Hank E wrote in post #4236943
I didn't realize shooting at a higher ISO meant fewer images per CF card. For a H.S. football game last Friday night, I bumped the ISO up to 3200 and noticed that it dropped my available images down to 379. Usually I shoot at 1600 and get 422 shots. My thought is there is more "information" being recorded for each image, therefore fewer images per card. Any ideas, guys?
Also, why do you only get the amount of images the card says are initially available, since images are all different sizes?? When I format a card and shoot jpeg large, fine at an average of ISO 400, my size files are anywhere from 3000-5000kb. If all my images were 3000kb, I'd still only get 481. Yet, if they were 5,000 kb, I'd still get 481 (I believe). Any thoughts on this?
Just curious....
Hank
Some images are more compressible than others. JPEG is a compressed format, so each image will be compressed as its written to the card. Due to the compression algorithm, some photos will be able to be reduced by a greater factor than others, and hence you will get smaller files, and more images per card. The "shots left" counter seems to be quite pessimistic on this. I shot a day's motorsport, and took 407 frames, on a card which at the beginning of the day said had capacity for 390. I don't remember how many frames were estimated to be left, but I think it was still double-digits.
Where the surprise for me came in is that there is a degree of compression applied to the storage of RAW files. Naturally, this is not a lossy compression like JPEG, but it does still contribute to a fairly wide variance in file size. The above-referenced day was shot in RAW with no JPEG.
If you want to understand more about what makes an image compressible, you need to know in detail the workings of the compression algorithm. I don't know enough detail to comment authoritatively, but I do know that JPEG does best when it can map the image sine waves, which is why it's so bad at compressing very high contrast.
The link between ISO setting and # images is not down to detail, it's down to the noise in the image. Sensor noise isn't very compressible because it is random colour speckles that have little or nothing in common with near neighbours, and therefore cannot be "aggregated".
