Hi there,
I've seen self-noise figures for the PCM-M10 of EIN = -122dB (A weighted), for example at the following:
http://www.theatreofnoise.com/2009/11/choosing-audio-recorder-for-ultimate.htmlI'm trying to measure this noise floor myself. Now forgive me if this is a silly question, but whenever I record silence with the mic (e.g. with an external battery mic with the power off, or a stereo cable with nothing connected to the other end) I get a wav file (recording at 24/96) that has a noise level of -93dBFS at best when I process in Audacity using the stats.ny plug in (available
here). This best-case scenario occurs when I record at level 0 on mic or line input.
Can anyone tell me whether I could expect better than this? Is this really the EIN of the device at zero gain?
I know that as EIN is the noise floor of the recording + the analog gain of the recording.
So as the gain value increases, the noise floor won't increase as much so the EIN would be better at higher gain values. I was just expecting a better EIN than I'm seeing at lower gain values (e.g zero gain).
Specifically, I was hoping to find that the noise floor of any recorded file, whatever the gain value, to be better than the -93dBFS I'm measuring. Otherwise there would presumably be no point in recording anything at 24-bit resolution as this noise level is already higher than the quantisation noise of 16-bit recording (-96dBFS)
Have I missed something?