Again! Another camera goes soft with a pixel jump!!
RichA
Olympus E-410/510 to E-420/520 (no pixel jump but more NR) images went
soft.
Now, the newest Canon, same thing. It's obvious something, high
powered in-camera NR, whatever, is softening up the images.
Get lost
Oops! Sorry. Some other posters on dpreview have now pointed out
that it appears the samples provided by the new Canon have...focus
problems! Big ones!! What a surprise! So we can't really judge the
camera until they fix that issue. Come back to it in about 8 months
then. :)
Bruce
>Pentax K10D to K20D, 10 to 15 megapixel jump, images got soft.
>Olympus E-410/510 to E-420/520 (no pixel jump but more NR) images went
>soft.
>Now, the newest Canon, same thing. It's obvious something, high
>powered in-camera NR, whatever, is softening up the images.
Also true of the 1Ds Mk III and especially the 5D Mk II.
The level of detail obtainable from the 5D Mk II is way below what you
would expect of a 21.8 MP DSLR, and not so much better than the output
from the original 5D.
Martin Brown
> RichA <rande...@gmail.com> wrote:
>
>> Pentax K10D to K20D, 10 to 15 megapixel jump, images got soft.
>> Olympus E-410/510 to E-420/520 (no pixel jump but more NR) images went
>> soft.
>> Now, the newest Canon, same thing. It's obvious something, high
>> powered in-camera NR, whatever, is softening up the images.
>
>
> Also true of the 1Ds Mk III and especially the 5D Mk II.
It is more likely that the sensor is outpacing the capabilities of the
glass and end user combined. Are the images still soft on a tripod and
taken in good light at the optimum f ratio for diffraction limited imaging?
Sharpness sells so a lot of the lower end cameras by default do some
fairly brutal in camera sharpening to add extra punch. The high end
cameras tend to record more accurately what the lens actually saw.
>
> The level of detail obtainable from the 5D Mk II is way below what you
> would expect of a 21.8 MP DSLR, and not so much better than the output
> from the original 5D.
The search of ever larger numbers of pixels is running out of steam. If
the pixels are too small you get poor signal to noise (not that it will
stop the marketeers fooling Joe Public with their uber-megapixel race).
Regards,
Martin Brown
Bruce
>Bruce wrote:
>> RichA <rande...@gmail.com> wrote:
>>
>>> Pentax K10D to K20D, 10 to 15 megapixel jump, images got soft.
>>> Olympus E-410/510 to E-420/520 (no pixel jump but more NR) images went
>>> soft.
>>> Now, the newest Canon, same thing. It's obvious something, high
>>> powered in-camera NR, whatever, is softening up the images.
>>
>>
>> Also true of the 1Ds Mk III and especially the 5D Mk II.
>
>It is more likely that the sensor is outpacing the capabilities of the
>glass and end user combined.
Nonsense.
>Are the images still soft on a tripod and
>taken in good light at the optimum f ratio for diffraction limited imaging?
Absolutely.
Ray Fischer
>Martin Brown <|||newspam|||@nezumi.demon.co.uk> wrote:
>>Bruce wrote:
>>> RichA <rande...@gmail.com> wrote:
>>>> Pentax K10D to K20D, 10 to 15 megapixel jump, images got soft.
>>>> Olympus E-410/510 to E-420/520 (no pixel jump but more NR) images went
>>>> soft.
>>>> Now, the newest Canon, same thing. It's obvious something, high
>>>> powered in-camera NR, whatever, is softening up the images.
>>>
>>> Also true of the 1Ds Mk III and especially the 5D Mk II.
>>
>>It is more likely that the sensor is outpacing the capabilities of the
>>glass and end user combined.
>
>Nonsense.
What an impressively well-reasoned rebuttal.
>>Are the images still soft on a tripod and
>>taken in good light at the optimum f ratio for diffraction limited imaging?
>
>Absolutely.
Zoom lenses don't count, and especially not the cheap ones.
