Mailinglist Archives:
Infrared
Panorama
Photo-3D
Tech-3D
Sell-3D
MF3D
|
|
Notice |
This mailinglist archive is frozen since May 2001, i.e. it will stay online but will not be updated.
|
|
Re: slide vs neg
Gabriel Jacob writes:
>>If negative film can record a far greater range of contrast than reversal
>>film, then why don't those computer types use negative film then scan the
>>negatives, as if they were positives, then let Photoshop, or a comperable
>>program mathamatically reverse them, thus keeping all the contrast?
>>
>>Grant
>
>Actually they do! Problem with amateur type scanning of negatives is getting
>rid of the orange tint. Don't know what the pros do about that.
>There was a good article (pg.42) about "keeping all the contrast" in the
>Sept 96 issue of Popular Photography. This was the How To issue, that also
>had a short article about how to achieve your first 3-D fuse and make your
>own stereographs. This article was discussed here on P3D at the time.
By "the orange tint" I assume you mean the color in the film base. Yes,
this is one of the problems in scanning negatives: Not only is the
overall
transform non-linear, but the red, green, and blue channels are nowhere
near equivalent!
It should be pointed out here that handling these anomolies requires
more than 24 bits per pixel: If you do a "raw" scan at 24 bits and then
adjust the curves to get a suitable positive image there will be a great
deal of missing and duplicate codes, effectively reducing the image to
far less than 24 bits. I use a 36-bit scanner for negatives. This is
more
expensive than the 24-bit scanners, so in addition to being easier to
do,
getting a good quality scan is cheaper with a transparency.
BTW, Photoshop and every other image editting program I've seen on
PCs or MACs only handle up to 24 bits, so you have to do all the
major adjustments in the scanning process, not after.
Someone questioned the numbers of "stops" available on a scanner.
While there certainly is a limit on the density range of a scanner, the
bigger question is how many bits (how many grey levels) are needed
per stop? To be virtually lossless in scanning slides I've heard you
need
64 levels per stop. So a 6-stop range would be 384 levels, which is
8 and a half bits. But this is in the linear world (slides). If your
"bits"
are linear and the film isn't (negatives) you need to match the worst
part of the slope....
Ah heck, I'm rambling again. Summary: Negatives are good, but
hard to deal with, and expensive.
Greg Marshall
------------------------------
|