I don't think it would be scaling of the data that is already there in the .txt file, but rather the data that the .txt file was generated from, that has been incorrectly processed in some way. At least that's what Tony @ minidsp seems to be implying.I plotted three graphs using three different calibration files in the rangen 20–800 Hz(ignoring the first sample).
The graphs looked like random noise. I don't think scaling could fix the 20–800 region.
It'll be interesting to see once new calibration files come through whether it's just the 20-800hz range that gets changed or if its the 800+ , or both.
Despite the dodgy calibration file I've managed to use REW + Umik1 + minidsp to turn my horrible boomy, muddy, bloated Audioengine 5+ speakers into delightfully clean sounding "monitors". They had real issues with 100hz-300hz range booming away at +12dB which was easily sorted with REW's automatic EQ correction biquads which I then imported into minidsp:
Apparently the UMIK-1 is pretty accurate in that range anyway even without the cal file, which is probably why I'm hearing such a dramatic improvement in the bass :sn:
One question though, where should the Umik's volume be set in Windows 7 properties? The default is "10" on my laptop. Should I leave it at 10 or should it be at max? Does it make any difference to REW? I feel that my Umik is reporting levels in REW lower than they actually are. For example when the mic says 80dB SPL, it sounds very loud to me. I actually got sore ears from playing that sine sweep at 80dB, and REW kept telling me to increase it to 90dB!