--
Ray Fischer
rfis...@sonic.net
Don Stauffer
Indeed, that is what I was going to comment on. The higher the
resolution of a new film or a new camera, the more focus and other
imaging faults show up. And, of course, with digital, we can always blow
up the images so simply with fat pixels in image editing programs, so we
can really see any imperfections.
Now, that to me is not a reason to AVOID high-res cameras, it is just
tht the higher the resolution you use, the more craftsmanship it takes
to make use of that higher res :-)
This is assuming, of course, that the rest of the system can make use of
the higher pixel count- i.e, it really is a higher res system.
David J. Littleboy
"Don Stauffer" <stau...@usfamily.net> wrote:
>
> Now, that to me is not a reason to AVOID high-res cameras, it is just tht
> the higher the resolution you use, the more craftsmanship it takes to make
> use of that higher res :-)
Exactly! Faster shutter speed (if hand held), f stop selected from a
narrower range, sturdier tripod. High res is a pain, and you don't get the
high res unless you embrace that pain.
--
David J. Littleboy
Tokyo, Japan
Charles
"Martin Brown" <|||newspam|||@nezumi.demon.co.uk> wrote in message
news:dVwyl.102692$Rg3....@newsfe17.iad...
> It is more likely that the sensor is outpacing the capabilities of the
> glass and end user combined. Are the images still soft on a tripod and
> taken in good light at the optimum f ratio for diffraction limited
> imaging?
The tried and true weak link maxim. The emphasis on sensor resolution was
once important, but is now mostly a tiresome marketing ploy.
> Sharpness sells so a lot of the lower end cameras by default do some
> fairly brutal in camera sharpening to add extra punch. The high end
> cameras tend to record more accurately what the lens actually saw.
Wow, how true this is with beginners who see sharpness above all factors but
seem to miss (ignore?) the ugly artifacts.
> The search of ever larger numbers of pixels is running out of steam. If
> the pixels are too small you get poor signal to noise (not that it will
> stop the marketeers fooling Joe Public with their uber-megapixel race).
Noise really shows up with the crammed sensors. The fixes fuzz the picture
and the EFFECTIVE resolution is less than what can be achieved with a
reasonable design.
Alas, the horsepower race and the watts per channel race and the digital
cameral MP race will never completely die because consumers are too often
keyed in one or two issues.
Get lost
> >>Bruce wrote:
> >>> RichA <rander3...@gmail.com> wrote:
> >>>> Pentax K10D to K20D, 10 to 15 megapixel jump, images got soft.
> >>>> Olympus E-410/510 to E-420/520 (no pixel jump but more NR) images went
> >>>> soft.
> >>>> Now, the newest Canon, same thing. It's obvious something, high
> >>>> powered in-camera NR, whatever, is softening up the images.
>
> >>> Also true of the 1Ds Mk III and especially the 5D Mk II.
>
> >>It is more likely that the sensor is outpacing the capabilities of the
> >>glass and end user combined.
>
> >Nonsense.
>
> What an impressively well-reasoned rebuttal.
>
> >>Are the images still soft on a tripod and
> >>taken in good light at the optimum f ratio for diffraction limited imaging?
>
> >Absolutely.
>
> Zoom lenses don't count, and especially not the cheap ones.
Some modern zooms are equal or better than any old prime designs. Old
crap from the 1980s don't count.
Wolfgang Weisselberg
> Bruce wrote:
>> The level of detail obtainable from the 5D Mk II is way below what you
>> would expect of a 21.8 MP DSLR, and not so much better than the output
>> from the original 5D.
> The search of ever larger numbers of pixels is running out of steam. If
> the pixels are too small you get poor signal to noise (not that it will
> stop the marketeers fooling Joe Public with their uber-megapixel race).
The 20D (8Mpix, crop) and the 5D II share the same pixel density.
What that means re "the 5D II has not enough detail" or "too
small pixels" is an easy exercise for the intelligent reader.
-Wolfgang