naturerecordists
[Top] [All Lists]

Re: Re: Measuring EIN

Subject: Re: Re: Measuring EIN
From: "Bruce Wilson" <>
Date: Wed, 5 Apr 2006 19:24:27 -0600
Well, here are my first measurements. There are some puzzlers here.

All are measured with no mic rolloff, level controls turned up all the way,=

with a 150 ohm resistor in the mic input. Resulting recordings were
FFT-filtered to a flat 20-20k passband, and the noise analyzed to calculate=

the average rms power in dB relative to the full-scale power, using a
sine-wave standard.

I'll space-format a table of the results. Hopes it comes out okay.

Recorder            dB-FS     gain     A-EIN-FS      FS       EIN
Marantz 670        -62          48         -110           -8        -118
SD722                -78          70         -148
Sony NHF800     -72          75*        -147
Jukebox 3           -56          48         -104

dB-FS is the 20-20k unweighted noise level in dB relative to full scale.
gain is the gain of the preamp, if known.
A-EIN-FS is the Apparent EIN relative to the Full-Scale level.
FS is the full scale level in dBu (0 dBu=3D0.775v).
EIN is the measured EIN value in dBu.

*I used the value of 75 dB for a different Sony MD recorder for this figure=
.
Sony doesn't say what it really is, so I've used the value Rob mentioned in=

another post (http://groups.yahoo.com/group/naturerecordists/message/21105)=
.

So, I hoped you noticed the problem with the SD722 and the MD. Both are wel=
l
below the theoretical minimum EIN of -136 dBu for a 150 ohm resistor. My
only conclusion is that either my technique sucks, or both must have a full=

scale level that is well in the positive dBu range. To make the SD722 match=

the published specs for this test ("-128 dBu max (-130 dBV), 150 ohm source=
,
20 Hz-20 kHz BW flat filter, gain fully up") the FS level must be +20 dBu.=

Frankly I doubt this, and am uncomfortable with the results.

Thoughts?

At first I thought it was a problem in the way Audition measured levels.
There everything is called "power", but upon reflection I think everything=

we call "levels" really are powers because the old reliable dB=3D20log(v1/v=
2)
is a power form of the dB calculation (based on P=3DV^2/R). The -78 dB for =
the
722 is much lower than it was before the bandwidth was cropped.

By the way, the Jukebox3 has a very funny noise signal: it's really a 90 Hz=

sawtooth waveform with some noise on top of it. Got to be the disk drive.

Bruce Wilson KF7K
http://science.uvsc.edu/wilson




________________________________________________________________________
________________________________________________________________________

<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU