Using my D200 with 1 Gig Kingston Card, in J Peg, Large, Fine, the camera shows 112 in the numbering system. In use the Card only indicated that it was 'full' when it reached 243. Any suggestions please. Can you also assist with regard to the indicator panel on the SB800 which on the right hand edge shows a camera outline wth a twin ended arrow above it. I cannot find any reference to it,it does not show up when the gun is used on the D100. Many thanks.
Pretty sure the camera counter is based on an average picture size. If the image sizes are variable (i.e. more or less color) then the number you can squeeze on the card will also vary. An all-black JPG will take up less space than all-white.
>No, both would compress to exactly the same size file. > >-Ade
Not in-camera they won't. Black = absence of any color information (0,0,0). White = color information present (255,255,255). Just verified w/ an informal test. One shot with lens cap on and one of a white door completely out of focus (to remove detail). Lens cap = 201 Kb. White door = 2+Mb. "Color" yields even larger images. Test this in the "real world". Stand in a windowless room (inside bathroom or closet works well), close the door, take a shot (say, 1/80). Now, take the same shot w/ the flash. More color information = larger file (201Kb vs 2.8Mb for me).
That test is completely invalid, because two completely different subjects are used. Find a black lens cap and white lens cap from the same manufacturer (same size, style, etc.) Take a picture of each, properly focused and exposed in good lighting. Their jpeg sizes would be very close to each other.
In fact here's an easy test everyone can do: a) take an existing real-world photograph, properly exposed. b) open in Photoshop (or similar tool) and save a copy as full-color jpg. c) open the original file again, and convert the image to black & white (ie., gray scale). d) save the gray scale image as jpg, using the same compression settings as step b. Repeat the test for a few different images to discard anomalies.
You'd find that even by dramatically throwing away ALL color information, the average full-color jpg vs. b&w jpg sizes are within 15%-20% of each other.
On the RGB (0,0,0) vs (255,255,255) thought, here's a real-world hint: there is a reason jpeg encoding customarily uses YCbCr, not RGB.
No, it's not invalid because a white lens cap would be black to the sensor if it was on the lens .
I think the disagreement lies in the JPG compression setting on the D200. Mine is set to "Optimal Priority" where I imagine yours is set to "Size Priority". Since an all-black image has less information than an all-white, the D200 can compress the image significantly. An all-white image would not fare as well under extensive compression (from the camera's point of view). The "bathroom test" I mentioned above resulted in more uniform image sizes on "Size Priority", though there was still a 20% increase in file size when the flash was used to illuminate the room as opposed to total darkness.
Simply put, black takes less space in a JPEG than any other color because the JPG algorithm compresses black most efficiently (because there's no information for the pixel). As the number of black pixels in a JPG rises, the file size will decrease. You yourself verified this with your test, yielding file sizes that had a 15-20% size difference. This is significant. That's 15-20 more shots for every 100.
So I stand by my original statement, which is supported by both your experiments as well as mine: "An all-black JPG will take up less space than all-white." Perhaps my mistake was not beginning the statement with "In the D200". I assumed it was implied by the OP's topic.
PS, I might also add that black & white, from the perspective of the JPG compression algorithm, is NOT greyscale. Black and white uses only two colors. Greyscale uses at least 4, commonly 256+. Try running your experiement again with true black and white (i.e. "on or off")
>Simply put, black takes less space in a JPEG than any other >color because the JPG algorithm compresses black most >efficiently (because there's no information for the pixel). >As the number of black pixels in a JPG rises, the file size >will decrease. You yourself verified this with your test, >yielding file sizes that had a 15-20% size difference. This >is significant. That's 15-20 more shots for every 100.
I am sorry but the above statements are absolutely incorrect. Perhaps one should review the way jpeg streams are encoded before making further claims.
The test I proposed above compares full color jpgs against grey-scale jpgs, when the chrominance channels are discarded. It does NOT pertain to the compressibility of any color (black included) against another.
True. I stand corrected. Did not understand JPG like I thought. Thanks for making me "look it up".
That being said, I need to eat some crow and rephrase my original statement so that it's more accurate and addresses the OP's original question. Images that are largely uniform in color will take less space than those that are "complex". An image of a blue sky will compress better than an image of a rainbow. As such, if you take a large number of picutes of grass, you'll be able to squeeze more on a card than if you took pictures of a jar of marbles...
Would that be more accurate?
And honestly, thanks again for not just moving on Ade. I'm a hard-headed SOB, but in the end, I'd more interested in learning what's correct than maintain some silly sense of pride. I learned about JPG many many years ago when it was still new, so whoever "taught" me had it wrong.
The other question asked was about the symbol in the SB-800's display. That symbol means the flash is being used with a camera compatible with Nikon's creative lighting system (CLS). The D100 isn't compatible with CLS and thus no symbol on the SB-800 when mounted to that camera.