35 mm Summaron/2.8 samples/opinions

[onnovisser]

Brian,

man.. I speak Chinese, Swahili, French, Dutch, Frysian, German and English but I have no idea what that last sentence means But hey if you wrote it in an add I'm sure I would feel the need to buy your product!

 

[zeos 386sx]

With all due respect to the future, given their current state of development, camera CCD's look limited next to film.

Brian states, "It would be possible to make the amps and/or A/D's have a response curve in them, but I do not know if this is commonly done." I would bet it is expensive to do that or the camera companies would have already done it.

Would I be correct in assuming that B&W will be the biggest loser in the digital revolution? Color film has always required limited contrast and photographers have adjusted to that fact, using fill flash, reflectors or whatever to reduce contrast. B&W photographers, on the other hand, have enjoyed much more exposure latitude. It sounds like those days are quickly passing.

I can see an increased need for spot metering. The photographer is going to have to know exactly where image starts and "clipping" leave off - on both ends - and judge which end is the most important to the picture.

 

[Sonnar Brian]

I'll ask around. 3dB is ~2:1, or one more bit per pixel. so 8 bits is 256 levels of grey, or 24dB. 12 bits is 4096:1, or 36dB. Normally 80dB would require 27bits of resolution using a linear amp. You just can't get 27Bits!

I'll ask the guys in the lab next door how their digital camera is doing. A few years ago they built a custom 100MPixel camera to fly on an F14.

Uh Oh. Did I make peoples Heads Explode? The optics guys used to do the same to me with 10 to the power of numbers over 10. Then I found that 3dB was a factor of two, and every bit that you add to a number increases its range by 3dB.

 

[doubs43]

Then I found that 3dB was a factor of two, and every bit that you add to a number increases its range by 3dB.

+3db is a 100% increase while -3db is a 50% decrease. In audio, the human ear requires a 10db difference to actually hear a 100% increase in volume or a 50% decrease in volume. I remember this from my Air Force Technical School training. Applying it to pixels is something I'd have to study to really understand.

Walker

 

[phototone]

Phototone,

Thanks. If I understand you a CCD behind a contrasty lense is like transparency film. You have to expose for the highlights - with the caveat that you have to make a second exposure for the shadows. Yikes, this is the future of photography???

It sounds like the camera companies are going to have to work overtime to make CCD's with the versatility of negative film.

What kind of conditions cause "clipping"? I assume from what you have said that it would be high constrast situations - with high contrast lenses just making it worse.

What is the truth is about film v. CCD contrast/density ranges. With the zone system and controlled development film can have a broad density range. I assume from what Brian said (i.e. "The response of the CCD is linear") that CCD's are strictly limited to a built-in range. Are there other work-arounds, like Fuji's, to get contrast rendition closer to film?

Well, different camera companies have different approaches to exposure and contrast with their digital cameras, and also different imagers have different levels of range before they clip.

I have a Kodak SLR/n 14 mp digital slr, nikon mount. It has a very wide dynamic range, perhaps wider than some films, and you can recover highlight detail very nicely from overexposed areas up to a point. The downside is that the Kodak is really only a low ISO camera, as its imager is "noisy" and if you underexpose and then try to bring up the image in Photoshop, the shadows get "blotchy" looking. The base ISO of this camera is 160. This is one of the only digital SLR cameras that it would be better to "overexpose" than to "underexpose".

It is true, as a general rule, that it is possible to get clipping on both highlights and shadows. However, you can, with film, overexpose and underdevelop and get a wider tonal range than possible under normal conditions. Of course you would have to do this for the entire roll. In that event, possibly some of your images might be low contrast, but if you scan them you can bump them back up.

Films ability to record shadow detail is somewhat dependent on the base density of the film stock, as a film that has some "grey" to the base, such as Tri-x, the darkest shadow detail could get lost.

As a general rule though, the modern Digital SLR's and the Epson RD-1 (which uses the same imager as many Digital SLR's) are not as limited in tonal range as transparency film is, but are not quite up to dealing with correction from gross overexposure as negative films are. If your exposures are correct, then the tonal range of the images from digital files from good digital cameras are every bit as good as film, and sometimes better. You just have to get your exposure correct, just as you have to do for film transparencies.

As far as your remark above about making two exposures to get full tonal range with digital. While this is possible for extreme circumstances, it is not needed nor desired for common subject material. The tonal range of good interchangeable lens digital cameras is good right where it is. The images may not look just the way you want them right out of the camera, but can be modified in Photoshop to achieve excellent results.

Never-the-less, I still prefer film for my personal work. I like to work with film cameras, and I like to develop film, and I like to make scans from film.

 

[zeos 386sx]

I have a friend in the wedding photography business. Taking candids of a sun-lit bride in a white dress against a dark background has to qualify as an "extreme" circumstance. Watch your background!

One more question...

Does changing the ISO affect the contrast range of a CCD or just shift the "clipping range" up and down?