AVS Forum banner

REW and UMIK gain in Windows oddity

2K views 6 replies 2 participants last post by  alextr75 
#1 ·
While working on trying to setup my two UM18s with the iNuke6000DSP I noticed something strange, that I was hoping if anyone might have an idea what is causing them.

My UMIK is the one that comes with 18dB gain (or so it shows in Windows).

I was using the ASIO driver in REW

First - I setup the UMIK mic level in the Windows recording properties to 10% (~5dB) as per the Guide to REW document I followed. While testing with this, I decided to try to play a Pink Noise with the Generator at -30dBFS and randomly raised my AVR volume to a loud level but nothing specific, and wanted to compare REW SPL Meter (dBC slow SPL ) readings with my external SPL meter I bought from The Source. The reading in REW were about 10dB different. Playing with the Gain slider in Windows did not seem to have any effect on what the REW SPL was showing.

I then opened the logger from the REW SPL meter and what I noticed is that the logger does seem to be affected by the gain setting from Windows, e.g. I had to change the gain to about +11.3 dB in order to match the REW SPL Metter Logger readings with my external SPL meter, but the reading on the REW SPL screen (not the logger) were still the same (about 10 dB off).

To summarize:
- I had to bump the mic gain in Windows to about +11.3 dB to get the same reading on my external SPL meter as the one showing in the REW SPL Logger screen. Is this normal .. the guide mentions +5 dB, but maybe my UMIK comes with a different gain ?

- Even when I get the REW SPL Logger graph to match the reading on my external SPL, the REW SPL main reading screen was still showing something lower (even though the logger wasn't). I have no idea why they would be different ?
 
See less See more
#2 ·
I am even slowly starting to question whether I have a proper calibration file.

I ordered straight from miniDSP's web site. This is what it says and how I downloaded the calibration file:

"To download the calibration files, look for the 7 digit Serial Number found on the body of the microphone."

Well .. I tried 701-2345, but do you think this is what that says ?

I do not know if maybe contacting support and giving them my order confirmation info, if they will be able to help.

Maybe I should have ordered from Cross Spectrum .. it was just cheaper straight from miniDSP .. sigh



 
#3 · (Edited)
Try using Pink Noise 100Hz to 5kHz (it will be more-accurate with a C-weighted profile meter), or Pink PN.

The SPL logger uses Leq which is a different formula for calculating the SPL, the RTA seems to be using another slightly different method as well, the SPL meter uses dbC(F/S) formula.

I use +1.1db.
Mine is a CSL UMIK (which probably helps a lot.)



Mine is within +-1db of my CM-140 meter and TermLab.


Mine isn't miles off, it's all within a very tight range.
If you trust your external mic, you could adjust the UMIK calc file's sens factor to match it closer instead of using the windows mic volume, you'll have to reload the calc file or restart REW for the change to be applied.
 

Attachments

#4 ·
I emailed Support @ miniDSP, so hoping they will be able to help, will have to wait for now.

If not maybe I can try to remove the sensitivity parameter and calibrate using my external SPL meter.

Whether I should trust my SPL mic, not really sure, how would I know whether to trust it :)

It is this one:

http://www.amazon.com/Scosche-SPL10...TF8&qid=1454831454&sr=8-12&keywords=spl+meter

If I calibrate using my external SPL meter (and removing the sens param first), does it matter at all what the Level should be set in Mic's properties in Windows Recording Devices ?
 
#6 ·
how would I know whether to trust it :)
You can never really know for sure.
The Galaxy meters have a calibrator (sold separately here, for those that are paranoid about true accuracy.)
http://www.amazon.com/CM-C200-Calibrator-Meter-Audio-Metering/dp/B0010H1PTC

This one is cheaper:
http://www.amazon.com/gp/product/B008S0OVR2/
You might be able to find a used one on ebay or alibaba.

You could get this one which is solid-state, no moving parts, never falls out of calibration (factory set.)
http://spllabusa.com/us/products/bass-meters/usb-bass-meter.html
It's the same as a termlab (more or less.)
 
#5 ·
Technically it should be 0db boost on the control panel, but my windows doesn't allow it for this device for some reason.

Anything over 0db is digital boost, it's like digital zoom on a camera, it's a horrible way of zooming. You limit the dynamic range of the mic and it clips artificially that much lower. I believe miniDSP sets the default dip switches inside the UMIK to 18db gain which has a clipping point of ~132-18 or 114db peak, but mine was clipping at ~107db with RMS sinewaves.

If you add 10db of boost the noisefloor will go from 30-40db to 40-50db and you'll clip at 97db instead of 107db.

I recommend setting the UMIK dip switches to 12db as that allows 120db test signals free from clipping, which is high enough for testing THX level.
 
#7 ·
How is the CM-140 bought from Cross Spectrum ? Says they verify calibration, so maybe it will be more trustworthy ? I believe its $125 + Shipping, so more acceptable than > $300.

I did spent today measuring with +1.1 dB Windows gain, could not set it any lower using the slider, one more step to the left seemed to set it to a negative gain already.

I am not sure I trust my SPL metter actually, I did auto AVR calibration using the AVR mic's. That should set the channels to 75dB. When I verified the levels using my SPL meter it showed around 68dB for each ?! Checking with the UMIK and REW I could see around 78dB.

So based solely on this, I would only guess the UMIK is still more accurate than the SPL meter.

I would have thought that my SPL meter would be more accurate than the AVR mic, but that may not be the case.

I also tried to open the UMIK to change gain to 12dB, but could not find a way to open it. There are those two small holes, and I tried some needles etc .. but could not open it. Just gave up at the end, I haven't measured high enough yet to get it clipping.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top