Hi Vicki--
Thanks! I agree that there's no difference in the noise between these
two samples. I think, however, that both the metronome and the room
ambience behind it are considerably too loud for quantizing noise to
become audible. We need a _very low_ level subject like minimal
background ambience at the quietest time of night-- if you want-- a
very soft clock tick. We'd also need to study clips from the ORIGINAL
16 and 24 bit recordings-- no gain changes, no mp3 compression-- just
an 2-3 second excerpt from each of the .wav originals. May as well
record the same, very soft subject using High Sens "10" and "4"
settings too and the test will be more comprehensive? There's a
chance folks might be able to use Low Sens/"10" settings to improve
noise performance even with much quieter subjects-- but I'm not
certain, as yet. Do you have to change the sample rate when you
change the bit rate? If not, record both at 48K. We could discuss
this study off-list and get back to the list with anything
interesting we learn? Rob D.
At 3:05 PM +1000 7/10/08, Vicki Powys wrote:
>Hi Rob,
>
>OK since you asked! I will upload to the naturerecordists web page,
>my LS-10 beeper test that compares 16 bit with 24 bit.
>
>I was using an ME67 with the LS-10, set 3.5 metres from an
>electronic metronome. Settings were Low sensitivity, record volume
>was on 10. Meter was registering around -18 dB. Low cut was off.
>First clip is 44/16, second clip is 48/24, as set via the LS-10
>menu. After the transfer to my computer, I boosted the volume of
>both files by 10 dB (using Peak), then converted the 48/24 file back
>to 44/16. I combined the two files, with a brief silence to separate
>them. Then converted to MP3.
>
>At every stage of this process, I could not detect a difference in
>noise levels between the two files.
>
>The first section is 44/16, the second section is 48/24.
>
>If you can suggest a different way of doing this test, I am happy to
>try it!
>
>Vicki Powys
>Australia
>
>On 09/07/2008, at 2:42 PM, Rob Danielson wrote:
>
>> Hi Vicky--
>>
>> Thanks for the carefully confirmed findings. I'm sure those with
>> LS-10's are following along closely. Is there a chance you could
>> share short clips of your low sensitivity level "10" recordings made
>> at 16 and 24 bits? Do you happen to recall what the record meter
>> activity was like at Low Sens "10." Its possible that one could push
>> the improved noise performance at Low Sens level "10" even a tad
>> further. Rob D.
>>
>> At 10:51 AM +1000 7/9/08, Vicki Powys wrote:
>>> Rob & Raimund,
>>>
>>> With the ME67 and LS-10 recorder, I've now tried comparing volume 3
>>> and volume 4 at high sensitivity setting, against volume 10 at low
>>> sensitivity setting, boosting the low-10 file (by 5-6 dB) to match
>>> the levels of the other two files.
>>>
>>> It does seem to me that you are still better off to use LOW SENS
>>> setting on 10, there is less noise, compared to 4 or less on high
>>> sens.
>>>
>>> I then compared Low 10 at 24 bit and 16 bit settings but could not
>>> really detect any big difference in noise levels for the 2 settings.
>>>
>>> I might add that using the ME67 with the LS-10, on low sensitivity
>>> (if you have a good signal), really does give an excellent result,
>>> comparable to what I could get with my old TCD D10 DAT recorder (the
>>> DAT still has a slight edge but the LS-10 is a hell of a lot
>>> lighter!)
>>>
>>> Vicki Powys
>>> Australia
>>>
>>> On 07/07/2008, at 6:26 PM, Rob Danielson wrote:
>>>
>>>> Vicki--
>>>> Based on your tests so far, is this rule of thumb correct for
>>>> when to
>>>> use High of Low Sensitivity on the LS-10?
>>>>
>>>> Use the High Sensitivity setting on the LS-10 if the meter (in High
>>>> Sens mode) indicates that level of "4" or higher is required. Use a
>>>> Low Sensitivity setting of "10" if the meter (in High Sens mode)
>>>> indicates a level of "3" or lower is required. Rob D.
>>>>
>>>> = = = = =
>>>>
>>>> At 4:51 PM +1000 7/7/08, Vicki Powys wrote:
> >>>> Sorry that my last message to Rob was not quite right! I herewith
> >>>> make a correction.
