Secret editing of EXIF data for photographs using freeware

Nomen Nescio wrote:
clandestin_=C3=A9cureuil wrote:

Cyberiade.it Anonymous Remailer wrote:
clandestin_=3DC3=3DA9cureuil wrote:
=20
As the time is dependent on the camera owner setting it correctly and =
mos=3D
t=3D20
cameras don't have geo-data capability, the only real value in EXIF da=
ta =3D
is=3D20
to identify the camera and its settings.
=20
Wow.
=20
You obviously have no knowledge what so ever with respect to what
EXIF is, does, and/or means.
=20
=20
=20
I have enough knowledge to use exiftool and a beta version of Photome on =
a=20

Yes, as I said the nonexistant nature of your "knowledge" is
self-evident.

You said it, but it isn't so. I imagine that happens quite a lot with you.
I am not an expert, that is why I am here, hoping to learn more, but I
certainly have a good amount of general knowledge and experience.


What is your point

That toys like Exiftool and Photome don't even begin to scratch the
surface of the information modern cameras store in images.

Ok, if Exiftool and Photome are toys, enlighten me, what are the "real
deal" in such software?


That even competent tools won't help one "forge" EXIF data because much
of that stored information is proprietary, binary, and probably
undecipherable by anyone who doesn't have an NDA from the camera's
manufacturer in their pocket.

I have no interest in, nor have I suggested or implied that such a thing is
possible with any software program. I was asking someone else why they
wanted to change data other than time and place data. That data is usually
concerned with camera id and specific settings for each shot. You seem to
disagree. Is that simply because you are such a disagreeable person or do
have something valid to add?

Aside from camera id - model, make etc., and settings, what is it that you
find me to be so alarmingly ignorant of? What mystical data is there that
"toys" like Exiftool and Photome can't isolate? Is the Secret of Colonel
Sander's Eleven herbs and spices there? The whereabouts of Jimmy Hoffa? A
map leading to Elvis' whereabouts?



And that you're completely oblivious to all of that, as your
laughable "camera and it's settings" bloviation so succinctly
demonstrates.

Bloviation? I was quite succinct. My use of succinct is valid, you on the
other hand seem to be implying that I was both verbose and concise. Make up
your mind.


Now would there be anything else I can clear up for you today?

Yes. Are your parents going to invite you to their wedding?


Secret Squirrel

--

Ingrid Rose

clandestin.ecureuil(insert missing symbol here)gmail.com

Cyberiade.it Anonymous Remailer wrote:


HOWEVER, all this crap you are talking about should be
removeable by removing ALL non-picture data (e.g. by decoding
to a pure bitmap raw format (e.g. using Imagemagick) (not a
"camera raw" or a Windows .bmp) but a file containing ONLY the
bits themselves, creating a jpeg from that, and adding in
only what the user wants in a exif.

That method, of course, does not hide the fact that the exif
is "synthetic".

Of course, another method is to place the image data alone into
a file stolen off the net, complete with exif data. It won't be
bogus exif data ... just WRONG exif data. The human readable part
can then be changed at will.

Doug MCDonald

Cyberiade.it Anonymous Remailer wrote:
Ok, if Exiftool and Photome are toys, enlighten me, what are the "real=20
deal" in such software?

There's tools like exifprobe and exiv2, but even they won't ferret
out everything (although they do a much better job than the toys
you're using). To be thorough you'd need to use software specific
to the camera in question. Most of it costs big pesos, and you
still can't be sure you're seeing everything.

I'm not familiar with Photome. Claiming /exiftool/ to
be a toy compared to /exifprobe/ and /exiv2/ is an
interesting concept...

Well, as an example all my Nikon DSLR's newer than about 2 years
read information from lenses and "encode" it in a 32-byte value. It
changes image to image even when the images are snapped in rapid
succession. What it consists of is "undefined". And there's several
fields where even the field name itself is a HEX value, so you can't
even know what that field pertains to let alone what the value
might mean.

You are confusing the information displayed by a program
with the Exif specification.

*All* records (called an Image File Description) are
keyed to a integer value in the ID Tag (index field).
There is no "field name" in the data kept with an image.
Rather there is, in the specification, a text title
assigned to each index tag value.

