Paid too much money for my mic to sound this bad

[arcaxis]

I just did a test recording with my Behringer X1204USB into Audacity. Used a Behringer B1 mic. On the mixer I preset the channel level at '0' as well as the Main Out at '0'. In the Windows7 audio/sound record properties for the the USB MIC CODEC I set the level to 100 (max). Brought up the gain for the channel so the VU LED's were hitting about three greens (more LED's on the 1204 than the 802). This pushed the VU in Audacity to about -20 to -18dB for normal speaking. Made a recording and normalized it to bring it up to a good level. Reasonably good recording with some background noise which I believe to be the PC's fan and the wife downstairs in the kitchen, otherwise quiet.

Went against the grain with Dave's suggestion to knock down the gain level in Windows properties, but certainly worth playing with a bit to find a good balance. Also look at other software that may interact with the audio, such as something the computer manufacturer may have for the sound system (my ASUS uses Waves Maxx Audio to 'improve' the sound)

Has the OP said if this is on a laptop either battery powered or noisy switchmode wall wart or a desktop that is likely properly grounded. A USB cable with a ferrite choke might be worth a try.

Attached is the MP3 I did and also someones rendition of the USB added to an 802 mixer block diagram.

 

[ecc83]

Morning Mark,

I have attached screen shots* of RTAs of the last 'noise' part of your recording and two plots of my ZED10 mixer the latter just noise, no mic.

The results for the mixer are with Win mic gain at 100% and another at 5%. The spikes are still there but some 27dB below the 100% version and in practice the noise floor is about -85dBfs. The A&H result is in fact cleaner than I remember it but then I have not done this test for quite a while.

The spikes are at different frequencies for the two mixers (slight clock difference in the two converter's? "Don't get technical with ME Artoo!") but the trend is the same, such a noise plot of any 'decent' AI would give a smooth curve with barely a ripple.

Can I prevail upon you Mark to do a 'silent' recording? Sans any input so we can see the mixer's noise floor?

I am now going to drag my zed10 into my living room to test this hub business (the fings I do for.....!)

*NB the noise floor of the mixer is NOT -120dB! That is a feature of Right Mark Analyser. I guess there is a way to 'calibrate' it but I can't be arsed atmo. I just look at the pictures!

Dave.

 

[ecc83]

All the previous plots including the mic waffle, were done through a hub* but with no external power (could not be a'ed to find the rat) but the plot attached now is with a 5V supply connected.

The trace has cleaned up slightly but there is still a spike at about 1kHz.

Dave.