>>>>>
>>>>> More indoor tests today (using a softly beeping electronic
>>>>> metronome): For loud sounds e.g. lyrebirds, if you have a strong
>>>>> signal at high sensitivity volume 2, you will get a better (less
>>>>> noisy) result if you switch to low sensitivity and volume 10. The
>>>>> signal will be about the SAME, but the noise a lot less. This rule
>>>>> applies if you are using ME67, WL-183s, or built-in mics.
>>>>>
>>>>> However if you need the recorder on high sensitivity volume 5
>>>>> to get
>>>>> a decent signal, it is not worth trying to use low sensitivity on
>>>>> volume 10 then boosting the signal later (if you do, the noise
>>>>> levels
>>>>> will be very roughly the same).
>>>>>
>>>>> I have done more tests with the WL-183s, on high sensitivity
>>>>> volume 5
>>>>> (a good average level for recording birdsong). They give a very
>>>>> slightly LESS strong signal than the built in mics, but when the
>>>>> signal levels are equalized, the 183s are still a bit less noisy.
>>>>> The 183s seem to work best (less noise) with the low cut ON,
>>>>> alternatively they need a bit of low roll off below 500 hz.
>>>>>
>>>>> Vicki Powys
>>>>> Australia
>>>>>
>>>>> On 07/07/2008, at 9:17 AM, Vicki Powys wrote:
>>>>>
>>>>>> Hi Rob,
>>>>>>
>>>>>> Yes I have tried lining up two files, one recorded in high
>>>>>> sensitivity volume 2, the other on low sensitivity volume 10. I
>>>>>> then
>>>>>> increased the levels of the low sens. file by 5 dB (also 7 dB) to
>>>>>> match the levels (of the recorded bird sounds or whatever) of the
>>>>>> high sens. file. The exact amount of increase needed I am still
>>>>>> working on, but it seems to be about 5-7 dB. Even with this
>>>>>> increase, the noise levels are still lower with the low sens.
>>>>>> file,
>>>>>> especially the FIZZ noise.
>>>>>>
>>>>>> Vicki
>>>>>>
>>>>>> On 06/07/2008, at 2:01 PM, Rob Danielson wrote:
>>>>>>
>>>>>>> At 11:52 AM +1000 7/6/08, Vicki Powys wrote:
>>>>>>>> All,
>>>>>>>>
>>>>>>>> Using an ME67 connected to my LS-10, I have done a bit more
>>>>>>>> testing
>>>>>>>> in the field (and indoors) of noise levels using high vs low
>>>>>>>> sensitivity. For lyrebirds, I do seem to get a better result
>>>>>>>> using
>>>>>>>> low sensitivity at volume 10, instead of high sensitivity at
>>>>>>>> volume
>>>>>>>> 2. There is much less FIZZ.
>>>>>>>>
>>>>>>>> But also, surprisingly, I seemed to get better results
>>>>>>>> yesterday
>>>>>>>> when
>>>>>>>> recording Regent Honeyeaters (they have a soft melodious call),
>>>>>>>> using
>>>>>>>> low sensitivity at volume 10, then raising the levels to match
>>>>>>>> what I
>>>>>>>> had recorded at high sensitivity at volume 5. Again, much less
>>>>>>>> FIZZ
>>>>>>>> on low sens.
>>>>>>>
>>>>>>> Did you confirm this with matched playback levels? If not,
>>>>>>> place the
>>>>>>> tests made with high and low sensitivity one after the other on
>>>>>>> different tracks in a time-line. Boost the playback volume of
>>>>>>> the low
>>>>>>> sensitivity one to match the playback level of the louder (high
>>>>>>> sensitivity) one. You should be able to hear the difference
>>>>>>> in the
>>>>>>> "fizz" noise level where the two files join.
>>>>>>>
>>>>>>>>
>>>>>>>> This seems to be a very complex issue! More testing needed!
>>>>>>>>
>>>>>>>> I haven't tried these tests with the WL-183s, but would
>>>>>>>> suspect a
>>>>>>>> similar result.
