Hi--
Sorry about the delay. No electricity here today which was sort of
refreshing,.. and revealing.
Correct:
MR, EK, RS, PJ,
Incorrect:
TW, JT
See answer near bottom of inserts into Mike's email below:
At 5:27 PM +0000 8/18/10, Mike Rooke wrote:
>
>...the [DR680] has very poor rejection of 50hz mains or cabling
>issue? (9 dB at 50hz worse)
I'm not sure about the source of this hum. I suspect its
"environmental" to the extent that most or all of it wouldn't be in
the recordings made by both recorders away from electric uses. This
would have to be tested to be sure, of course.
>
>Considering both pre's are at max gain (thus exhibiting their lowest
>noise) and the Mic is 8 dBA self noise? - the ticking clocks are
>+40something dB above the noise which may make it difficult to
>discriminate the noise, which is below the mic self noise.
I agree that the ticks are at a faster rate and louder than ideal for
analysis. The mic self- noise is very audible to me as-is. The high
tones in the clock tick do mask some of it though.
>We're trying to determine a 2dB? difference between pre's with 8 dB
>of mic self noise on top?
Raimund's measured 3 dBu noise difference [-130dBu (A) for the 702
and -127dBu (A) for the DR680]. This input noise difference appears
to be hard or impossible to detect "behind" the considerably louder
mic self-noise.
> >From 300hz -1000hz, I measured just 0.349 dB difference between
>the recordings. Above 10Khz its around 1.365 dB (Averaged from 2
>second sections of the first two) - I'd assume the figures could be
>from slight gain level differences in the recorders.
These differences come from several places; even the mics have
slightly different outputs. Note that playback levels are adjusted to
match for one to evaluate other differences by ear. I use a
combination of measurement and perceptual methods in making the
"matches." Basically, its impossible when the Hz responses of the
mic preamps differ a lot. This test is borders on the problem as Paul
points out.
I assembled the test blind and could not ID the recorders until I saw
the <50Hz response in a sonogram. I think the line hum is masking a
good bit of this LF performance difference-- at least it doesn't
stand out in speaker playback. Extended LF response is a SD pre
trait that I'm sure several people recognized.
The SD702 is B-- the first and last samples in the time line.
=3D =3D =3D
>Note to Dan Dugan, make an XLR plug "PreGenerator" alternate between
>two or four resistive loads at a predetermined rate that can be sold
>for testing pre's.. Ticking clock, distant bird call available as
>options. :)
A measurement of -127 dBu (A) or better seems to be a reliable
breaking point at which we can stop worrying about recorder input
noise and start focusing on other quality factors. We've reached
this conclusion with separate 680, FR2-LE, HDP2 and MP2 tests. See
as one example, http://tinyurl.com/yg2oopb.
If the joke is about also doing resistor load testing in the same
instant, I would agree that these pre tests using very low noise mics
serve most of all as a reminder that mic self-noise largely takes
over around -127 dBu (A). Input noise differences through resistor
load testing can be found, but based on the listening tests so far,
its very difficult or impossible to hear them with the lowest noise
mics we have at common disposal. Indeed, there are very cold, still
nights in January when I may as well be recording resistors, but its
all I can do to try and understand how things sound before how they
measure. :-)
>Mind you all this is [moot] if the recorder doesn't do its job of
>recording, which is a shame considering the 680's pre's are this
>quiet - brings lots of potential for multichannel arrays in the
>future.
>
>Thank you for the testing.
Thanks to Emil! :-) Rob D.
>
>BR
>Mike.
>
>
--
|