RTL Utility or Cakewalk. Which is truth.

[Michael Fogarty]

Hi. Just doing tests to see whether the RME or MOTU Mk4 has better latency. Cakewalk reads them both, but the RTL utility won't give me a reading on the Babyface. Says it might be up to 50% off, do it again, and then a different prompt shows up. I am testing at 96buffer. Both will record down at the low levels but it seems the Babyface likes it better. I usually record at 128 once things are up and running and tracks are filling up and only need these low latencies (96 or 64) when I am comping digital pianos. Not worrying about the RTL utility here, can I pretty much bet that the onboard Cakewalk utility is accurate? Why is it when something is stand alone or only does one thing we give it more credence?

Edited March 23, 2021 by Michael Fogarty
sp

[Matthew Sorrels]

Cakewalk doesn't measure anything.  It's reporting the setting by taking the buffer size and using the project's sample rate to compute the latency.  RTL measures the actual round trip time of a generated audio signal (that you feed back into your inputs).  This can vary from the setting because of how your audio interface works (and your computer and your OS and the drivers, etc).

Important to point out that Cakewalk has no way of knowing if you can actually get that level of performance without drop outs or the audio engine stalling.  It just reports the setting.  RTL gives you a way to test and see if that rate can be actually used.  But just because your audio interface/computer can run a low buffer rate that doesn't mean you won't have to increase it a lot when you start increasing the processing on the signal (plugins/soft synths/effects).

Edited March 23, 2021 by Matthew Sorrels