D
Doug Kerr
Guest
. . . an image.
A little philosophy and a little semantics this morning, while I wait for Carla to return from her morning convention session (a so-called "breakfast") so we can go to (real) breakfast.
We often discuss here the implictions of processing photographs beginning with either (a) a raw data file or (b) a JPG image file.
It's is important that we remember that the properties of (b) with which we are most often concerned here is not that the image is represented in JPEG form but rather that it is an image. That is, the file describes the color of every pixel in an image. By contrast, (a) does not describe an image. It is a suite of data from which we can infer an apprximation of the color of every pixel in an image representing the taken scene.
We do that, of course, through the process of demosiacing, a central part of the broader process we often (and aptly, in my opinion) speak of as "developing" the image.
Of course, most digital cameras only provde for the delivery of an internally-developed image in JPEG form, (although I have a camera that also allows delivery of an internally-developed image in TIFF form).
I don't mean to suggest that the outlook I point out should have any inflence on how we proceed. I just think it is just important to note that, for the most part, the properties we ascribe to the JPEG output of a camera, in the context of how we process phtograph, come not from its JPEG-ness but rather from its image-ness.
Similiarly, there is a tendency to think of the external software we use to "develop" our images from the raw data as "converting the image from raw to JPEG form". Actually, what they do is convert the results of our phtography from raw form to image form. We can then of course choose to save that image (if we care to) in a number of forms. (In many cases, we may make substantial further changes to it before we save it in any form.)
Well, my breakfast date is here!
A little philosophy and a little semantics this morning, while I wait for Carla to return from her morning convention session (a so-called "breakfast") so we can go to (real) breakfast.
We often discuss here the implictions of processing photographs beginning with either (a) a raw data file or (b) a JPG image file.
It's is important that we remember that the properties of (b) with which we are most often concerned here is not that the image is represented in JPEG form but rather that it is an image. That is, the file describes the color of every pixel in an image. By contrast, (a) does not describe an image. It is a suite of data from which we can infer an apprximation of the color of every pixel in an image representing the taken scene.
We do that, of course, through the process of demosiacing, a central part of the broader process we often (and aptly, in my opinion) speak of as "developing" the image.
Of course, most digital cameras only provde for the delivery of an internally-developed image in JPEG form, (although I have a camera that also allows delivery of an internally-developed image in TIFF form).
I don't mean to suggest that the outlook I point out should have any inflence on how we proceed. I just think it is just important to note that, for the most part, the properties we ascribe to the JPEG output of a camera, in the context of how we process phtograph, come not from its JPEG-ness but rather from its image-ness.
Similiarly, there is a tendency to think of the external software we use to "develop" our images from the raw data as "converting the image from raw to JPEG form". Actually, what they do is convert the results of our phtography from raw form to image form. We can then of course choose to save that image (if we care to) in a number of forms. (In many cases, we may make substantial further changes to it before we save it in any form.)
Well, my breakfast date is here!