Re: sony alpha a100 for a beginner? What lens?

David Kilpatrick wrote:
J. Clarke wrote:


You've made a lot of assertions about noise reduction. Do you have
the images or measurements to back them up?


I have some images which have been widely enough seen. I don't make
measurements, but Anders Uschold's measurements are good enough and I
find that nearly every observation I can make from examining actual
files ends up supported by Uschold's published results.

See:

http://www.pbase.com/davidkilpatrick/digital_photo_samples&page=2

This includes some comparisons of Dynax 7D, Sony A100, Nikon D80 and
Canon 400D. I have subsequently tested the K10D but not repeated the
exercise. My Sony A100 pages on pbase also include many different
noise, dynamic range and related comparisons:

http://www.pbase.com/davidkilpatrick/sony_alpha_100

Many of these are images which were described in dpreview threads and
don't have captions, sorry; also there's almost 500MB of images on the
pages now and I must weed them soon.

To make a convincing case you really need to organize those and describe
your methodology.

I have had enough discussions with engineers responsible for image
processing systems (face to face) and enough long discussions via
various groups to know that I'm not talking through my hat. Others
here who know more about the inside workings of these systems will
confirm that noise reduction and sharpening processes which are not
connected with the controllable NR or sharpness settings of user menu
screens take place in all digital cameras.

So what specific "noise reduction and sharpening processes" are
different in the Sony?

I've been an engineer long enough to have a very healthy respect for the
ability of a non-engineer to listen to a bunch of engineers talking for
a week and then regurgitate what he learned and get it all wrong.

This is one reason why there is so much variation in performance
between cameras using the same basic sensor - Sony A100, Pentax K10D,
Nikon D80 (the D200 is not identical, just nearly so).

David

--
--
--John
to email, dial "usenet" and validate
(was jclarke at eye bee em dot net)


.

Relevant Pages

  • Re: sony alpha a100 for a beginner? What lens?
    ... I have some images which have been widely enough seen. ... I don't make measurements, but Anders Uschold's measurements are good enough and I find that nearly every observation I can make from examining actual files ends up supported by Uschold's published results. ... I have had enough discussions with engineers responsible for image processing systems and enough long discussions via various groups to know that I'm not talking through my hat. ... Others here who know more about the inside workings of these systems will confirm that noise reduction and sharpening processes which are not connected with the controllable NR or sharpness settings of user menu screens take place in all digital cameras. ...
    (rec.photo.digital.slr-systems)
  • Re: NI Vision Assistant
    ... NI Vision Assistant 7.1, and I'm collecting an avi file from the camera ... average 90 images that I upload into the browser. ... I perform three different measurements on each image. ...
    (comp.lang.labview)
  • Re: Nikon D80/200 - Canon 30d
    ... unequivocally that CMOS sensors generate *more* noise than CCDs, ... physics and statistics --- you don't even try to understand ... that they are not measurements, but claim that nobody does have ... Basically, they take two images shot one right after the other, then ...
    (rec.photo.digital.slr-systems)
  • Re: precession of mercury
    ... telescope was not designed to take images. ... The 120,000 stars in the ... photometric B-V and astrometric measurements for over 1 million ...
    (sci.physics.relativity)
  • Re: 20 Years of JPEG Celebrated With Software Launch
    ... but I and other people in application are looking *real* images. ... everyone in the community knows that PSNR isn't a visual metric. ... As said, and as already posted, it's a bit too early to post results, but on the measurements we have at the current time, things work very much alike like PSNR, interestingly. ... If you have proposals for objective, full reference visual metrics, please let me know. ...
    (comp.compression)