At 6:04 PM +0200 6/20/05, wrote:
>I just made a quick comparison between the new Marantz PMD 671 (that is
>capable of recording at 24bit/96kHz) and the SoundDevices 722 recorders in
>conjunction with a pair of Rhode NT1-A microphones (sensitivity: 25 mV/Pa,
>equivalent noise: 5dBA).
>Inspired by Rob Danielson's test arrangement, I placed the two microphones
>on the desk in my office (microphone spacing about 10 inches). There was a
>ticking alarm clock at a distance of about one meter and a small tweeter
>producing a constant reference tone of 1 kHz.
>I first recorded the reference tone. I then switched the speaker off and
>subsequently replaced the left microphone by a resistor of 180 ohms
>connected to the pins 2 and 3 of the XLR connector (and additionally two 6=
>resistors between pins 1 - 2 and 1 - 3 in order to draw a current from the
>phantom power supply circuit). I did this procedure with both recorders
>under exactly the same conditions. Unfortunately, there was some additiona=
>ambient noise (hum and distant jets).
>The recording settings were as follows:
>- sample rate: 44.1 kHz
>- bit depth: 16 bit (dithering enabled on the SD722)
>- approximate recording level of the 1 kHz reference tone on both recorder=
>- gain setting on the PMD 671 : 5
>- gain setting on the SD722 : 68
>- no high-pass filtering selected on the mic pre-amps
>Except of some cutting, I did not change the original sound files:
>To me, the differences are very small. The dummy resistor test section
>reveals the lower inherent noise floor of the 722 preamps. However, as Kla=
>already mentioned, when using such high-output microphones, these
>differences should become nearly irrelevant.
>The most important difference is perhaps the higher (hard-wired) cutoff
>frequency of the PMD 671 preamp, which might reject a portion of the bass
>components that might be useful for a "spacy" sound. I don't know how the
>new PMD 671 exactly compares to the PMD 670 (the MIC Signal-to-Noise Ratio
>specification on the Marantz website is 65 dB for both models).
>Nevertheless, it makes much more fun to use the more professional
>SoundDevices recorder (high-quality metal housing, very nice battery
>powering concept, large internal harddisk, ...).
Thanks for the test! Seems like both recordings are noisier than what
I'd expect. In addition to more hiss than I'm accustomed to with
NT1A/722, I'm hearing (and seeing on a graph) a pronounced line hum
that's either in the room, in the signal paths, or both. I usually
have to shut down every appliance in the room and run entirely on
batteries to rid all RFI. I can also detect that a good portion of
higher hiss is centered suggesting its not from the mics. One should
be able to hear the noise from the mics themselves with the 722 which
is under 5dBA effective self noise at full gain- and its usually
assymetrical. The hiss noise is so similar in both of your tests/pres
that I'd guess its coming from a third source.
I was curious if my memory was right about the amount of noise so I
complied and uploaded a short mp3 with three segments in this order:
1) NT1A->671 @ "5" by RS
2) NT1A->722 @ 68 by RS
3) NT1A->722 @ full gain by RD
Tried to match levels of my test with your tests by ear. Rob D
University of Wisconsin-Milwaukee
[Non-text portions of this message have been removed]