Megapixel war just beginning with DSLRs

Pentax K20D, 1.5 crop and 15 megapixels, now Sony?? I've got to
admit, a 25 megapixel 1.5 crop camera would be hilarious.

http://masterchong.com/v2/sony-alpha...dy-photos.html

On Jan 28, 7:40 pm, RichA wrote:
Pentax K20D, 1.5 crop and 15 megapixels, now Sony?? I've got to
admit, a 25 megapixel 1.5 crop camera would be hilarious.

http://masterchong.com/v2/sony-alpha...-camera-body-p...

I'd just like to add to this that there is (obviously) a pixel size at
which the gains of resolution are outweighed by the losses in DR and
sensitivity. So, what specifically can we say (given the performance
we've seen with the latest DSLRs) is the smallest practical pixel size
for a DSLR?

How many people on this newsgroup wrote that APS-c sized sensors could never
achieve the noise levels seen on the new Sony wunderkind in the D300 and new
Sony dSLR?
In truth most users of consumer grade dSLRs do not need more than 6mps.

RichA wrote:
On Jan 28, 7:40 pm, RichA wrote:
Pentax K20D, 1.5 crop and 15 megapixels, now Sony?? I've got to
admit, a 25 megapixel 1.5 crop camera would be hilarious.

http://masterchong.com/v2/sony-alpha...-camera-body-p...

I'd just like to add to this that there is (obviously) a pixel size at
which the gains of resolution are outweighed by the losses in DR and
sensitivity. So, what specifically can we say (given the performance
we've seen with the latest DSLRs) is the smallest practical pixel size
for a DSLR?

That only matters when legacy lenses are used. Lenses of higher
precision can be made, and if sensors with smaller pixels come into
use, better lenses will be made to exploit them.

--
Chris Malcolm DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]

RichA wrote:

I'd just like to add to this that there is (obviously) a pixel size at
which the gains of resolution are outweighed by the losses in DR and
sensitivity. So, what specifically can we say (given the performance
we've seen with the latest DSLRs) is the smallest practical pixel size
for a DSLR?

12M on an APS-size sensor (Nikon D2x) is quite demanding of lenses. It
really shows when a lens is less than perfect, and benefits from using
the best ones you can use. Smaller pixels will make this problem more
significant -- but the problem can be solved by making better lenses.

Another issue is that pixels that small make focus errors, camera shake,
etc., that much more pronounced. This is worked around by better technique.

The tricky one is that, at this size, you have a practical minimum
aperture of f/11 before you start losing resolution to diffraction.
With smaller pixels this will get worse, and I wouldn't want to use
an APS-size camera that can't go below f/8. You can't really solve
this one with better lenses.

--
Jeremy Nixon | address in header is valid
(formerly )
http://www.flickr.com/photos/100mph/

Jeremy Nixon wrote:
RichA wrote:

So, what specifically can we say ...is the smallest practical pixel size
for a DSLR?

The tricky one is that, at this size, you have a practical minimum
aperture of f/11 before you start losing resolution to diffraction.
With smaller pixels this will get worse, and I wouldn't want to use
an APS-size camera that can't go below f/8. You can't really solve
this one with better lenses.

Interesting point. f/11 or 16 is where a 12MP D300 or D2x becomes
diffraction limited so for 15MP maybe f/16 isn't worthwhile (for that
type of tripod landscape/studio work anyways).

On Jan 28, 10:32 pm, Paul Furman wrote:
Jeremy Nixon wrote:
RichA wrote:

So, what specifically can we say ...is the smallest practical pixel size
for a DSLR?

The tricky one is that, at this size, you have a practical minimum
aperture of f/11 before you start losing resolution to diffraction.
With smaller pixels this will get worse, and I wouldn't want to use
an APS-size camera that can't go below f/8. You can't really solve
this one with better lenses.

Interesting point. f/11 or 16 is where a 12MP D300 or D2x becomes
diffraction limited so for 15MP maybe f/16 isn't worthwhile (for that
type of tripod landscape/studio work anyways).

Like Olympus, diffraction kicks in at around f8.

RichA wrote:
Paul Furman wrote:
Jeremy Nixon wrote:
RichA wrote:
So, what specifically can we say ...is the smallest practical pixel size
for a DSLR?
The tricky one is that, at this size, you have a practical minimum
aperture of f/11 before you start losing resolution to diffraction.
With smaller pixels this will get worse, and I wouldn't want to use
an APS-size camera that can't go below f/8. You can't really solve
this one with better lenses.
Interesting point. f/11 or 16 is where a 12MP D300 or D2x becomes
diffraction limited so for 15MP maybe f/16 isn't worthwhile (for that
type of tripod landscape/studio work anyways).

Like Olympus, diffraction kicks in at around f8.

I guess that's what large format is for... like a studio tabletop shot,
you can ultimately stop down more if the subject holds still.

I wonder what the '35mm equivalent f/stop' of a stitched pano would be?

Paul Furman wrote:
Jeremy Nixon wrote:
RichA wrote:

So, what specifically can we say ...is the smallest practical pixel size
for a DSLR?

The tricky one is that, at this size, you have a practical minimum
aperture of f/11 before you start losing resolution to diffraction.
With smaller pixels this will get worse, and I wouldn't want to use
an APS-size camera that can't go below f/8. You can't really solve
this one with better lenses.

Interesting point. f/11 or 16 is where a 12MP D300 or D2x becomes
diffraction limited so for 15MP maybe f/16 isn't worthwhile (for that
type of tripod landscape/studio work anyways).

Isn't worthwhile if using the full resolution. But if you have 12MP
you might well be content with say a close-up product photograph of
3MP resolution, for which you could either use in-camera or editor
resizing. In that case would you gain anything, such as increased DoF,
by being able to stop down further than the 12MP diffraction limit?

--
Chris Malcolm DoD #205
IPAB, Informatics, JCMB, King's Buildings, Edinburgh, EH9 3JZ, UK
[http://www.dai.ed.ac.uk/homes/cam/]

On Jan 29, 8:44 am, Paul Furman wrote:


I wonder what the '35mm equivalent f/stop' of a stitched pano would be?

The same: stitching amounts to using a larger sensor with the same-
sized pixels, if you think about it. The focal length and aperture
diameter also don't change.