naturerecordists
[Top] [All Lists]

Re: documenting recordings...

Subject: Re: documenting recordings...
From: "Walter Knapp" waltknapp
Date: Mon May 7, 2007 9:34 am ((PDT))
Posted by: "Gianni Pavan"

>          I'll like a lot to have a recorder able to log all possible
> info about the recording... at present, when I record on a laptop or
> on a tablet (very soon I'll test a Samsung Q1 for field recording) I
> use my own software that logs date, time, gps position (if a gps is
> connected) and technical info I set at the beginning of the
> recording. For maximum safety I embed in the filename as much
> information I can, including sample rate, bit depth and number of
> channels. This is useful in case something goes wrong and the header
> results corrupted. With all that data in a filename (that can be
> corrupted too) and in a txt logfile, I normally can restore the
> corrupted files (if it happens...). Moreover, all that info in the
> filename helps a lot when I browse through thousands of files, or
> when I give a copy to someone.

I think everyone would like a system that would do the task of recording
all data so we did not have to. I don't think it will ever happen. The
data needed to be recorded varies for each of us.

There is a big push for cameras with built in GPS. On the surface it is
a good idea in that it would give you a location and calibrated time for
each shot if it worked. What is a problem for my mind is that they still
have not produced a GPS that's not a battery eater. And GPS, even the
newest ones take time to find themselves each time they are turned on.
Of course there is also the problem that GPS cannot find it's location
everywhere. Shots taken inside buildings, forests, deep valleys and so
on would likely either take a long time for the GPS to be ready or it
would never be ready. I've used GPS steadily since it first became
available, so have less faith that it will magically give me a location
that's usable all the time everywhere.

What I use the filenames for is the date of the shot (or audio
recording). I've tried other uses for the filename, but that one has
consistently been the most useful. I don't believe in very long
filenames in most disk or folder displays they get truncated or
seriously interfere with what you can see in the window. Each day's
shoot has it's own folder also date named, and then each file in the
folder, that's the basics. My master backups are in chronologic order
too. In fact my film slides are also organized the same way.

I'm not convinced that recording the position of gain knobs and such
like on the uncalibrated recorders we use is of much value. Those values
are not that precise, I know I don't depend on that for setting any
future recordings. I use headphones and the level metering for every
recording to determine settings. Of course if you are using a calibrated
system and calibrate it before and after each recording (the way you
would use a calibrated sound meter to make a legally valid measure),
then knowing the exact settings would be something to do.

As I've noted all the data ends up in a Filemaker database in my system.
Not only is the program designed entirely around not losing the data,
but it's also designed so you can easily find it. That's the biggest
failing of metadata, it's not quickly accessible across thousands of
files occupying numerous disks. A central database is needed for this.
The other big failing of metadata is poor support, it's not protected by
the programs we use and is easily lost with no warning. You can't depend
on it surviving no matter what.

I can't remember a time when my in depth archive system and database
backups did not take care of any file corruption. In fact it's been a
very long time since I've even had any file corruption. That's what a
multilayer archive system is about, however.

Walt





<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