[justinr]

I think it slightly nonsensical to plant oneself too firmly in either the RAW or jpeg camp. Both have their uses, RAW for the quality and extended potential for manipulation, jpeg for the convenience when the lighting is right and you know there won't be much post processing required on the 500 shots you've just taken.

Coming back to horsey shots and the K10, I had hoped for an improvement on these which were taken with the ist a few weeks back, ( I think I may have posted this link before, but not too sure)-

Pony club.

[GoremanX]

I think we are chasing each other in circles here.

Yes indeed jpegs are movable feasts to a certain degree but their approximate size can be decided upon by setting limits within the compression algorithms which is what we are doing when choosing the jpeg quality level.

That "approximate" file size can vary greatly depending on the amount of detail in the scene. Blowing highlights is one way to reduce detail.

I just took 2 pictures of the exact same thing. Camera is set to 14.2 MP, 4 star quality JPG. Exposure settings are identical for both pictures with the exception of shutter speed.

One is obviously blown, the other is not:





The blown one was 8.2 MB in size. The "good" one was 10.8 MB in size. That's a 25% difference. Even now that they're resized to 800px wide, the blown one is 20KB smaller than the "good" one. Both were re-saved with identical settings.

If the file is 10 MP in size, then it was captured, processed and recorded as a 10 MP file. The camera doesn't fudge a 10 MP file after using fewer pixels.

But again, since we don't have a file with all retained EXIF data, this is all guessing.

[justinr]

Given the above I've been back to the folder of files are sure enough the darker ones tend to be bigger jpegs, until they become obviously underexposed and then once again they shrink. Still no sharper though whatever the file size. However, the largest jpeg I have found in the bunch still has blown whites in it!

I must put my hand up and admit that this larger one is around 3.8mb, which is still a lot smaller than average suggested by the manual.

[GoremanX]

I think it slightly nonsensical to plant oneself too firmly in either the RAW or jpeg camp. Both have their uses, RAW for the quality and extended potential for manipulation, jpeg for the convenience when the lighting is right and you know there won't be much post processing required on the 500 shots you've just taken.

RAW+! RAW+! RAW+!

That's my cheer. Memory is cheap, even the *ist D can use SDHC cards (with a firmware update). The biggest advantage to JPG-only is that you can fire off more shots before filling the buffer.

My 8 GB card can hold almost 300 shots in RAW+. I expect you'd be able to hold twice that easily on a K10D.

I routinely use both formats for my needs. I quickly resize and upload JPG files for this forum (as I did a couple posts ago), and I always have the RAW file if I happen to have accidentally captured some award-winning shot (which has yet to happen).

Now we're getting away from our topic again.

[GoremanX]

As for the sRGB/AdobeRGB discussion, I asked my tutor's opinion on it. His reasons for not advising use of the sRGB setting in the camera is :- Because sRGB is a smaller colorspace than AdobeRGB, if you set it to that, there's no changing the colorspace if you want to for editing purposes. There should be no noticeable difference in an uploaded to Web image originally shot in AdobeRGB as long as it is set to sRGB before being uploaded.

That would technically be true if the JPG files generated by your camera weren't so crap to begin with. Even at 4 star quality, Pentax uses lower-grade subsampling for the JPG compression algorithm. There's a whole lot of reasons why having the camera set to AdobeRGB is a bad idea. For some of them, I refer you to this thread (which you've had to moderate extensively):

https://www.pentaxforums.com/forums/pentax-dslr-discussion/86233-regarding-adobe-rgb.html

For some of them, I refer you to this thread (which you've had to moderate extensively):

https://www.pentaxforums.com/forums/pentax-dslr-discussion/86233-regarding-adobe-rgb.html

I'm not going there again.

All I know is, I was shooting Jpegs with sRGB setting, I am now shooting RAW and using AdobeRGB setting. When I convert the RAW files to JPEG and load them here for example, I set them to sRGB and I'm not seeing anything different than I did before. If anything, I think the pictures look better than before. I have also worked with JPEGS shot in Adobe RGB and see no problem with the images when set to sRGB.