For example, there is a IFD with the tag name
"Orientation of Image" and with a "field name" of
"Orientation". The ID TAG value for this IFD is 112 in
hex (274 in decimal). Also included in the record is a
Type indicator which in this case will indicate the
value field will have a 16 bit integer, and a Count
indicator which will be 1. That single 16 bit value can
be decoded only as one of the following:

Value Row 0 Column 0
--------------------------
1 -- top left
2 -- top right
3 -- bottom right
4 -- bottom left
5 -- left top
6 -- right top
7 -- right bottom
8 -- left bottom

That IFD uses only 11 bytes of data in the image file
primarily because it does not include the ascii text for
the tag name or the field name.

One effect of that form of indexing is that a programmer
can write a display routine for "unknown data" (the
definition of which the programmer simply does not have
at the time). Typically such a display will provide the
(in hex) instead of a title in text. However, the value
for an unknown field often cannot be properly decoded,
and is usually displayed as one or more integer values
of unknown significance.

One of the reasons to use Exiftools is that the package
is well maintained and the current version is likely to
have appropriate titles and decoding schemes for what
will be labeled only with a hex value in other programs,
which is apparently what you are seeing and assuming the
displayed values mean it is unspecified. In fact those
values are specified, but your program is not up to date.

Your toys probably don't even show much of this, let alone give you
a clue how to interpret or manage it.

If you used Exiftool you probably would not have that
problem... (which is ironic because you addressed that
comment to someone who apparently does use Exiftool).

Even my 10 year old P&S cameras have EXIF data that's undefinable
by name. And values in a dozen or so fields that's essentially
meaningless.

Well... make an image from an old P&S available for
download, and indicate which records in the Exif data
are "essentially meaningless", and lets find out if that
is true, or just ignorance... :-)

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

Cyberiade.it Anonymous Remailer wrote:
clandestin_=C3=A9cureuil wrote:

Yes, as I said the nonexistant nature of your "knowledge" is
self-evident.
=20
You said it, but it isn't so.

It is so, and you self-demonstrated it so.

Ok, if Exiftool and Photome are toys, enlighten me, what are the "real=20
deal" in such software?

There's tools like exifprobe and exiv2, but even they won't ferret
out everything (although they do a much better job than the toys
you're using). To be thorough you'd need to use software specific
to the camera in question. Most of it costs big pesos, and you
still can't be sure you're seeing everything.

So if it is almost impossible to ferret out this information then it
really isn't necessary to worry about it. Is it?


Bottom line is... EXIF is a wide open spec. It lets camera
manufacturers and software authors play fast an loose. Idiotic
statements like "camera and settings" are about as far from fact as
you can get and still be speaking about digital images generally.

Well I must be idiotic as well, as my only interest in Exif info is to
find out what camera, lens, place date, etc. plus those details that
show how the shot was composed. In fact almost everyone I know who uses
Exif data must also be idiotic, as that seems to be pretty much
universal as far as interest in Exif in my experience.

What does it feel like to be the only non-idiotic person among all us
idiots? Or otherwise as is more likely.


I have no interest in, nor have I suggested or implied that such a thing =
is=20
possible with any software program. I was asking someone else why they=20
wanted to change data other than time and place data. That data is usuall=
y=20

I couldn't care less what you think your "motives" are here kid,
I'm addressing a patently false statement you made. Period.

I must have missed that, and looking back I can't find it. Would you
like to point it out?

I see that she (or he, as in the case with Rita you never can be sure)
did say "the only real value in Exif data is to identify the camera and
its settings". Is that what you are taking exception to? I can't
understand why if that is the case. To her and many others, that is
probably quite true. To lawyers in a copyright law suit that may not be
true.

Are you taking issue with her mention of the techniques used to alter
it? Others have posted in the past few hours saying the same thing. You
can't do it with an off the shelf app, but you can sure alter any file
that uses fixed length fields, and with a little more effort, those that
have variable length fields and even checksums. I can do it, though I'd
be damned if I can see a reason to do so on more than one or two
occasions. If I don't want Exif data in my files I'll just strip them,
not bother fudging them.

I am also curious as to why the OP wants to spoof a file, but I doubt if
he'd give an honest answer.

