Not MP but DR, stupid

Now every camera maker have cameras with lots of pixels. Even Samsung.

It is time to concentrate on Dynamic Range. There is a lot to do in that.

I want to hear about Canon cameras able to distinguish and record 16bit images in 11-12EV range.

--
Andy
http://www.flickr.com/photos/8170099@N02/

Amen.

alap said:

I want to hear about Canon cameras able to distinguish and record
16bit images in 11-12EV range.

Click to expand...

if it would be 1 fps I wouldn't care (think before you click), but an increase of dynamic range would help so much more, and yes I wouldn't mind 16MP, but dont necessarily need it.
--
http://www.richardbakker.com

The megapixel war is ridiculous. Make it stop.

DR is where it's at. I use a couple of Fuji S5's that rock in that area. I love to see a Canon 5dII or whatever with DR that equals my S5's. I'd be on board in a flash.
bl

and while I am sometimes frustrated about it's lack of fine detail resolution (it's, after all, a 6+6MP cam, not a 12MP), I have to admit each time I compare it's pictures with the one of my 5d that the Fuji makes just more appealing, more photogenic pictures, click by click. Up to A3 (11x16) size, those photos are plain gorgeous.

Would Fuji come with a true 12MP high DR camera, possibly with FF sensor, I'd be happy with it for years to come and had very little need in participating in the ever- ongoing "upgrade- game".

Unfortunately Fuji's commitment to the camera market seems very doubtful. Their biggest weakness being that they have no camera body and no lens mount of their own...

Bernie

One could make your argument with regard to the 1Ds mkIII - maybe, but I'd like to see 16 MP or better in my next 5D. Since Canon doesn't make dedicated cams for just portrait or just landscape etc, their cams have to exceed in some cases some folks' requirements. To me, these side by sides illustrate the value of extra MP especially for landscape shooting.

http://www.pbase.com/image/90999838/large

http://www.jimroofcreative.com/II-vs-III.html

alap said:

Now every camera maker have cameras with lots of pixels. Even Samsung.

It is time to concentrate on Dynamic Range. There is a lot to do in
that.

I want to hear about Canon cameras able to distinguish and record
16bit images in 11-12EV range.

--
Andy
http://www.flickr.com/photos/8170099@N02/

Click to expand...

alap said:

Now every camera maker have cameras with lots of pixels. Even Samsung.

It is time to concentrate on Dynamic Range. There is a lot to do in
that.

I want to hear about Canon cameras able to distinguish and record
16bit images in 11-12EV range.

--
Andy
http://www.flickr.com/photos/8170099@N02/

Click to expand...

--Practically speaking could it be even fit into the box, and would it be worth the added time and electrical energy needed? Even at 14 bits current Canon models dynamic range exceeds that of film.
-nothing beats a fast lense, except a fast girl-

--
-- richard

I don't thin upresing the 1DS II image is fair to that camera. Upresing anything would make it blurrier.

Maybe you wanted to down res the 1DS III instead to show that the lens can still outresolve the pixels?

Kaz

Rick Knepper said:

One could make your argument with regard to the 1Ds mkIII - maybe, but I'd like to see 16 MP or better in my next 5D.

Click to expand...

Indeed. For all practical purposes, the 1Ds III can be considered to have better resolving power than most 35mm film stocks (and even approaches medium formats), but it's really the only one.

The goal, for me, is to completely replace film, and if the camera doesn't even have the resolution to do that, then it's useless to me, personally. Without at least 14-bit color and 16MP, I'd think of all of the great shots that could have been better served by film, and it would haunt me.

The so-called "megapixel wars" won't end until all DSLRs exceed film's capabilities. That seems reasonable. (16-bit color would be nice, too.)

--

'Passion will make you crazy, but is there any other way to live?' - Kara Saun

You have to up the resolution to see the actual difference. Plain and simple.

There was a post a few months back where two shots, one from a 5D and one from a 1DsIII were compared and a lot of people were fooled into thinking that the 5D showed more detail. That was only because the 5D image did not have the same image features shown at the same size as the 1DsIII.

Look at it this way. You have a poster to print and your printer requires the file to be roughly 3744x5616. The 1DsIII makes this size natively. The 1DsII will need to be resized to get to the printer's specifications.

So, my illustration shows the difference in the detail between the two cameras when the size of the image is to be reproduced at 3744x5616. Of course, that gives the 1DsIII the advantage... but it is a real advantage. If we scaled the 1DsIII down to the 1DsII then we are throwing away detail. Now that wouldn't be fair at all since that discards one of the strengths of the 1DsIII.

