CMOS vs CCD Meanderings

One is a landscape camera and the other is a people camera. M series cameras have always been people cameras...
Have they? I would say that if they can be classified as limited as that (which I doubt very much) that it is more of a reportage and travel camera.
 
Here is a comparison of photos and the ensuing discussion on L Camera forum that some may find interesting. http://www.l-camera-forum.com/leica-forum/leica-m9-forum/275508-new-leica-m-vs-m9-daylight.html

I downloaded both RAW files and did a basic test:

*The M9 shot is overexposed, and the EXIF data for M9 lacks f stop number while for M10 its f8

M10 EXIF:

5oxi53.jpg


M9 EXIF:

x6g4cz.jpg




I used RPP to convert the DNG files with L curve and and everything else on default. This is how files look after conversion in RPP:

M10:
n3m6gk.jpg


M9:
2vmubef.jpg
 
Have they? I would say that if they can be classified as limited as that (which I doubt very much) that it is more of a reportage and travel camera.

A people camera will always be great for landscape as well, a landscape camera not so much - in digital photography and also with color film.
 
I have to say that from what I am seeing in these pictures, the highlight rendition in the M10 is enough to make me want it. This aspect of digital has always bothered me. I believe the color can be adjusted to render proper skin tones if people are seeing problems in this area. If Leica perceives the same thing. I always preferred Kodachrome but that is not going to get me anywhere.
 
I have to say that from what I am seeing in these pictures, the highlight rendition in the M10 is enough to make me want it.

*The M9 shot is overexposed and there is no number for what f stop was used in the M9 shot.

Even with overexposure, the M9 has more gradation in the highlights and a more neutral color rendition.
 
The histogram for the RAW files:

M10:

adkw14.jpg


M9:

vq5rx3.jpg


The histogram from M10 RAW file looks a little too nice and processed for a RAW file. The M9 histogram is clearly overexposed, at least by 1.5 to 2 stops.
 
-The histogram shots are from the DNG RAW files in faststone image viewer before conversion in RPP, Photoshop cannot open RAW files... you need acr.

-The M10 shot is properly exposed and processed... the clouds look sharpened.

-The M9 shot is overexposed and the f stop number omitted from the EXIF data.

-RPP knows very well how to deal with M9 RAW files, its an old camera and M9 users love RPP. RPP should have a problem with M10, but it does not seem to be the case here.


From a suspicious perspective, something stinks in these two shots but even despite all the "suspiciousness", the M9 still looks 'better' imo.


That comparison tells us a few things.

1. The native white point of the sensors is slightly different, and RPP64 doesn't have a clue how to deal with that.

2. The M10 has substantially better dynamic range, just as DxO tests had already demonstrated.

3. Using auto-contrast to compare files from two sensors is not a very good idea.

4. That Photoshop histogram tool was written when 8-bit files predominated (256 gray values). It's not telling you much that is useful.

5. A same-ISO comparison is somewhat unfair to the M9. A fairer test would be to operate both sensors at base ISO.
 
-The histogram shots are from the DNG RAW files in faststone image viewer before conversion in RPP, Photoshop cannot open RAW files... you need acr.

-The M10 shot is properly exposed and processed... the clouds look sharpened.

-The M9 shot is overexposed and the f stop number omitted from the EXIF data.

-RPP knows very well how to deal with M9 RAW files, its an old camera and M9 users love RPP. RPP should have a problem with M10, but it does not seem to be the case here.


From a suspicious perspective, something stinks in these two shots but even despite all the "suspiciousness", the M9 still looks 'better' imo.
-It is incorrect to capitalize RAW. It is not an acronym.
-There is not much interest in including f-stops in digital M files as they are guesstimates only.
 
Back
Top Bottom