RAW for K10D

Singinman · 02-26-2008, 03:22 PM

When I first got my K10D, I opted for the Pentax RAW option using the RAW converter with Elements 5 for processing. When I got Elements 6 with the updated Adobe Camera RAW, I took one series with the camera set for RAW DNG. I was blown away with the difference in the final result. The final images were richer and sharper than anything I had ever done before. I thought about this and decided that since Adobe set the specs for DNG they should have better algorithms for post processing.

Has anyone else player with DNG. If so what do you think?

*WMBP* · 02-26-2008, 04:07 PM

Has anyone else player with DNG. If so what do you think?

Not sure I understand the question or what you're getting at.

The sensor sees what it sees, in the way it sees it. This is not influenced by the choice of format for the subsequent storage of the data thus "seen". I have been told by an engineer at Adobe that a DNG file contains exactly and fully the same data as a PEF. Now, I was asking whether the files store the same data because I wondered if it made any difference if I shot PEF and converted, and he assured me that it did not. The conversion (which I do in Adobe Lightroom) retains every piece of data in the original PEF, which is everything the sensor originally saw. This makes sense to me. Note that this does not really speak to the question of whether different data might be stored if I shot to DNG in the first place, but I find it inconceivable for several reasons that the choice of file format would make any difference to what is stored. It matters only to how its stored. On a K10D, DNG is not compressed, which means that it writes slightly faster and the files are slightly larger. But other than that, no difference.

Now, the color space (sRGB vs Adobe RGB) can make a difference of some sort. I've read that, in theory, Adobe RGB is better in some potentially significant way. But Adobe RGB seems to cause other problems, with online printing services and other things. So I never use it.

I might add that different programs do seem to render raw data slightly differently. I noticed the other day that a photo that looked quite nice in Lightroom looked less nice in Light Crafts LightZone. Different programs use different presets to interpret the raw data. But again, this doesn't make any difference to the underlying data. I think you should be able to get an image to look pretty much the same in just about any program.

Finally, keep in mind that the camera's jpeg settings for contrast, etc., do make a difference to the jpeg preview that is embedded inside your raw files. However this is simply a matter of the initial preview of the photo when you get it to your computer, and it does NOT make any difference at all to the actual raw data.

Not sure whether this speaks to your question or not. If it doesn't, let us know and perhaps somebody will understand better than I.

Will

*WMBP* · 02-26-2008, 04:15 PM

By the way, I might mention my own workflow here, while I'm just throwing info at the wall and hoping some of it sticks. ;-)

I almost always shoot PEF on my K10D. Later, when I get the files on my computer, I open Adobe Lightroom and convert the images to DNG. It's the first step in my workflow.

Why bother with this two-step hassle? One reason only: file size. On the camera, because the PEF files are compressed and the DNG files are not, the PEF files are significantly smaller and I can get more shots per 2 GB card. But once I get 'em to the computer, in the conversion process, Adobe Lightroom apparently compresses the DNG that it creates, so these converted (and presumably compressed) DNG files that I end up with are even smaller than the original PEFs. If I use DNG in the camera, not only are the files bigger on the card, I'm stuck with 'em at that size, because Adobe Lightroom won't convert uncompressed DNG files to DNG files.

Shooting DNG in the camera seems to have only one advantage: Because the files are not compressed, there's less calculation involved in writing them to the card so you can shoot very slightly faster. I proved this to my own satisfaction in tests (which are discussed somewhere else in this forum) but I don't find the speed benefit anywhere near as important as the file-size benefits of PEF originals, so I never save as DNG on the card.

Will

*christinelandon* · 02-26-2008, 09:35 PM

Since I print most of my own stuff, I've been shooting in DNG and find the result works fine with my established Photoshop workflow. If I send a file out, I always ask first what requirements the lab has, and, if necessary, tweak the file before saving in sRGB.

chrisman · 02-27-2008, 03:28 AM

I always use DNG. Can't say that I've ever done a direct comparison with PEF, but I chose DNG because it is likely to outlast most other RAW formats and will probably become the standard for RAW. The results are always excellent, even if my photographic skills are not!

It also saves having to convert PEF to DNG later - anything that saves time in post-processing gets my vote. Files are bigger, of course, but that doesn't worry me since I am not 'trigger happy' with my K10D, or any camera for that matter

procyon · 02-27-2008, 04:24 AM

Same thing has been noticed by other people, I remember someone claimed that it was all about Adobe using different processing for DNG files (possibly to promote their format) that tends to give more pleasing results.
Actually it should be quite easy to test.
Shoot in PEF, use Adobe DNG converter to convert PEF to DNG and then process both files with exactly the same settings. If there are differences then Adobe does some magic in case of DNG files. I don't have photoshop myself, so someone care to test it ?
By the way I remember I have seen somewhere (here ?) a test of PEF vs. DNG (I mean the in-camera DNG here) and the results tended to be very software specific.