Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: Can M8ers learn from PhaseOne?  (Read 2736 times)
scott kirkpatrick
Newbie
*
Offline Offline

Posts: 22


« on: November 30, 2006, 05:21:43 PM »
ReplyReply

Michael, when you were working with the extreme wide angle cameras that employ a PhaseOne digital back (P25 and P39, I believe), you described calibrating the "lens cast" so that Capture One software could remove the cyan fringe that resulted at the edges of the frame.   This sure does sound like exactly the problem that was such a surprise to Leica when they underfiltered IR at the sensor.

Am I correct that the big digital backs have no IR filter?  The Kodak spec sheet for the 39000 chip makes no mention of it.

How well does "lens cast" removal work?  Was there any residual under- or over-"cast" when focussing very near or at very wide apertures?  Is there any user control over this or is it completely automatic once set up?

Is there a piece of technology in this that Leica should be picking up?

scott
Logged
michael
Administrator
Sr. Member
*****
Offline Offline

Posts: 4782



« Reply #1 on: November 30, 2006, 05:57:21 PM »
ReplyReply

These are different problems with different causes that require different solutions.

Michael
Logged
BernardLanguillier
Sr. Member
****
Online Online

Posts: 7752



WWW
« Reply #2 on: November 30, 2006, 09:17:17 PM »
ReplyReply

Quote
These are different problems with different causes that require different solutions.

Michael
[a href=\"index.php?act=findpost&pid=87950\"][{POST_SNAPBACK}][/a]

And even if it were the same issue, I don't think that the type of photography typically done with M8s would be compatible with having to shoot a calibration image on a white disc between every shots.

Regards,
Bernard
« Last Edit: November 30, 2006, 09:19:04 PM by BernardLanguillier » Logged

A few images online here!
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7247


WWW
« Reply #3 on: December 01, 2006, 12:19:20 AM »
ReplyReply

Hi,

My guess is that PhaseOne has an IR-filter. I think that "green cast" arises when light passes trough the IR-filter at oblique angles. I guess that this was the major reason for Leica to omit the IR-filter.

Leica has two problems in this area:

1) Short distance between exit pupil and sensor on wide angles, because there is no need for mirror.

2) Leica lenses are intended to be used at large apertures, which I think aggrevates the problem with oblique rays.

When IR-filters are added in front of the lens there will be a "green cast" on the edges. The camera has some firmware change to handle that. That is the probably the reason that the "fix" with external IR-filter needs coded lenses. (This speculation is coming from Sean Reid Reviews.)

My guess is that Leica will redesign the M8 with a sensor with IR-filter, and I guess that it will still be 0.5 mm thick.

Best regards

Erik


Quote
Michael, when you were working with the extreme wide angle cameras that employ a PhaseOne digital back (P25 and P39, I believe), you described calibrating the "lens cast" so that Capture One software could remove the cyan fringe that resulted at the edges of the frame.   This sure does sound like exactly the problem that was such a surprise to Leica when they underfiltered IR at the sensor.

Am I correct that the big digital backs have no IR filter?  The Kodak spec sheet for the 39000 chip makes no mention of it.

How well does "lens cast" removal work?  Was there any residual under- or over-"cast" when focussing very near or at very wide apertures?  Is there any user control over this or is it completely automatic once set up?

Is there a piece of technology in this that Leica should be picking up?

scott
[a href=\"index.php?act=findpost&pid=87946\"][{POST_SNAPBACK}][/a]
Logged

scott kirkpatrick
Newbie
*
Offline Offline

Posts: 22


« Reply #4 on: December 01, 2006, 06:54:11 PM »
ReplyReply

Quote
Hi,

My guess is that PhaseOne has an IR-filter. I think that "green cast" arises when light passes trough the IR-filter at oblique angles. I guess that this was the major reason for Leica to omit the IR-filter.

Leica has two problems in this area:

1) Short distance between exit pupil and sensor on wide angles, because there is no need for mirror.

2) Leica lenses are intended to be used at large apertures, which I think aggrevates the problem with oblique rays.

When IR-filters are added in front of the lens there will be a "green cast" on the edges. The camera has some firmware change to handle that. That is the probably the reason that the "fix" with external IR-filter needs coded lenses. (This speculation is coming from Sean Reid Reviews.)

My guess is that Leica will redesign the M8 with a sensor with IR-filter, and I guess that it will still be 0.5 mm thick.

Best regards

Erik
[a href=\"index.php?act=findpost&pid=87999\"][{POST_SNAPBACK}][/a]

I did some reading of PhaseOne's product literature -- there is an IR filter supplied by Phase One in both the P45 and P25.   The Kodak chips that they use don't describe the filter in their specs so it is not bonded to the chip but must be attached in a frame by Phase One during the camera manufacture.  It's not clear whether they chose an IR-cut filter or a weaker version, but the IR-cut class of filters is much more strongly angle dependent.  The CaptureOne literature on "lens cast" talks about both green casts on one side of the image and magenta casts -- I have no idea how the magenta part might happen.  

Analyzing Sean's experiments, I find that there is little aperture sensitivity to the amount of green cast at the edges, and this is apparently predicted to be insensitive in more elaborate calculations.  Large aperture increases overall vignetting, but not the color effects.


Michael, in dismissing my question so completely in his response above, may have thought I was asking about IR-induced false colors, but the problems of wide angle lenses close to the imager occur in MF cameras as well.  Phase One seems to know some things about the problem and Capture One about the solution.

scott
« Last Edit: December 01, 2006, 07:01:48 PM by scott kirkpatrick » Logged
Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad