The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

Digital Sensor "Centerfold" correction

GrahamWelland

Subscriber & Workshop Member
It's bland white sky season again here and so a lot of my images are basically high-key in nature, at least as far as the sky is concerned. As a result I'm shooting and processing a lot of images to B&W, using C1 Pro 7 as the raw converter and Photoshop/Silver Efex Pro 2 for the B&W treatment. This has the capacity to stretch delicate tonality to a high degree and as a side effect it also really emphasizes any subtle sensor mosaic centerfold differences. (Thank goodness for PS's content aware fill / correction!!) It's always been there in colour but very very rarely visible unless a file was pushed & pulled.

I was wondering if other folks had noticed their backs becoming more susceptible to centerfold calibration over time? My IQ160 is now some 15 months old and I'm beginning to notice the calibration differences more, exacerbated no doubt by me doing a lot more B&W conversions these days.

Have you had to get your back recalibrated? Any feedback on turnaround times if so?

I've submitted a case with Phase One and contacted my dealer already although it's holiday season so I don't really expect to hear much until next week. However, I was curious to know what experiences others might have had with this.
 

Pemihan

Well-known member
I only notice it with my Cambo, with the 645 AFDIII there seems to be nothing.
I think it has something to do with the angel of light from the large format lenses.

LCC seems to take care of the issue for me...

Peter
 

GrahamWelland

Subscriber & Workshop Member
Doug - Sure - will do.

This particular case was with a Alpa/Schneider 150 XL Digitar on my Alpa STC. 8-10mm rise. I'll send you the raw and example B&W rendered image today when I'm home. The images were processed with ICC as I was actually putting together a pano. Normally I seldom bother with ICCs with the 150 XL.

I've noticed it a few times before though but it's strictly been with bright portions of images. Not sure if from memory whether any have been from my DF since 90% of the time I use the IQ with my Alpa.
 

Paul2660

Well-known member
Graham:

Is the centerfold the effect where you see the lines that correspond to the areas where the actual sensor is divided into 8 segments? I have never really known what the lines are from, stitching together of 8 sep sensors or coating or what.

If it is the line effect, I often this with my 160 when viewing a LCC on the camera LCD. Much more rarely I will see it in a image file. Mainly in a situation like you describe white sky or us sky. So capture one seems able to adjust these out.

On a recent loaner 180 I got this issue is spades. Especially with the 43 Schneider. All on a blue sky series. Even capture one could not fix them and they turned out to be very hard to remove with content aware. Problem did not manifest itself with any other tech lens or anything taken with a DF.

Just got my 160 back but not sure it's the same camera as before. Mainly in the higher iso ranges. Still need to test it but weather conditions here are not that great at the moment.

Paul
 

GrahamWelland

Subscriber & Workshop Member
Paul,

Yes, centerfold effect is the very small difference in read out value between segments which may be only 1-2 levels in RGB but sufficient to show as a gradient line in the image. With B&W processing it becomes more obvious but I've had it in colour images too as you describe. If you stretch the tones then obviously the effect becomes more visible too.

Btw, I see it both with shifted and non shifted images, with and without LCC.
 
Last edited:

studio347

New member
a little different topic... but is the recalibration recommended in general?
I noticed that the censor is not even... I use LCC all the time and I thought that the unevenness is caused mainly by lens. But nowadays, I'm thinking that the censor also is not perfect too. Is there someone who did this recalibration? Do you want to share the experience and the process for doing it? And is it better after it? Maybe I'm too obsessed with it... maybe it's ok to have a little not-perfect censor.. :) since I know very well... that the perfect censor doesn't result in better images in general...
 

GrahamWelland

Subscriber & Workshop Member
Here's what it looks like. The recalibration is typically done back in Denmark I believe. Luckily I have a VA warranty so not really an issue. It takes certain types of images to make this visible and the attachments are small crops from a much much larger scene.

LCC doesn't correct this unfortunately.

It's getting sorted out by dealer / Phase One and so :thumbs: to both of them for responding even over the holiday weekend / new year.
 

MGrayson

Subscriber and Workshop Member
A blast from the past - this is from one of my very favorite cameras of all time, the Canon 1D. Not s, not II or III or X, just "1D". Only 4MP, but each pixel was ELEVEN microns wide.

