At 10:21 AM +0100 8/28/08, Michael Oates wrote:
...
>It is also interesting that you record at 16/48K, I found that I can add
>more gain to 24bit recordings without bringing up the noise floor,
I'm not one of those who can discern the 16/24 difference audibly.
:-) If anyone has come across a blind test that clearly confirms
this ability, I'd love to try and take the test! I keep failing my
own tests,...
Some recordists might have recorders/mic pres where the input noise
increases at a faster rate than the gain increase towards the end of
the record level scale. In this case, using a lower gain setting at
the 24 bit rate could help improve noise performance for very quiet
subjects. As best as I can tell, recording at 24 bits might lower
noise only if the level of the background sound in the recording
drops under -55dB (96dBFS). Quantization noise can crop up when
_very quiet_ locations and very low pre gains are used-- lower than
most of us would choose to use. IMO, that was the fallacy of Sound
Device's reasoning behind their argument for recording at 24 bits.
(Most, understandably, would prefer to forget the discussion we had
about this and the 16/24 comparison tests:
http://www.uwm.edu/~type/audio-reports/LowSaturation/LowFieldSaturation2.ht=
ml
The summary seems to be fairly consistent with the math calculations
other folks have performed.)
There is definitely a lot more headroom or dynamic range with 24 bit
files. In the fairly quiet place I live (45 lat in North America), I
find that I typically only need 25 dB above the background level for
natural subjects. Those who live closer to the equator seem to
require more dynamic range. I can set my pre gain at 60dB with mics
with 16 mV/Pa sensitivity and record everything including very close
encounters with wildlife without over modulation. Sometimes. I have
to turn the gain down for thunderstorms for close lightning/thunder.
My background levels are considerably above -55dB on the quietest
winter nights.
> it also gives me more headroom in case something gets rather close
>to the mics. I
>also find 96K better especially for frequency plots to produce sonograms a=
s
>I seem to miss a lot of the plot if I don't. So 24/96K files are rather
>large, hence saving 50% is very welcome, so thanks again :)
The .flac files seem to be require about 33% of the storage space. I
was hesitant to toss the full res files for a long time. I checked to
see whether AIFF files rebuilt from the saved .flac files were
identical to the original AIFF files and they were. Some PC apps
might record and play .flac files too. In practice, some kind of
compression will be used for most kinds of delivery as time
progresses. It goes against my instincts about "quality," but its
probably just as "safe" to preserve the recordings in .flac as full
res. Rob D.
>
>Mike
>
>>Kevin & Mike, et al--
>>I've been recording all-night (8-12 hour) takes, 5-7 nights/week
>>since last Fall and have come-up with some "improvements" in my
>>procedures that might be of interest to you.
>>
>>1) I record in 16/48K with an application called, Amadeus Pro
>><http://www.hairersoft.com/AmadeusPro/AmadeusPro.html>http://www.hairerso=
ft.com/AmadeusPro/AmadeusPro.html
>>directly to a
>>computer hard-drive. One can get 500GB internal hard drives for
>>laptops today. I used to record at 24/48K bits but I found that I
>>could not discern any difference if my background levels in the 16
>>bit originals did not drop below -55dB based on 96 dBFS. All post
>>work is done in 24 bits.
>>
>>2) When I click "stop" or use Amadeus' timer to stop the recording,
>>the application creates a waveform "overview" file that shows the
>>_whole night_. I have found this view really useful for gaining a
>>sense of the night's rhythms, sensing the passage of time and finding
>>events-- especially compared with working with numerous segments. I
>>take a screen shot of the night's overview and place it in my log. I
> >ca use the the waveforms "see" seasonal developments like the onset
>>of insect species, the affects of weather changes, man-made
>>intrusions, etc., by comparing waveform views.
>>
>>3) Storage is no longer a hassle for me. After I make "outs" or
>>smaller sound files I want to collect and study later, I use, "save a
>>copy as .flac" and the _entire_ 8-12 hour sound file is saved as a
>>compressed (lossless) file that is only 1.8 - 3.0 GB in size. That
>>file, along with the "outs" and my log notes will burn to an 26 cent
>>"archive" quality DVD-R in 24 minutes at 2X speed. I scan the DVDdisk
>>to my master disk catalog file (Disktracker) so I can find them
>>anytime by numerous search methods. Then I can free-up my hard drive.
