Any Hope for Canon?

[BJL]

I find it quite bizarre that people are seriously speculating that Nikon, #2 in DSLR sales and profits with daylight in third place, is threatened with extinction by the second tier SLR systems of  Konica-Minolta/Sony/Zeiss, Samsung/Pentax/Schneider or even Olympus/Panasonic/Leica.

the success of an SLR system depends greatly on factors like its lens selection, camera components like auto-focus and auto-exposure, accessory systems like flashes and so on. In these areas, both Canon and Nikon are solidly clear of the competition, notwithstanding even the the smaller SLR makers' new partnerships with old-guard German lens makers Schneider, Leica and Zeiss.


P. S. Nikon collaborates with Nikon on sensor designs for Nikon SLR's, not to mention providing Sony with the steppers that it needs to make those sensors. (Would Sony try to abandon Nikon and go to the other major stepper maker, Canon!?) To some extent, Nikon might be indirectly reselling some of its sensor technology for use in Pentax and Sony SLRs.

This is probably related to the exclusive period that Nikon seems to get on each new sensor model. It is for now not clear if the Sony A100 uses exactly the same sensor as the D200 or a lower cost version, stripped on the multi-channel read-out needed to reach 5fps, and perhaps lacking the ISO 100 and 125 capability (the A100 does 3fps and has minimum ISO speed of 160).

And as a variant on pom's ideas, Fuji is one of various alternative sensor making partners for Nikon; Fuji is already generating sales for Nikon lenses, accessories and the numerous Nikon components used in Fuji DSLR's.

[John Sheehy]

I hope the successor to the 5D will not only have a greater pixel count but a real ISO 3200 setting with less noise than ISO 1600 (with same exposure).
[a href=\"index.php?act=findpost&pid=67496\"][{POST_SNAPBACK}][/a]

It does seem like a real 3200 could be better than under-exposing at ISO 1600, based on the trends going from 100 to 1600, although I don't have the data for the 5D.  With the 20D, the standard deviation of blackframe noise (the noise that determines dynamic range) goes 2.1 -> 2.2 -> 2.5 -> 3.2 -> 4.7 -> 9.4 from ISO 100 to 3200, and if these are normalized for ISO 100, they go 2.1 -> 1.1 -> 0.63 -> 0.4 ->  0.29 at 1600, and then the same at 3200, since it is only an arithmetic gain.  Based on the trend for the analog gains, one would expect something like 0.22 or so for 3200 with the same technology, which is almost a half stop better than 0.29.

[Ray]

I do not understand what anyone stands to gain from this. As I understand it, you are comparing images taken the same exposure level (same f-stop and shutter speed) but with different ISO speed settings, and thus with different amounts of exposure compensation in post-processing. Specifically, ISO 100 with about four stops of underexposure and thus massive "brightening" needed in PP compared to ISO 1600 setting with exposure giving roughly the desired output levels with little or no PP adjustment needed.

What is the point? Surely in general one would use the latter, higher ISO speed option, and never have reason to about whether or not it is slightly better than the never used option of "ISO 100 and +4".
[a href=\"index.php?act=findpost&pid=67551\"][{POST_SNAPBACK}][/a]

BJL,
The comparison using 'same exposures' is to demonstrate graphically and clearly that there is a real and unequivocal advantage to using a high ISO (with Canon cameras) that is not just a software fudge. This has not always been the case. The following image taken with my D60 at ISO 400 demonstrates this point very well.

[attachment=667:attachment]

It was necessary to use ISO 400 for this shot in order to get the White Herons sharp (1/500th at f5.6). The exposure was pretty close to a full ETTR, requiring a minus 0.5EC adjustment in ACR. But have a look at the shadows.

[attachment=668:attachment]

There's room for improvement here and the improvement has been delivered by Canon, in spades.

Image Stabilisation and high image quality with a small amount of light are high priorities for me when choosing a camera, and Canon has it. I think it's fantastic I can use my 5D, handheld, for street photography at night without a flash.

[John Sheehy]

What is the point? Surely in general one would use the latter, higher ISO speed option, and never have reason to about whether or not it is slightly better than the never used option of "ISO 100 and +4".
[a href=\"index.php?act=findpost&pid=67551\"][{POST_SNAPBACK}][/a]

It is good to know what the compromises of under-exposure are with your equipment.  For example, if you know that your camera under-exposed 2 stops at ISO 100 isn't any worse than ISO 400 (not under-exposed), why even use the 400 if there are highlights like speculars or light-sources in the frame?  You might as well capture them correctly if there is no loss to the shadows.  If, however, you know that your camera is cleaner as measured in absolute exposure terms at higher ISOs, you might decide to forego the specular highlights and optimize for the shadows with the higher ISO.