>>>>>>>
>>>>>>> Perhaps. The WL-183's are fairly sensitive mics but not in the
>>>>>>> same
>>>>>>> ballpark with the ME-67.
>>>>>>>
>>>>>>>>
>>>>>>>> Re my binaural setup, and your comment Rob (see below), where
>>>>>>>> do I
>>>>>>>> find Aaron's DSM temple mount set up?
>>>>>>>
>>>>>>> Aaron uses a WHB/N Windscreen
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>><<<http://www.sonicstudios.com/access.htm#whb>http://www.sonicstudios.com/access.htm#whb>http://
>>>>>>> www.sonicstudios.com/access.htm#whb>http://
>>>>>>> www.sonicstudios.com/access.htm#whb
>>>>>>> I'm not sure if Leonard
>>>>>>> sells just the mount-windscreen or not but you can see how his
>>>>>>> design
>>>>>>> works in the pictures. Rob D.
>>>>>>>
>>>>>>>
>>>>>>>
> >>>>>>>
>>>>>>>> Vicki Powys
>>>>>>>> Australia
>>>>>>>>
>>>>>>>> On 03/07/2008, at 12:31 AM, Raimund Specht wrote:
> >>>>>>>
>>>>>>>>>
>>>>>>>>> Rob Danielson wrote:
>>>>>>>>>>
>>>>>>>>>> A metronome at 5 metres might be a pretty "hot" reference
>>>>>>>>>> signal.
>>>>>>>>>> Its possible that the "Low Sensitivity" setting might not
>>>>>>>>>> provide
>>>>>>>>>> sufficient gain for quieter sounds and environments. Its
>>>>>>>>>> worth
>>>>>>>>>> further investigation, for sure.
>>>>>>>>>>
>>>>>>>>>> Coincidentally, earlier today, I thought of Aaron's DSM
>>>>>>>>>> temple
>>>>>>>>>> mount
>>>>>>>>>> preference in relation to your WL-183 rig priorities. Rob D.
>>>>>>>>>
>>>>>>>>> Yes, this is a relatively loud test signal. Though, I believe
>>>>>>>>> that
>>>>>>>>> this is an adequate model for a loudly singing Lyrebird.
>>>>>>>>>
>>>>>>>>> For recording softer sounds, one should perhaps better use the
>>>>>>>>> HIGH
>>>>>>>>> MIC SENSE setting. The problem with the Lyrebird was obviously
>>>>>>>>> that
>>>>>>>>> the REC LEVEL was turned down to 2, which caused the higher
>>>>>>>>> overall
>>>>>>>>> noise level. In other words, it is not appropriate to
>>>>>>>>> attenuate
>>>>>>>>> the
>>>>>>>>> microphone signal and then amplify it again in the next stage.
>>>>>>>>>
>>>>>>>>> Note that the MIC 2 input of the TASCAM DR-1 also provides a
>>>>>>>>> relatively low gain (input clipping level: -28 dBu), while its
>>>>>>>>> inherent noise floor (-115 dBu(A)) is still good enough for
>>>>>>>>> the
>>>>>>>>> K6/ME6x series.
>>>>>>>>>
>>>>>>>>> Regards,
>>>>>>>>> Raimund
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>>>>
>>>>>>
>>>>>> ------------------------------------
>>>>>>
>>>>>> "While a picture is worth a thousand words, a
>>>>>> sound is worth a thousand pictures." R. Murray Schafer via Bernie
>>>>>> Krause
>>>>>> Yahoo! Groups Links
>>>>>>
>>>>>>
>>>>>>
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>>
>>>>
>>>>
>>>>
>>>> ------------------------------------
>>>>
>>>> "While a picture is worth a thousand words, a
>>>> sound is worth a thousand pictures." R. Murray Schafer via Bernie
>>>> Krause
>>>> Yahoo! Groups Links
>>>>
>>>>
>>>>
>>>
>>>
>>
>>
>> --
>>
>>
>>
>>
>> ------------------------------------
>>
>> "While a picture is worth a thousand words, a
>> sound is worth a thousand pictures." R. Murray Schafer via Bernie
>> Krause
>> Yahoo! Groups Links
>>
>>
>>
>
>
--
|