usb performance

I was comparing the performance of my 2 RME devices babyface pro and fireface UC both up to date with the latest drivers and firmware on a razor laptop with Intel(R) Core(TM) i7-8750H CPU @ 2.20GHz 
While running an ableton session at 64 samples with a cpu hungry setting i noticed that using Asio4ALL drivers and the internal Realtek audio I got better performance (no clicks and pops) than with either of the RME devices. I was not using a usb hub and I tried different usb cables. I tried different usb ports on the computer as well. With ASIO4ALL the cpu meter in ableton could climb up to 60% without any clicks or pops or dropouts. With the RME devices once the CPU was past 30% i would get clicks amd some dropouts. I have optimized the computer for audio. Is this normal? to have superior performance with an internal soundcard?
is this a problem with usb? any suggestions for optimization?

2

Re: usb performance

When using ASIO4All you are NOT using 64 samples.

The UC is not only USB, but has more channels than the Realtek, so again comparison is not valid.

For optimization tips search this forum for posts from user ramses.

Regards
Matthias Carstens
RME

Re: usb performance

my intention here is not to say that asio4all is better than RME. my point is i get less crackling with asio4all when my computer is working hard. i know that rme has better sound quality than realtek internal sound cards and i also know that there are proper multi channel outputs and inputs with RME. to say that asio4all cannot run at 64 samples i dont think is true. it is 64 samples of latency reported by ableton in the preferences and i can tell the feel of playing a softsynth and it does feel like 64. i will certainly check the optimization again. i was looking for constructive help so i can achieve possibly the same clean audio performance from the RME gear with the same latency and workload. Maybe it is not possible.

Re: usb performance

It's worth checking the ASIO4ALL setting, as they include extra buffers and buffer offset values. You need to zero those to get a fair(er) comparison.

Sorry if you already knew that!

Re: usb performance

thanks for that bit of info. buffer offset was at 4 ms. i will test again at zero.