>>
>>4) I monitor with wireless headphones as I work in the evenings and
>>as I sleep to increase the chances of detecting events. I also enjoy
>>taking 30 minutes in the morning to "spot play" the master recording
>>and make excerpts before I trash the full res 16k/48 files. The
>>lossless .flac file retains everything for future uses. The only
>>thing I'd like to automate is the location of the "outs" I make on
>>the master file-- the saved .flac master doesn't retain these
>>regions, I have to enter them in my log manually.
>>
>>For me, I prefer creating a record of the whole night and it has
>>turned out to be less of an headache than dealing with a bunch of
>>fragments that take more work to identify/make relevant. Amadeus is
>>$40USD, but, unfortunately, its only for Mac. Rob D.
>>
>> =3D =3D =3D =3D =3D
>>
>>At 1:13 PM -0600 8/26/08, Kevin Colver wrote:
>>>Mike,
>>>The decision on how much to store depends on the purpose I was
>>>recording. If I am after a specific species or sound I will scan and
>>>save the recordings of that sound while splicing out and discarding
>>>the trash. Some recordings are done for soundscape archival and
>>>future scientific reference, these are saved intact with full
>>>documentation of location, weather, time, ect.
>>>
>>>I just hope the price of memory storage keeps dropping as fast as my
>>>collection grows. I remember being excited to have a 1 gigabyte hard
>>>disk not too long ago, now the 500 gig is filling up.
>>>
>>>Kevin
>>>
>>>On Aug 26, 2008, at 12:59 PM, Michael Oates wrote:
>>>
>>> > Kevin,
>>> >
>>> > Thank you, what puzzled me was you said "set to start a new file
>>> > every 60 min", I was
>>> > aware of the option to start new files after a set size limit was
>>> > reached, but not after a
>>> > time limit. The menu number is 41 on the 744T.
>>> >
>>> > So I guess you either record at 24/48 stereo or use single tracks
>>> > and use 24/96
>>> >
>>> > Good idea!
>>> >
>>> > Now then once you have gathered these all night recordings, how do
>>> > you go about checking
>>> > them for sounds you want to keep. Do you actually listen to it all,
>>> > or do you skim through
>>> > looking at a spectral display so that you can see where possible
>>> > sounds may be and just
>>> > listen to parts.
>>> >
>>> > Going on from there, these files take up masses of storage space,
>>> > once you have gone
>>> > through the files and taken cuts out of the good bits, do you then
>>> > delete, or do you
>>> > archive the whole lot?
>>> >
>>> > Thanks,
>>> >
>>> > Mike
>>> >
>>> > >Mike,
>>> > >On my 722 menu there is option 33: File: Max Size which I set at 95=
0
>>> > >megabytes which is about one hour. There are also options of about
>>> > >30 min and 2 hours. At the end of the megabyte limit the 722
>>> > >immediately launches into a new file with continuous recording.
>>> > >There is no loss of data or pause between the blocks of sound. I ca=
n
>>> > >splice them together if needed.
>>> > >
>>> > >BTW, I also have a Sony PCM D1 but haven't been able to figure out
>>> > >how to set a limit on file size or duration. It will go on recordin=
g
>>> > >more than 2 hours in one block and finally stop when memory is full
>>> > >or battery dies. Anyone know if the file size can be set on this
>>> > >machine?
>>> > >
>>> > >Thanks,
>>> > >Kevin
>>> > >
>>> > >
>>> > >On Aug 26, 2008, at 10:56 AM, Michael Oates wrote:
>>> > >
>>> > >> Kevin,
>>> > >>
>>> > >> I am new to SD recorders, I have a 744T. How do you setup the
> >> > >> recorder to start a new file
>>> > >> every hour. And are there any gaps in the recording or can they b=
e
>>> > >> pieced together without
>>> > >> any gaps or glitches.
>>> > >>
>>> > >> Thanks,
>>> > >>
>>> > >> Mike
>>> > >>
>>> > >> >I agree, it's nice to hear a bit of a blog on Dan's recording
>>> > >> >adventures. When we hear stories like this we pick up little tip=
s
>>> > >> >that can only be learned by experience, ours or those of someone
>>> > >> >else. Please all, send in your stories.
>>> > >> >
>>> > >> >Agreed; that leaving the recorder on is helpful. Bernie taught
>>> > that
>>> > >> >to me and I've captured some great material by recording hours
>>> > at a
>>> > >> >time. My SD722 is set to start a new file every 60 min, so in th=
e
>>> > >> >studio at home I can review the recordings in manageable 60 min
>>> > >> >blocks. Sometimes, when something great comes out of the night,
>>> > you
>>> > >> >want to have the preceding several minutes to frame the
>>> > interesting
>>> > >> >sound.
