High ISO on M9 better than m8.2?

I originally shot the sample shots using uncompressed DNG. I just compared the Histogram of the original sample image using Lightroom 3.3, and it is fully populated. It covers more of a spread than many of the "Live Images". Has a good tail to it, in the areas of black in the hair and pupil of the eyes.
 
M9, Elmar 134/4, iso 2500

full frame
U25074I1300841656.SEQ.0.jpg



100% crop, minimal noise reduciton in lr3 (colour 25, luminance <10)
U25074I1300841647.SEQ.0.jpg


Mike
 
That´s a very nice example!

Definitely not like the great m8. (noise helps quite a bit on B/W conversions)
 
The " grain" works very well here; it could be reduced even more with ACR 6.0 (CS5) but there is no need. It will print beautifully :)
 
Have a look at DXOMARK. The M9 is better but not much better. note that the pixel pitch is essentially the same.
I wouldn't use ISO 80 as its a pull-ISO (not a the base sensor ISO, so it requires extra processing which degrades the picture).
 
My issue with M8 high iso (alas I have no M9, nor will I in the foreseeable future) is banding in the shadows, much more than noise, and I see what looks like banding or something similar in the dark areas above the woman's head in Brian's 2500 shot.
 
JSU said:
Brian--
Which version of PS are you using, I currently have CS3 and I'm unfamiliar with the dust filter? It isn't listed under filters, that I know of, is it elsewhere or only available in newer versions?
Click to expand...

I am using Photoshop 7.0, circa 2002. I can export 16-bit/color PSD's to it from Lightroom and it works fine.

I'm surprised a Dust and Scratch filter is not in the newer versions.

Also- I am surprised by how "Flimsy" Lightroom 3.3 is with respect to crashing on start-up. It created a corrupt "Catalog" file, and crashed without recovery options. De-install and re-install did fix it. Finally, changing the Directory name of the path it was looking in allowed recovery. Apparently it "Hid" a corrupt catalog file somewhere in the directory. Once the directory was inaccessible, gave me the option to "create new catalog". Let's just say I am underwhelmed by Adobe Products of the last few years, after using them since 1994.
 
under LR3 high iso raw shots of the M9 are usable

M9, iso 2500
not a dslr result, but not so bad...

25001c.jpg
 


35/1.2 Nokton, wide-open, 1/6s, hand-held. ISO 2500 and -1 EV compensation. Straight JPEG conversion in LR 3.3.

At the Planetarium in Columbus, GA.
 
Found this ISO 2500 shot done with the M8.

Canon 50/1.4, wide-open, 1/125th sec ond.

picture.php


No issues with banding, color is decent, noise is under control. I have a late-run M8.
 
I have an R-D1 and consider getting an M9 later on, and I worry about banding as well, mostly on scenes with bright lights on the background, leading to banding on the shadows. So when I pull up the backlight tab I get a purple stripe. I'm not sure if it's something that is intrinsic to digital sensor or if it can be fixed.
 
ISO 2500, 35/1.2 Nokton wide-open, 1/6s, hand-held.



Raw mode, Auto white-balance, no corrections required for LR3.3 to convert to JPEG. About as straight from the camera as you can get.

I used the ISO 2500 setting quite a bit in two "no Flash Photography allowed" museums last week. I was happy- far better than my experience with Kodacolor 800.
 
My M8 was very similar in noise level to my Canon 20D. I find my M9 is similar to my 5D MK-I. So, improved but still a generation behind Canon (and so I'm told by my Nikon-shooting pals, 2 generations behind Nikon :D). But although that's speaking of noise levels, my personal feeling is that noise quality is more pleasing in the M9 and M8 than my prior-gen Canons, particularly in b&w. As with film, the predominance of my photos are shot either in color in daylight (hence at ISO 320 max), or in low-light in b&w. So I'm at peace with the noise performance of the M9, and even think that if I had a D700 I might end up having to grain-effect in PP for those moody b&w shots I enjoy.
 
Problem is, everybody is comparing the incomparable when quoting Canon/Nikon. With those brands noisereduction is standard in- camera, with the Leicas it is DIY in the computer. And we are comparing CMos with on - sensor reduction with CCD which does not do that.
As for M8/M9, the M9 has double the area, double the pixels and half the magnification, thus half the noise.
 
I wish Kodak would publish the long sheet for the KAF-18500. The sensor element size is the same. the Signal to Noise ratio for a CCD would normally be higher for a larger CCD type sensor as the electron charge must shift through more elements to make it off the chip. I suspect that Kodak substantially improved the system noise by a factor of 4 on the KAF-18500 as compared with the KAF-10500. I am not using any noise reduction on the shots done in the Planetarium- and needless to say, it was very dark. The projections are not very bright, hence the slow shutter speed with an F1.2 lens and ISO 2500.
 
You must log in or register to reply here.
Back
Top Bottom