PK

"Anonymous Remailer (austria)" wrote:
Floyd L. Davidson wrote:

Well, as an example all my Nikon DSLR's newer than about 2 years
read information from lenses and "encode" it in a 32-byte value. It
changes image to image even when the images are snapped in rapid
succession. What it consists of is "undefined". And there's several
fields where even the field name itself is a HEX value, so you can't
even know what that field pertains to let alone what the value
might mean.

You are confusing the information displayed by a program
with the Exif specification.

Wrong. I'm pointing out the fact that the EXIF spec is open ended
enough to allow camera manufacturers to include non-human-readable
and proprietary information.

That is not Exif data, and of course an Exif
specification cannot prevent a manufacturer from writing
data other than that specified by Exif to the same file.

*All* records (called an Image File Description) are
keyed to a integer value in the ID Tag (index field).
There is no "field name" in the data kept with an image.
Rather there is, in the specification, a text title
assigned to each index tag value.

Quite a flowery way of saying "field name".

There *is* a "field name", and it is *not* the ID Tag.
The point is that the text name is *not* in the file.
Never, ever. Only an ID Tag is, and that is the way
a specific record is identified.

In other words, it it has a text name it is necessarily
a specified record type, known to anyone who looks up
the current specification, but cannot be known just by
reading the data from the file.

And the spec be damned, My Nikons use non-text values as field
names. Which really highlights the point I was making all along.

Not only do Nikons use non-text values, *all*
identifiers in Exif data are non-text! Not a single one
of them has a field in the data for a text name of that
particular Exif data record.

Try using the "strings" utility to find various "field name"
entries in files that have Exif data. They aren't there.

If you used Exiftool

I do. And about a dozen other EXIF manipulation softwares. And
probably 20 different general image manipulation softwares. I've
even written a few special purpose tools from scratch, myself.

If you want to start trading "If you ever" barbs there kiddo, be
ready to pay the piper. NFGAA

There is no point in "barbs", traded or otherwise.

There is also no point in you making claims about having
"written" tools, because it is very clear that you do
not have that kind of knowledge of Exif data structures,
or even a passing familiarity with the Exif
specification.

Even my 10 year old P&S cameras have EXIF data that's undefinable
by name. And values in a dozen or so fields that's essentially
meaningless.

Well... make an image from an old P&S available for
download, and indicate which records in the Exif data

Not worth my time dragging one out of an archive. Sorry. Anyone who
has their own images and *decent* quality software tools can verify
it themselves. And they might learn something in the process.

are "essentially meaningless", and lets find out if that
is true, or just ignorance... :-)

If you can explain field names like 0x0029 and completely binary
data in the field itself, be my guest. Otherwise quit wasting my
time trying to blow smoke.

There is NO extry with a "field name" of 0x0029. Field
names are *not* hex numbers. That would be a ID Tag
(and there is none with the number 0x0029).

Whatever, I have no interest in wasting more time
explaining this. You want to trade barbs, I don't
care...

--
Floyd L. Davidson http://www.apaflo.com/floyd_davidson
Ukpeagvik (Barrow, Alaska)

Anonymous Remailer (austria) wrote:

Floyd L. Davidson wrote:

Well, as an example all my Nikon DSLR's newer than about 2 years
read information from lenses and "encode" it in a 32-byte value. It
changes image to image even when the images are snapped in rapid
succession. What it consists of is "undefined". And there's several
fields where even the field name itself is a HEX value, so you can't
even know what that field pertains to let alone what the value
might mean.
You are confusing the information displayed by a program
with the Exif specification.

Wrong. I'm pointing out the fact that the EXIF spec is open ended
enough to allow camera manufacturers to include non-human-readable
and proprietary information.

*All* records (called an Image File Description) are
keyed to a integer value in the ID Tag (index field).
There is no "field name" in the data kept with an image.
Rather there is, in the specification, a text title
assigned to each index tag value.

Quite a flowery way of saying "field name".

And the spec be damned, My Nikons use non-text values as field
names. Which really highlights the point I was making all along.

If you used Exiftool

I do. And about a dozen other EXIF manipulation softwares. And
probably 20 different general image manipulation softwares. I've
even written a few special purpose tools from scratch, myself.

Is there *ANYONE* here who believes this guy's BS?

Cal