D200 Compression?

[Coresounder]

O.K. after two days I broke out the manuel. I'm finding cool things like this can do WB fine tuning & bracketing. How to use the buttons, instead of going into the menu, ect..FUN!

A question though, I understand compression, like jpeg 1:4 or 1:8, ect.
However, there is another setting I don't quite understand. The menu is called
"Jpeg Compression" and there are two settings.

1. Size Priority - images ars compressed to produce relativity uniform file size. Quality varies with scene recorded.

2. Optimal image quality. File size varies with scene recorded.

Would someone please explain this? Thanks!

One thing I do know, when the setting is on #2 it cuts the memory in half!

 

[Murray Bowles]

compression

What I've read is that "quality priority" is what Canon has always done, and "size priority" is what Nikon used to do. From the text in the manual it seems that "size priority" adjusts the JPEG quality parameter to force more equal size files. But is "quality priority" just the JPEG algorithm run with a constant quality parameter? (I'm just inferring from the Photoshop GUI that there is such a thing as a JPEG quality parameter.)

I've been meaning to do some experiments to see in particular how "fine, size priority" compares to "normal, quality priority".

 

[AFS]

murray has it pretty well covered.
Simply:
Size priority: varied amounts of compression to get all files to be roughly the same size.
Optimal Quality: Minimizes compression (within the parameters you set via fine/normal/basic) and allows it to jump around in filesize in the interests of maintaining the most detail possible.

 

[Coresounder]

Thanks guys! Now I can see why the #2 setting takes up much more space.
So, the #2 setting would be the best setting, with less compression, if a person has the card space (and wants to use jpegs).

 

[AFS]

yep.
Since when I shoot JPEG I want the best file I can muster, I go for optimal quality myself, on all of my shooting banks (I've got all 4 set up- main NEF, main JPEG, B&W, and Night/long exposure) even when shooting NEF+JPG L/Fine.
Not good for my gallery's dial up friendliness though :smile:

 

[naisan]

Most image compression works by determining which pixels are, for instance, the same color in a row or column, and by only describing the colors in that image, row, or column. So a huge jpg of a single-color image would be tiny, where a jpg of an image with mutiple colors would be large, and also whether color variations ocurr in columns or rows of pixels also is important.

So that's why you have to choose which priority - there can't really be any guaranteed sizes, just preferences as to which takes precedence.

 

[Murray Bowles]

still remaining

(jeez i always have to make up my own title?)

is
1) is Canon mode (= quality mode) really a constant JPEG parameter?
2) how does Nikon mode (= size mode) decide what JPEG quality parameter to set for a given image?

 

[dhudspith]

Just a side note on the "Quality image mode". From my experience, the camera is using a farily generous file size estimate to calculate the remaining images on your CF Card. I guess if your subject has a great deal of detail in it with a lot of variation (meaning harder to compress) then the remaining shots is closer to correct. But if you are out shooting landscape type of shots with 30%-50% blue sky, they are overestimating. YMMV, just don't get too discouraged when the remaining shots drops in half - it may not be nearly that bad.

 

[naisan]

Answers?

(jeez i always have to make up my own title?)

is
1) is Canon mode (= quality mode) really a constant JPEG parameter?
2) how does Nikon mode (= size mode) decide what JPEG quality parameter to set for a given image?

I believe, not with any Nikon-camera specific knowledge, that:
1) Yes - here you are choosing a certain quality %compression. You could express it as percentage of original colors, for sake of explanation. So if you had 100%, then every pixel would have it's own color, and the size for a complex image would be, let's say 100 kb. Then at 50% quality, you would have a 50kb image because it chooses close colors for compression. So one very color-complex image might be bigger than another with very similar colors, but the percentage of color in the compressed images would be the same relative to the originals - The end results is setting it at 80, let's say, means you retain 80% of the colors in all situations with this setting.
2) size mode no doubt has a calculation appled in the processor to ratchet down the quality setting until it gets to the right file size. So if the target file size is 800k, it keeps moving down the quality until that file reaches less than 800kb.

Hope that helps, and again I don't know exactly how Nikon works this, but I'd assume the above is pretty darn close.