[GoremanX]

All I know is, I was shooting Jpegs with sRGB setting, I am now shooting RAW and using AdobeRGB setting. When I convert the RAW files to JPEG and load them here for example, I set them to sRGB and I'm not seeing anything different than I did before. If anything, I think the pictures look better than before. I have also worked with JPEGS shot in Adobe RGB and see no problem with the images when set to sRGB.

If you're shooting RAW, it really doesn't matter what your camera is set to. It doesn't change the RAW file at all. RAW files have no colour space. The colour space setting of the camera affects only the JPG output, nothing else.

So essentially what you're saying is, your pictures look better now that you're shooting RAW instead of JPG... mega-duh, I could've predicted that result without resorting to my battery-powered crystal ball

If you're shooting RAW, it really doesn't matter what your camera is set to. It doesn't change the RAW file at all. RAW files have no colour space. The colour space setting of the camera affects only the JPG output, nothing else.

So essentially what you're saying is, your pictures look better now that you're shooting RAW instead of JPG... mega-duh, I could've predicted that result without resorting to my battery-powered crystal ball

Reread my last line, the same applies to JPEGS shot in AdobeRGB and set to sRGB.

[justinr]

That's my cheer. Memory is cheap,

Maybe, but time isn't. which probably means I've lost a fortune hanging around forums.

[GoremanX]

Reread my last line, the same applies to JPEGS shot in AdobeRGB and set to sRGB.

oh I agree, if you convert AdobeRGB JPGs to sRGB, they work fine... they even work fine as AdobeRGB! But now you're getting into the technicalities of colour space embedding methods, and you don't want a technical discussion. Just keep in mind that Photoshop is not the only graphics application out there, the number of them are vast and cover all kinds of needs, from advanced editing to simple viewing, and a lot of them do not support the way Pentax (and most camera manufacturers) embed the AdobeRGB colour profile. It's nice to be able to say "well *I* don't see the difference when I use *MY* application", but that doesn't apply to everyone.

If someone can be bothered to convert an AdobeRGB file to sRGB for posting on the web, they can also be bothered to process a RAW file into sRGB in the first place and get better results (without having to recompress the JPG in the first place). That's my first generalization of the day since I woke up.

But now you're getting into the technicalities of colour space embedding methods, and you don't want a technical discussion.

Checking a box is technical? Wow, I'm more advanced than I thought I was.

[GoremanX]

Checking a box is technical? Wow, I'm more advanced than I thought I was.

I'll pass that off as sarcasm and NOT type up a 1000 word rant!

[sterretje]

Attached are four very basic examples of jpegs created with the gimp (640x400 pixels).

The plain white one is 1805 bytes, the one with one black block is 2870 bytes, the one with the 2 black blocks is4396 bytes and the one with the black and red block is 5903 bytes.

This illustrates that the filesize of the jpeg compressed file depends on the content.

I will add two more examples in the next post as well as a conclusion.

EDIT: the filesize have changed as I have added a border around them and are now 6043, 7980, 8083 and 9666 bytes

[sterretje]

Previous examples are not quite real life, so I also created 2 grey scales of the same size.

The one with the white part at the top 'simulates' your photos with blown out sky and is 34775 bytes. The other one is 50160 bytes.

The conclusion that I draw is that the sizes that you get are small because the pictures are lacking detail.

[justinr]

The above shot is from the K10. I took a series of them in the various focusing modes with both the K10 and a couple from the ist.

The k10 was set to record 10 million pixels at 3 star quality. The ist was also set to record at maximum quality.

The k10 jpegs ranged from 2.87mb to 3.25 mb from the same spot on a tripod, same focal length but differing shutter speeds as the sun put in an appearance. As it got brighter the file size increased.

The ist with 4 million less pixels produced files of 2.9 and 2.92mb.

Just to complicate matters the the ist consistently produces darker (underexposed) images compared to the K10 at similar settings and I wonder if the ist jpegs would have been bigger if they were brighter.

The sharpness (which was the original point of the test) was pretty much indistinguishable between cameras.