It was extraordinarily fast for its day, and read the data off each side of the sensor. Sometimes (very rarely) this happened:



It was either post this, or the results from a GIS of "Digital Sensor Centerfold"

:ROTFL:

--Matt
 

GrahamWelland

Subscriber & Workshop Member
Amr,

All the images taken in the same light. Unfortunately I only really set up for one scene so basically the light was the same. I only see this with certain high-key situations. More contrasty lighting tends not to exhibit the behavior as the tones tend to be spread further I think.
 

GrahamWelland

Subscriber & Workshop Member
Now THAT's a centerfold :eek:

Cute house Matt. That NY real estate certainly looks drafty though.

A blast from the past - this is from one of my very favorite cameras of all time, the Canon 1D. Not s, not II or III or X, just "1D". Only 4MP, but each pixel was ELEVEN microns wide.

It was extraordinarily fast for its day, and read the data off each side of the sensor. Sometimes (very rarely) this happened:



It was either post this, or the results from a GIS of "Digital Sensor Centerfold"

:ROTFL:

--Matt
 

Pemihan

Well-known member
Graham, what being done to your back? Recalibration?

I just got a DM33 which actually exhibits something similar. In some of the images LCC takes care of it on others unfortunately not.
I have both vertical and horizontal lines in the images

Peter

Here's what it looks like. The recalibration is typically done back in Denmark I believe. Luckily I have a VA warranty so not really an issue. It takes certain types of images to make this visible and the attachments are small crops from a much much larger scene.

LCC doesn't correct this unfortunately.

It's getting sorted out by dealer / Phase One and so :thumbs: to both of them for responding even over the holiday weekend / new year.
 

Woody Campbell

Workshop Member
This is actually very serious. Can it be traced to a firmware release? Are people seeing it in files that have not been tortured? In files that are exposed in daylight at base ISO? Does it affect both P1 cameras and tech cameras?
 

GrahamWelland

Subscriber & Workshop Member
Woody,

This is nothing to do with the camera and everything to do with the back. I know a few folks ran in to it with the IQ180 when it first came out. As regardsa the image conditions, yes, base ISO in daylight. It'll be more obvious at higher ISOs. It's not specific to technical cameras either, although perhaps the glass quality exacerbates it.

Basically it's manifesting what you've no doubt seen in live view where the different sections of the sensor are read but the same tonal values differ by 1 or 2 points or more between segments which leaves a perceptible edge between them. This may be both horizontal and/or vertical as the image is essentially built up from multiple square sections.

As mentioned, I only ever really see it with white/grey skies although snow would also show it. Stretch the tones in a B&W image and you'll exaggerate the effect.

It's not really just an IQ issue. Any camera can exhibit this if abused sufficiently. Since clouds, snow & overcast conditions are a big part of my environment and subject matter it's a particular concern for me. Once you see it, you'll start seeing it more and more ...
 

Woody Campbell

Workshop Member
Graham - I disagree with your first sentence (I almost never disagree with you). I lived with this issue with my 60 meg Hasselblad kit - it's ultimately why I sold it. Centerfolding was not a problem with images taken with the hasselblad body and Hasselblad lenses - presumably because the mirror requires a minimum distance between the rear element and the image plane, there are no shifts so the geometry of how light hits the sensor doesn't change, and the manufacturer has a chance to fully optimize firmware for issues like centerfolding in the controlled environment is its own lenses on its own body. That all goes out the window once you introduce all of the variables associated with a tech camera.
 

GrahamWelland

Subscriber & Workshop Member
I certainly can understand how the lens distance of a DSLR lens would cause light to strike the sensor in a cleaner, straighter angle and with a wide Schneider lens it makes total sense that the image would be less "balanced" across the frame and sensor boundaries. If i had an IQ180 I'd be all over the Rodie retro focus lenses for this exact reason.

The only thing is that my example is from a 150mm tech camera lens which has it's rear element perched literally a mile away from the sensor, arguable further away in fact than the rear element of a 150mm DSLR lens actually. Now it does have a huge image circle so perhaps my rationale is flawed and there's still a significant light angle involved even with this lens?

I have to wait until the weekend until I can give the system a full test with the DF in addition to the Alpa. I'll let you know if I can reproduce it with both systems on demand.
 

yongfei

New member
In the past, when we were discussing the possibility of a square (eg 48mmx48mm) CCD, it was often considered that the cost would be prohibitive. I think at that time, few of us realized that the CCD was actually being "stitched" together?

For this problem, the root cause is the stitching. What the software/firmware can do is just to "hide the line". Because there is no physical CCD area overlapping for this kind of "stitching", the tolerance must be very high.
 
Top