>>> > >> >
>>> > >> >And it's nice at night to have the mic at the other end of a lon=
g
>>> > >> >cord. Bernie's recordings of a bear mouthing the mic and a jagua=
r
>>> > >> >growling at it in pitch darkness are dramatic. Once I was startl=
ed
>>> > >> >out of a half sleep while monitoring at night by an angry loud
>>> > >> >hissing. I jumped because it sounded like something was right
>>> > in my
>>> > >> >face. Lucky for me, the patrolling bobcat was down at the other
>>> > end
>>> > >> >of the 100 foot cable and I was reclined safe and warm in my
>>> > parked
>>> > >> >truck.
>>> > >> >
>>> > >> >Kevin
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >> >On Aug 25, 2008, at 6:21 PM, Romilly Hambling wrote:
>>> > >> >
>>> > >> >> --- In
>>><naturerecordists%40yahoogroups.com><naturerecordists%40ya=
hoogroups.com>
>>>Dan Dugan <>
>>> > wrote:
>>> > >> >> >
>>> > >> >> > Sharon Perry and I did our second Yosemite sequoia groves
>>> > >> expedition
>>> > >> >> > August 8-10. ... To be continued.
>>> > >> >> > -Dan Dugan
>>> > >> >> >
>>> > >> >> >
>>> > >> >> Nice to resurface on the list and find Dan writing about
>>> > recording
>>> > >> >> adventures out in the wild again -- AND at night. I look
>>> > forward to
>>> > >> >> hearing samples from that 4 channel 3032 setup.
>>> > >> >>
>>> > >> >> Dan, run the thing all night, turn in and get some proper slee=
p!
>>> > >> Do I
>>> > >> >> know what it's like "sleeping" with headphones on and a finger
>>> > >> on the
>>> > >> >> pause button. Then you get a good day, and back home you go
>>> > through
>>> > >> >> the files knowing you've missed absolutely nothing. Well,
>>> > yes, it
>>> > >> >> takes time, but what a lot there can be -- some of which would
>>> > >> never
>>> > >> >> wake you up or . . sort of heck, it's 3.37 am and can I
>>> > really be
>>> > >> >> bothered with this 57th possible false alarm.
>>> > >> >>
>>> > >> >> Most I excerpt, but one or two I keep as they're so magical. A=
ll
>>> > >> those
>>> > >> >> tiny noises, things dropping (as Dan says), and oh so mysterio=
us
>>> > >> >> visitors to the mics. Can be quite creepy sometimes.
>>> > >> >>
>>> > >> >> I was forced into this (leaving equipment running while gettin=
g
>>> > >> some
>>> > >> >> sleep) because the dreadful piece of tinny plastic I'm current=
ly
>>> > >> using
>>> > >> >> has all of a two-second buffer. All but useless unless you're
>>> > >> prepared
>>> > >> >> to turn yourself into a nervous wreck with arm cramp! So I jus=
t
>>> > >> left
>>> > >> >> it running. Still haven't been through all the files!!
>>> > >> >>
>>> > >> >> As to jets, etc, I'm now resigned to Sats and Suns, tho jet
>>> > traffic
>>> > >> >> can start up un Sun in anticipation of Mon I find. Oddly some
>>> > >> weekdays
>>> > >> >> are passable, but you have to get to know which. I've no idea
>>> > >> why they
>>> > >> >> should vary. Round us the night to avoid is Monday -- invariab=
ly
>>> > >> >> awful.
>>> > >> >>
>>> > >> >> On another note, I'm fairly sure digifish got his 3032s at lea=
st
>>> > >> >> partly because he heard my samples. I got my 3032s (now 2 pair=
s)
>>> > >> >> entirely because of this list, especially of course Eric
> >> > Benjamin's
>>> > >> >> report. Use'm all the time. Now many are wailing because they'=
re
>>> > >> >> unobtainable. tee-hee.
>>> > >> >>
>>> > >> >> Three cheers for nature recordists, night recordists and 3032s=
!
>>> > >> >>
>>> > >> >> romilly
>>> > >> >>
>>> > >> >>
>>> > >> >>
>>> > >> >
>>> > >> >
>>> > >> >
>>> > >>
>>> > >>
>>> > >>
>>> > >
>>> > >
>>> > >
>>> >
>>> >
>>> >
>>>
>>>
>>>
>
>
--
|