P45

[Sandra]

Hi,
I'm planning to buy a MF camera. My first choice goes for a Contax 645 with a Phase One back. The P25 has great reviews. But I might be considering the P45.
I have 2 questions:
1- Are there any reviews online (i.e. internet)?   I've been searching for more than an hour and didn't find anything. Can somebody help?
2- A technical question: the P25 has 9micron pixels. the P45 has 6.8micron pixels!! Does this mean the P45 image quality is not as good as the P25'
Thanks!
S.

[kenstrain]

More pixels means smaller pixels or a bigger sensor.   If your lens can resolve sufficient to put detail finer than two pixel widths on the P25 - which seems quite likely for a good lens, at least over the cantral part of the image  - you gain resolution switching to the smaller pixels.  I imagine this is the main motivation for the P45 over the other.  

[IIRC there are either no or weak AA filters on both of these backs, so I said two pixels for the nyquist limit, as an initial guess - the limiting resolutions are roughly 55lp/mm and 73 lp/mm.  A very good lens should just be able to make a difference at f11 or f16.  I' expect you'd see a difference with a good 120mm f/4, for example.]

You are, perhaps, thinking that the noise is worse.  Well it could be, but
a) you may be able to drop one ISO step, this should restore the noise
b) the noise is finer in the image, which helps
c) you can use selective noise reduction to obtain the same noise where it matters and better resolution everywhere else.

But please wait for the reviews before believing any of this, as the financial element is considerable.

Ken

EDIT: disable emoticons if you type b)!

2- A technical question: the P25 has 9micron pixels. the P45 has 6.8micron pixels!! Does this mean the P45 image quality is not as good as the P25'???
Thanks!
S.
[a href=\"index.php?act=findpost&pid=53199\"][{POST_SNAPBACK}][/a]

[BJL]

2- A technical question: the P25 has 9micron pixels. the P45 has 6.8micron pixels!! Does this mean the P45 image quality is not as good as the P25'
[{POST_SNAPBACK}][/a]

No it does not mean that, because judging by pixel size alone ignores other differences. There have been a lot of other design changes from the older Kodak KAF-22000 sensor in the P25 to the new KAF-39000 in the P45. On paper at least, the reduction in pixel size has caused almost no reduction in dynamic range, which is the usual fear about smaller photo-sites. Also by the way the new smaller pixels generate 20% less electrons of read noise (16 vs 21), a fact often ignored in naive theoretical discussions of sensor noise levels.

I suggest you ignore any sweeping "bigger is better" generalizations and stay with your first idea of reading good reviews and comparisons.  And then seek a thorough hands-on try-out.

Or if you like reading technical spec's, visit
[a href=\"http://www.kodak.com/global/en/digital/ccd/products/fullframe/blueplusColorMain.jhtml?id=0.1.4.8&lc=en]http://www.kodak.com/global/en/digital/ccd...d=0.1.4.8&lc=en[/url]


P.S. If you care enough about noise at higher ISO speeds, consider another coming Phase One model, the P30 with "only" 31MP. It has a smaller, less expensive sensor (33x44mm instead of 37x49mm), the same 6.8 micron sized photo-sites, but it will be the first medium format sensor with micro-lenses, which about double its sensitivity over the P25 or P45. But it will be poorer for wider angle and tilt-shift work. At the Kodak site above, its sensor is the KAF-31600.

[BJL]

By the way, all these worries about the lower highlight headroom and dynamic range of smaller photo-sites will go out the window if new technologies like this work out:
http://www.dpreview.com/news/0512/05121201new_chips.asp

That project is at the University of Rochester, not coincidentally in Kodak's back-yard. It is another variant on repeatedly checking and resetting of photo-sites during an exposure so that wells never fill up (or run dry, since detection is by removing electrons from wells), with conversion to digital done on the fly during the exposure at each photo-site. For surveillance gear first, but maybe for normal cameras later.

[eleanorbrown]

Sandra i asked the same question to the Phase people about concerns regarding pixel size.  I have a Contax with a P25 back and am very happy with it.  It is resolving distant tree branches 1-2 pixels wide.  lenses are really first rate. do wait for the reviews and I would suggest testing one of the backs also.  In addition, check out the Phase One Forum on the P backs.  Eleanor

[BernardLanguillier]

I suggest you ignore any sweeping "bigger is better" generalizations and stay with your first idea of reading good reviews and comparisons. 
[a href=\"index.php?act=findpost&pid=53366\"][{POST_SNAPBACK}][/a]

Wondering if anyone is going to do an actual match up between the a 4*5 Imacon scan, Leaf 75, Phaseone P45, P30, P25, Mamiya ZD, 1ds2 and D2x and stitched images made from the 1ds2 and D2x.

I mean with pixel peeping detailed comparison of prints at A2 and A1 sizes on high DR landscape scenes with lots of small details both near and far (DoF check). Comparisons both in the center of the frame and in the corners (image uniformity check).

A simple review concluding that the P45 is great will have zero value as far as I am concerned, other than make us dream, which we can very well do on our own.

Pixel peeping is IMHO a must when dealing with equipment that is claiming to get these last 10% for you. A useful review would have to confirm us whether these 10% are actually gained or not.

Regards,
Bernard

[BJL]

... Pixel peeping is IMHO a must when dealing with equipment that is claiming to get these last 10% for you.
[a href=\"index.php?act=findpost&pid=53383\"][{POST_SNAPBACK}][/a]

Good ideas, but can we go for "print peeping" instead? Say 300ppi prints of crops from various parts of the images? The P45 gives about A2 or 18"x24" at 300ppi, or A3 at a bit over 400ppi.

P. S. My guess is that the winner would be well-done stitched images (for the situations where stitching is possible), by allowing one to use the higher absolute resolution of the best in smaller format lenses.

[BernardLanguillier]

I am 100% with you on this. I had prints in mind in the first place.

A2 is enough isn't it? No resampling done in PS, we just pick a 21 inch lenght (so as to be able to use an Epson 4000/4800 or HP 130) and compute the DPI needed without doing any up-sampling in PS. Then we let IP (for instance) do the job of printing the image the best it can.

The best would be to request the brands to send a representative able to use each piece of equipment at its max, gather everybody in one/two locale(s) and start shooting 4 or 5 scenes that everyone agrees on. For instance:

- one fully static scene in morning/evening light with a near/far composition featuring lost of small details both in the foregounrd and in the distance -> check DoF, overall sharpness, colors, WB and shadow noise,
- one scene with moving water (river or waves) -> this will put pressure on stitchers,
- one back lit scene including the sun in the image, branches,... -> check DR, highlight rendition and CA,
- one panorama shot,
- one macro shot.

I agree with you on stitching coming out as the sharpness winner, although you lose the DoF advantage of the smaller format all things being equal.

But there is more than sharpness. We should also take into account:

- The time needed to come up with the print,
- The DR achieved (overlaying being a valid technique),
- ...

For the test to be perfect, we would also need all these cameras to be mounted on the same tripod (say a Gitzo 1227/1257) with the same ball head (say a BH-55)... which makes things a bit comples. It should be possible to come close enough though.

Wondering if Michael would be interested in conducting this. If he is not, it could be fun to try to do it in Japan. I wonder if the back resellers would be interested in cooperating with such a test...

Regards,
Bernard