Nikon respond to RAW White Balance concerns

Scott Peterson writes:
Sure, JPEG. The specs are public and they are all over the place.
Unless all of civilization's knowledge is wiped out in some big icky
nuclear blowout, we will still know how to read JPEG files.

The specs are available for a lot of scientific data stored on tape in
the 1950's and 1960's. But there is no equipment available today to
read it.

That's a media and hardware problem, not a file format problem.

I didn't say it wasn't documented, I asked if there were any programs
around that would still support the format then or would it simply be
a historical curiosity? Different question.

There are programs published as source code today, written in C for
example, that support JPEG format. Do you think those programs are
going to disappear somehow, even if you keep copies of them along with
the JPEG files? Or do you mean will C compilers still be around? I
think C compilers will still be around. Heck, the x86 architecture
will probably still be around, at least in some kind of compatibility
mode, just like the 360 architecture is still around.

All well and good, but what's your equivalent of "processed" film,
that doesn't destroy any info from the RAW file?

At this point there aren't many. One that comes to mind, of course,
is Adobe's DNG.

I haven't examined DNG but I can see some logic to the notion that RAW
formats are really camera specific. Look at the Sony 828 with its
four color planes instead of three, the Fuji S1/S2/S3 pro DSLR's with
their hexagonal CCD grids, etc. Who knows what weirdness future
camera makers will come up with? For DNG to really be universal, it
has to cope with all of those future concoctions that nobody has
thought of yet. So, the best bet is for camera vendors to publish the
RAW formats for the individual cameras.

Just do a web search on DMCA and reverse engineering or if that's too
much, just click here and read a few articles by people who really do
know this stuff.

From http://www.chillingeffects.org/reverse/:

The Digital Millennium Copyright Act (DMCA) generally prohibits
circumvention of technological protection measures. On its face,
since circumvention is generally required for reverse engineering,
this prohibition would prevent reverse engineering of those
measures that control access to a copyrighted work.

The DMCA reverse engineering stuff is about circumventing access
controls to copyrighted works. Since the white balance data is not a
copyrighted work, my read is that the DMCA does not apply.

Scott Peterson wrote:

Paul Rubin wrote:


Can you guarantee that ANY currently used image format will be
readable in fifty years? No.
...


RAW is a format for storing unprocessed images. I can't help thinking
that a good simile is film. RAW is the equivalent of undeveloped
film. You certainly wouldn't keep that around. You process it into a
negative or slide and then work with it or archive it. .

All well and good, but what's your equivalent of "processed" film,
that doesn't destroy any info from the RAW file? JPEG is lossy
compression and has only 8 bits color. I'm not sure if 16-bit TIFF is
viable right now. It might be an alternative, but it still means the
RAW data has been through an interpolation process and you might want
to use a different one. A documented RAW format is the best bet so
far.


At this point there aren't many. One that comes to mind, of course,
is Adobe's DNG.


I don't see why various RAW files couldn't be processed into a standard
DNG. I mean it's just raw sensor data, WB & such is not critical, I
don't know what other kind of info they could be including in these
files other than various compression types, proprietary panorama tags &
other relatively unimportant junk. By converting to a common format, at
least the files will be readable in 10-20 years on 'Windows 2025' or
linux 2020.

Scott Peterson wrote:

Paul Rubin wrote:


Can you guarantee that ANY currently used image format will be
readable in fifty years? No.
...


RAW is a format for storing unprocessed images. I can't help thinking
that a good simile is film. RAW is the equivalent of undeveloped
film. You certainly wouldn't keep that around. You process it into a
negative or slide and then work with it or archive it. .

All well and good, but what's your equivalent of "processed" film,
that doesn't destroy any info from the RAW file? JPEG is lossy
compression and has only 8 bits color. I'm not sure if 16-bit TIFF is
viable right now. It might be an alternative, but it still means the
RAW data has been through an interpolation process and you might want
to use a different one. A documented RAW format is the best bet so
far.


At this point there aren't many. One that comes to mind, of course,
is Adobe's DNG.


I don't see why various RAW files couldn't be processed into a standard
DNG. I mean it's just raw sensor data, WB & such is not critical, I
don't know what other kind of info they could be including in these
files other than various compression types, proprietary panorama tags &
other relatively unimportant junk. By converting to a common format, at
least the files will be readable in 10-20 years on 'Windows 2025' or
linux 2020.

In article ,
Paul Rubin wrote:

There are programs published as source code today, written in C for
example, that support JPEG format. Do you think those programs are
going to disappear somehow, even if you keep copies of them along with
the JPEG files? Or do you mean will C compilers still be around? I
think C compilers will still be around.

Well quite. C is now over 30 years old, and is still extremely widely used
to write new applications. It's extremely likely to be around for another 30
years at least, and even if not, those of us who know it will be, and can do
hand translations into contemporary languages. It's not rocket science.

In article ,
Paul Rubin wrote:

There are programs published as source code today, written in C for
example, that support JPEG format. Do you think those programs are
going to disappear somehow, even if you keep copies of them along with
the JPEG files? Or do you mean will C compilers still be around? I
think C compilers will still be around.

Well quite. C is now over 30 years old, and is still extremely widely used
to write new applications. It's extremely likely to be around for another 30
years at least, and even if not, those of us who know it will be, and can do
hand translations into contemporary languages. It's not rocket science.

Paul Rubin wrote:
[snip]
I haven't examined DNG but I can see some logic to the notion that
RAW
formats are really camera specific. Look at the Sony 828 with its
four color planes instead of three, the Fuji S1/S2/S3 pro DSLR's with
their hexagonal CCD grids, etc. Who knows what weirdness future
camera makers will come up with? For DNG to really be universal, it
has to cope with all of those future concoctions that nobody has
thought of yet. So, the best bet is for camera vendors to publish
the
RAW formats for the individual cameras.
[snip]

The sort of technology change that would need a change to DNG is
relatively rare. If DNG had existed for a few years, the last such
change would have been a year and a half ago. (The Fujifilm SR sensor).
The Sony 4-colour sensor wouldn't have needed a change to DNG, because
4-colour sensors were already familiar. All the camera improvements
over the last year and a half did not need changes to a well-engineered
Raw format.

DNG has a version-numbering scheme designed to cater for major changes
of that sort. Each specification has a version number (the latest one
is 1.1.0.0), and both that number, and the oldest compatible version
number, are written to each file. If a new camera needs version
2.0.0.0, (they would discuss the requirement with Adobe), it will write
that number into the compatibility field. Then Raw processors can
identify whether they are able to handle that camera, or whether the
user needs an upgrade.

While DNG's mandatory fields contain all the image data, so any Raw
processor could handle the image, there is also an optional field that
manufacturers could use for extra information that would give their own
Raw processors an advantage.

As far as I can tell, the only disadvantage for camera manufacturers in
using DNG, (or some other common published format), is that they would
not be able to force photographers to buy their software by some form
of lock-up such as encryption. Essential image data has to be in a
published and unencrypted form.

But the advantages could be considerable, IF the manufacturer doesn't
intend to get lock-in. Apart from the publicity, and the savings from
not having to keep devising and documenting their own formats, they
would get rapid support by software that photographers typically demand
in their workflow. For example, it has taken Adobe some time to
catch-up with D2X, and other cameras, but if the cameras had output
DNG, either as native or as an option, support would already exist.

--
Barry Pearson
http://www.barry.pearson.name/photography/