--

'Truth is stranger than fiction, for we have fashioned fiction to suite ourselves.' G.K. Chesterton

http://www.jimroofcreative.net

Unless one is fixated on pixels and not on images, high DR and high pixel density are not in conflict.

Consider camera A whose pixels are half the size of those of camera B, but having the same quantum efficiency (light-gathering ability per unit area ). With equal area sensors, camera A has four times the number of megapixels as camera B. In the space taken by one pixel of camera B there are four pixels of camera A. The same number of photons captured by one pixel of camera B are spread over four pixels of camera A.

Naively, a measurement of DR would say that camera A has two stops less DR than camera B (since a single pixel only captures 1/4 the light, and the smaller pixels will max out at a light level proportional to their area for a given technology). Yet both sensors are capturing as many photons per unit area, the only difference being that they are resolved better spatially by camera A.

Camera A has the same light gathering capacity as camera B, and better resolution. Its main drawback is the size of its raw files and the processing overhead (both in cam and in post) that entails.

Dynamic range specs should be quoted on a per area basis if you want to compare them fairly across camera models with different pixel densities. Perhaps one should also divide by the square of the crop factor, to normalize DR relative to overall sensor area.
--
emil
--



http://theory.uchicago.edu/~ejm/pix/20d/

except for all the supporting circuitry around each pixel, which eats up a fair amount of that light-gathering area

--
http://www.pbase.com/thejaybird

thejaybird said:

except for all the supporting circuitry around each pixel, which eats
up a fair amount of that light-gathering area

Click to expand...

Not so:

http://forums.dpreview.com/forums/read.asp?forum=1000&message=24175679

Now of course the 1D2 used as reference in that post is actually one of the worst performing Canon DSLR's on a per area basis. With the 1D3, QE is roughly a draw with the FZ50 (on a per area basis).

--
emil
--



http://theory.uchicago.edu/~ejm/pix/20d/

Frankly speaking my personal felling is that resolution of 5D is beyond that of my old ElanIIe. But I would say that DR from slides was way better. Do not even mention B&W.

--
Andy
http://www.flickr.com/photos/8170099@N02/

These days, DSLR's are almost commodities.....in order for one brand to stand head and shoulders over the crowd, that brand will need to expand the DR lots, and I don't mean a tad like the Fuji's and Sigma's...I mean LOTS!

...like at least 3 stops wider for starters ;-)

And if you think about it, the DR of the 2001 Canon D30 is not much narrower then the 1DS Mark III...DR is one one attribute of DSLR's that has imrpoved the least, and this is sad.

--
Lanned For Bife

ejmartin said:

thejaybird said:

except for all the supporting circuitry around each pixel, which eats
up a fair amount of that light-gathering area

Click to expand...

Not so:

http://forums.dpreview.com/forums/read.asp?forum=1000&message=24175679

Now of course the 1D2 used as reference in that post is actually one
of the worst performing Canon DSLR's on a per area basis. With the
1D3, QE is roughly a draw with the FZ50 (on a per area basis).

Click to expand...

Yes, you can compare apples to oranges and draw any conclusion you like. The fact is that there exists supporting circuitry for each pixel, and it does take away from light gathering area. The more pixels you have, the more you pay for this overhead (slight though it may be).

--
http://www.pbase.com/thejaybird

thejaybird said:

ejmartin said:

thejaybird said:

except for all the supporting circuitry around each pixel, which eats
up a fair amount of that light-gathering area

Click to expand...

Not so:

http://forums.dpreview.com/forums/read.asp?forum=1000&message=24175679

Now of course the 1D2 used as reference in that post is actually one
of the worst performing Canon DSLR's on a per area basis. With the
1D3, QE is roughly a draw with the FZ50 (on a per area basis).

Click to expand...

Yes, you can compare apples to oranges and draw any conclusion you
like. The fact is that there exists supporting circuitry for each
pixel, and it does take away from light gathering area. The more
pixels you have, the more you pay for this overhead (slight though it
may be).

Click to expand...

Well of course circuitry takes space. The question is how big is the impact, and I merely gave a real world example that the FZ50 sensor, scaled up to APS-H size, would have on the order of 85MP and gather the same amount of light as the 1D3.

Those proclaiming it's all gloom and doom with 12 or 14 or 16MP could learn from this example that current technology is far from reaching the constraints imposed by fitting in the circuitry between the pixels. Or they might choose to ignore inconvenient facts.

--
emil
--



http://theory.uchicago.edu/~ejm/pix/20d/

Barrie