ARTA

[b_force]

please allow display of relative harmonic level in dBr instead of %

Please both, not instead please.

I don't like dBr (at all), but I know other people do

 

[JMB]

Good morning.
I would respectfully ask for the ability to leave the legend up on a graph for printing and review. I find that when comparing overlays, I have to repeatedly refer back to the legend to remember what each is representing. Thank you very much.

 

[SMABB]

Dear community,

I have bought the MOTU M4 as it comes highly praised on several forums.

When I make the initial tests of the sound card using loop cables, I simply don’t understand a thing.
Much to my surprise I find a THD+N level of 0.23 % for a -3 dBFS,, and if I investigate the sound cards frequency response, I also end up with something, that simply cannot be right.
Attached you can see the frequency graph in loopback mode. It is taken as a simple IR and converted to FR.

Setup on the M4 is loop cables out of Monitor 1-2 on backside of the M4 and into Input 1 and 2 on the front. Monitor buttons are not activated and Input Monitor Mix potmeter is turned all counterclockwise to "Input". ARTA box is not in the loop.

Any ideas on the correct setup for the M4 to work as intended?

 

[DcibeL]

Turn monitor mix knob the other way.

 

[SMABB]

Thanks - much appreciated.

I just did, but result is the same for the frequency response measurement.
I should note, that I have unticked the dual channel measurement mode as per description by @ente in his handbook 2.4

 

[DcibeL]

Ok, so loopback measurement is as simple as it gets, so your result is quite confusing. Perhaps a bit more detail is required.

Loopback cable is TRS type cable, from monitor out 1 to input 1 (left)? ARTA should be set to use ASIO driver in the preference, like this:

For measurement, you can use impulse response type of measurement, or realtime measurement "FR1". Disable frequency response compensation and measure.
Impulse response measurement window for sine sweep:

For impulse response, no gate applied, response should look similar to this:

 

[SMABB]

Thanks for taking the time to help figuring this out @DcibeL

I confirm that ASIO drivers are set to 1-2 output and 1-2 as input.
Frequency response compensation of microphone is disabled.
Input Monitor Mix is set to full Playback.

The result is the same as before.

 

[SMABB]

MOTU M4 settings as of current.

 

[SMABB]

If I use output 3 and 4 and loop back into inputs 3 and 4, I get the expected flat frequency curve.
I still have a rather high THD+N as before.
As expected this setup gives me no way to control levels on output and input.

 

[DcibeL]

Same result on both left and right front input ? My next comment would be to check your cabling, however I assume you've used the same cabling for input 3/4 test above.

Unrelated to your problem, but sweep FFT length is short, as is window FFT length. Increasing will provide higher resolution of frequency response, especially at low freq.

 

[Lojzek]

I would increase sampling rate.

 

[SMABB]

Same result on using right channel.
Yes - I confirm cables are the same.

I fully agree on the time window; these settings are just to get the measurement setup dialed in.

 

[SMABB]

Thanks @Lojzek

I have just made new measurement using sample rate of 96K and sequence length 128K

 

[DcibeL]

Get your RMA form ready...

FWIW If I set up my M4 the same way you have, input is clipping at -3dB output. It's hard to see your actual input dBFS since you are showing as SPL scale. In any case, just as above, unrelated to your FR problem, in fact clipped input should only be flatter

 

[SMABB]

 

[SMABB]

If I use outputs 3 an 4 and inputs 3 and 4 I will get a flat frequency curve as shown earlier in post #1,409

Even though I cannot make any adjustments to the levels, if I test the M4 with these outputs and inputs then I get a nice low THD+N.
Signal when estimating is some -10 dB lower than required, so not really representative.

 

[SMABB]

One question out of curiosity (and trying to understand):

When I use outputs 1 and 2 and turn the large monitor knob to the left of the display, shouldn´t the meters react?
Mine stays as if there is full output on 1 and 2. I can toggle the Monitor knob without any effect.

If I measure output voltage it is reacting to the adjustments, but meters are not..

 

[DcibeL]

Meters show signal level as digital dBFS level, volume knob is post-output. It it this way so the display meter shows useful information regardless of volume setting, and without requiring post-output feedback just for meter display.

 

[SMABB]

Thank you for explaining @DcibeL

- and thanks a lot for trying helping me out. The conclusion is, that channels 1 and 2 are defect, but 3 and 4 are working properly.

 

[blmn]

Motu are very good on specs: Ultralite MK5 Loopback bal_out>bal_in