Hi Paul--
Its assuring to see we can make some sense of the results, even if
not practical to apply. :-)
For me, it makes the significances of mic sensitivity simpler to
understand when I break it into three categories.
(1) High mic sensitivity cannot improve noise performance if one is
using a very quiet pre. Once we get into pres with measured input
noise of around -127 dBu (A weighted) and quieter, the pre is
effectively "transparent" and the higher sensitivity provides no or
little advantage. I had a MKH80 with 40 mV/Pa in the same test. The
amplitude/file saturation is higher but it's self-noise reigns.
(2) High mic sensitivity can improve noise performance by creating
more separation from the noise floor when the mics are used with a
"somewhat quiet" pre. I suggest for this group pres measured to be
-124 dBu and noisier (A weighted). The audible advantage is highly
impacted by the tonal balance, or "color" of the self-noise from the
mics and the input noise contributed by the mic pre and subject
matter. The separation and reduced noise will be considerably greater
for "foreground" sounds like robust animal vocalizations and louder
effects where low dynamic backgrounds are less a concern. For a
reliable sense of the noise performance with a "somewhat quiet"
recorder/pre, its best to try out specific mic/recorder combinations
than rely too heavily on self-noise and sensitivity ratings.
(3) Low mic sensitivity mics (>10 mV/Pa) can become a significant
liability especially if used with a "somewhat quiet" pre to record
lower sound-level sources/spaces. However, if used with a very quiet
pre for fairly robust sound sources, the noise impacts can be greatly
reduced if not eliminated.
I agree with you about the NT1-A's tonal balance. Several people
suspect its rating of 25mV/Pa is highly influenced by a "bump" at
<40Hz. It seems to have weak response to the lower mid-range and
useable bass registers than many mics, but this might be more common
than I was aware of. The bump/noise color of the NT1-A tends to
translate into considerable high Hz noise saturations when used for
full spectrum renditions of quieter locations where lower mids
(125-700 Hz) are important. That said, if we consider the rich
acoustic spaces that John Hartog and others have been able to portray
with NT1-A's, we know there's more to the story. As recordists we
tend to generate more evidence than explanation, but if we try to
make sense of both sides, our field work stands to improve
explanations. Rob D.
000 4/9/10, Paul Jacobson wrote:
>Hi Rob,
>
>Thanks for making these tests and uploading the results. The
>-130.7dB for NT-2A is with a 10dB safety margin so perhaps the shape
>of the noise floor of mic and preamp is a factor in audibility. It
>could be the flatter the noise floor of both mic and preamp the more
>likely the lower end of the 6-10dB range will be adequate. The
>calculated output noise figures would only ever be valid in an
>environment where the noise level was below the noise floor of the
>mic so these "tick" tests
>
>Looking at the original pink noise files the difference in level
>between the NT-2A and NT-1A is 1.5dB peak, 1.3dB average and 1.3dB
>rms which reflects the 2dB difference in sensitivity indicated by
>the manufacturers specs. Using a pink noise calibration tone to
>match levels is potentially more revealing that trying to match
>apparent levels. The differences in tonal balance are inherent in
>the design of mics and preamps so when comparing equipment we should
>be preserving these differences rather than trying to negate them.
>I'd personally boost the NT2-A by around 1.3-1.4dB to do comparisons.
>
>The differences between the mics are quite interesting - the NT-1A's
>sound rather "thin" compared with NT-2A which seems to have more
>emphasis on the low frequencies. I definitely agree the NT-2A has
>more noise but that is borne out by the specs.
>
>cheers
>Paul
>
>On 09/04/2010, at 7:33 AM, Rob Danielson wrote:
>
>>
>> Hi Paul--
>> Here are some test materials in a folder to download. I recorded both
>> mics with 58.6 dB of Gain with a SD 744T recorder.
>>
>> <http://tinyurl.com/y8haout>http://tinyurl.com/y8haout
>>
>> The Hz response on the two mics is very different. Mic self-noise,
>> high in register with both mics, is not significantly different in
>> color.
>>
>> Using RMS matching of pink noise is the only way I know of to match
>> playback levels when the tonality of the two mics differs as
>> significantly as these do. With the additional LF response it has,
>> the less sensitive NT2-A needed only .5dB of additional gain to match
>> the NT1-A (pink noise RMS). You can easily hear the additional LF
>> response in the room tone/tick section as well and there's not much
>> LF there to work with!
>>
>> I we assume as Rane Table 3 suggests that the mic pre in the 744T
>> [measured at -130dBu (A weighted)] is quiet enough for an NT1-A mic
>> (which we also know from other tests), then I don't understand why
>> Rane Table suggests a quieter pre is needed to become transparent
>> behind the NT2-A. When the signal output of the two mics is matched,
>> the sonograms I've made indicate that the NT2-A has slightly more
>> effective self-noise noise, not less. In my experience, this usually
> > means the pre can be slightly noisier, not less. If I've missed
>> something glaring, I'm ready fer educatin' Rob D.
>>
>>>
>>> At 5:03 AM +1100 4/1/10, Paul Jacobson wrote:
>>>>
>>>
>>>> On 31/03/2010, at 5:19 PM, Rob Danielson wrote:
>>>>>
>>>>> Rode NT-1A 25 5 -118.83 -128.8 dBu YES, A-weighted
>>>>> Rode NT-2A 16 7 -120.70 -130.7 dBu?? Side by
>>>>> side, the 2A is noisier than the 1A. In practice one could get by
>>>>> with a slightly "noisier" pre, not a "quieter" one and one would not
>>>>> need -130.7 dBu (A weighted).
>>>>
>>>> There is no question that the NT-2A has a higher level of
>>>> self-noise, but based on the published specs if you recorded both
>>>> the NT-1A and NT-2A at identical gain settings on the same recorder
>>>> the recorded mic noise floor of the NT-2A _should_ be lower in terms
>>>> of of dbFS than the recorded noise floor of the NT-1A. I'd be
>>>> interested to see your results of such an unadjusted comparison.
>>>
>>> Hi Paul--
>>>
>>> Its on the test "to-do" list. I'm not following how rec gain makes a
>>> difference in mic self-noise performance and the resulting recording
>>> if a sufficiently quiet pre is used. Such a pre would be necessary to
>>> compare the self-noise performance of the two mics.
>>
>>
>>
>> --
>>
>>
>>
>> ------------------------------------
>>
>> "While a picture is worth a thousand words, a
>> sound is worth a thousand pictures." R. Murray Schafer via Bernie Krause
>> Yahoo! Groups Links
>>
>>
>>
>
>
--
|