naturerecordists
[Top] [All Lists]

1. Re: Choosing the right sampling rate and sample size

Subject: 1. Re: Choosing the right sampling rate and sample size
From: "Rob Danielson" danielson_audio
Date: Tue May 25, 2010 8:15 am ((PDT))
At 6:26 PM +1000 5/25/10, Paul Jacobson wrote:
>
>On 24/05/2010, at 12:16 PM, justinasia wrote:
>
>>  --- In
>><naturerecordists%40yahoogroups.com>=
om,
>>Paul Jacobson <> wrote:
>>>  The FR2-LE uses an A/D chip with S/N of 95dB for example which
>>>  effectively makes it a 15-6 bit recorder.
>>
>>  Does this mean that with the FR2-LE there is no benefit using 24
>>bit over 16 bit? I have been using 24 bit up to now but if this is
>>true then I had better be using 16 bit and save space.
>
>That is something you would need to test and decide for yourself. As
>has been suggested it might be worth capturing recorder noise as it
>is possible to discern some detail in the noise floor. I've seen
>comments to the effect that some detail 1-3 dB below the noise floor
>can still be made out. You'd have to decide if this is worth the
>extra disc space.
>
>cheers
>Paul
>

I agree thoroughly with Paul. A comparison test can give one great
confidence in this decision- particularly if you frequent the same
places and are learning what kind of transient peaks are typical and
which of them you want to capture without over-modulation. If the
dynamic range of the setting isn't great (the majority of natural
settings have moderate dynamics), 16 bit recording _should_ match 24
bit recording performance, at least perceptually.

I'd use (fixed) max gain or very near max gain and make two, 10-20
minute recordings in the middle of the night or other time of day
when the background presence is at minimal levels. Take
quietest/lowest saturation passages from both the 16 ands 24 bit
recordings and cut them into a checkerboard A/B pattern in a
multi-track timeline. Add the same amount of gain to both to increase
audibility.

You might also want to take these 16 an 24 bit recordings and some
others with interesting (low level) content and run them through your
usual equalization/gain boosting practices and then compare 24 bit
masters made with these settings in effect.

I personally do such tests "blind" because it adds some drama and it
does reduce chances of bias.

Another compromise is to record in 24 bits and log/make excerpts with
Reaper or other flac compatible app, make 24 bit excerpts, convert
the full res PCM files to .flac and save the Reaper session. You can
then access your original material any time using the loss-less flac
files and toss the PCM's. This cuts your 24 bit storage demands in
half with no loss in quality. Rob D.
--









<Prev in Thread] Current Thread [Next in Thread>
Admin

The University of NSW School of Computer and Engineering takes no responsibility for the contents of this archive. It is purely a compilation of material sent by many people to the naturerecordists mailing list. It has not been checked for accuracy nor its content verified in any way. If you wish to get material removed from the archive or have other queries about the archive e-mail Andrew Taylor at this address: andrewt@cse.unsw.EDU.AU