Native Instruments Komplete 6 MKII and REW Levels

[rackmonkey]

I know a few folks here use Native Instruments Komplete Mk II sound interfaces with REW. I'm using the 6 channel version and I can't get the measurement levels high enough to make useful measurements. When I go to calibrate using the loopback, I can get my input levels dead even with the generator output level, whether I use the -12 dBFS default - or any level I choose, as you can see in the following image:



But when I go to take a measurement to actually create a .cal file, I get the following message:



I'm using the latest version of REW and the ASIO driver for NI on Windows 8.1. I've calibrated probably close to a dozen different soundcards and interfaces before, but have never run into this problem. I've tried it at every sample rate available with no change in level. It doesn't matter whether I configure Windows to allow applications to take exclusive control of the interface or not, and Windows recording and output level controls have no effect on the levels regardless of the exclusivity setting. I've tried using every combination of input and output. Switching between mic/line and gain settings done solve the problem. Setting the "Output 1/3 Volume" at its highest results in the same thing. I've used TRS->XLR and TRS->TRS for the loopback. I've seen some forum posts on other forums with users having the same issue, but none of the suggestions (all listed here) make any difference. I even tried using the headphone out back to the input and I still get the same results.

I'm starting to think REW and Komplete MKII just don't go together on Windows. Any ideas or anything I'm missing?

 

[ruffrecords]

I have a Komplete Audio 2 which I use with REW. I agree it can be temperamental in getting the past the cal stage. I usually end up flipping the ins and outs a few times until it works. I keep thinking it is me but maybe it is not.

Edit: I just tried it again. I use the Java drivers, set the output to Komplete spekers and the input to Komplete microphone. Below these two set default output and default input. Then it works.

Cheers

Ian

 

[rackmonkey]

Thanks, Ian. The Java drivers seem to be the key. Bummer that the ASIO driver doesn't work well. I didn't even think the Java driver worked on Windows. But it does. Thanks again!

 

[rackmonkey]

The other bummer is that max sample rate with Java on Windows is 48k.

Abby uses one of these on Windows, doesn't he? Wonder how he's getting 96k measurements.

 

[rackmonkey]

Okay it just hit me that since ASIO won't allow the "input device" and "loopback input" to be the same channel on the interface (the Java drivers DO allow them to be one and the same), you have to calibrate levels on the loopback channel, then physically move the input cable over to the input channel prior to the soundcard measurement, being careful to set the input level on the input channel to the same level as the loopback. Pain the arse and not very intuitive, but it is what it is and once I do that I can measure at max sample rate with a CAL file active for the card. Doesn't make much sense but nothing I can't live with.

Thanks for the help Ian!

 

[ruffrecords]

My pleasure. Glad you got it working. I use REW both on Linux (with a Focusrite 2i2) and Windows 10 (with the Komplete). I am pretty sure I measured the input bandwidth of the Komplete at 192Kss with the Java drivers.

Cheers

Ian

 

[abbey road d enfer]

Abby uses one of these on Windows, doesn't he? Wonder how he's getting 96k measurements.

TBH I never considered using anything else than ASIO.
For calibration, I set the generator at its max level (-3dBfs) and set the gain in my interface box to output +24dBu.
Then this is brought to the input of my interface box, set to 40dB attenuation. The actual level reaching the input of the soundcard is then -16dBu.
I do that because the compressor in the input starts to act soon after, which is probably good for recording music, but is a PITN for measurements.
As I mentioned earlier, measurements at 192k SR are accurate to about 90kHz, but the generator is very noisy between 40 kHz and 140kHz due to noise shaping; I had to install an 8th-order filter at 84kHz in the generator path, which allows accurate measurements up to 80kHz.

 

[rackmonkey]

Thanks all. As I related above, I got it all going with ASIO at 192k. A little counterintuitive getting there, but it works.