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!

CFVii50c strange artefacts and pattern

ErikKaffehr

Well-known member
So not only CCD sensors, but also most recent CMOS sensors are made of four tiles?
Hi,

The steppers used for lithography usually have apertures around APS-H size. So, anything larger than APS-H needs to be stitched. The segments are not tiles, but exposed using different masks. So think that the stepper exposes one part of the sensor for all sensor on the wafer, switches to another mask and exposes another part and so on.

Rumor says that Canon used to have a stepper that could expose a 24x36 mm frame in a single exposure, but it was limited to 500 nm designs rules. Canon's sensors used to use 500 nm rules. Go figure!

But, what I had read, Canon now has a fab with 180 nm design rules.

Best regards
Erik
 

Alkibiades

Active member
This is 20 mm shift image. I will check the 10 mm and 0 mm tomorrow.
schneider 35 xl on 50 and 100 MP sensors from Sony dont work at larger movements, even 10 mm is too much.
It is simply not working combination.
But I never see such effects on my 50 MP backs from Phase one.
The older hasselblad backs like H3D 22, 39, 50 has some issuess with copal shutters at faster speeds- Hasselblad give me the information to use speeds slowes than 1/15 and indeed all was fine. Maybe you try this on your back but with other lens- not with 35 xl!
 

usm

Member
What I found out meanwhile:
There is no sensor tiling when using the back turned to the right side (CCW270) even up to 20mm shift.
When turned to the left side (CCW90) I can see the sensor tiling above 15mm shift.
Same when the back is in landscape orientation (CWW0), sensor tiling visible at 20mm.

Regarding the sensor tiling: Movements until 15mm are ok. When rotated CCW270 I think 20mm are ok.

But there are other problems. I think it is the Phocus Software and not the Sensor.
There are, lets say "artefacts" when processing the files in Phocus. When using Capture One the artefacts are gone!
I converted the FFF fees via Phocus to DNG files and changed the exit data with a script to "Hassiblad".


Rise 20mm, rotated CCW270, middle and middle/top.
Left: processed in Capture One
Right: processed in Phocus
There are horizontal blue and vertical yellow stripes (view at 100%) in the Phocus version. Nothing in the C1 version.
20_CCW270-middle_top-C1.jpg20_CCW270-middle_top-Phocus.jpg

These stripes or artefacts are also very strong on the left and right side of the center. Like on the files:
Rise 15mm, rotated CCW270, left and right
Left: from C1
Right: from Phocus
15_CCW270-left_right-C1.jpg15_CCW270-left_right-Phocus.jpg
When rotating the back to CCW90 they are even stronger and more yellow instead of blue:
15_CCW90-left_right-Phocus.jpg
The same file from C1:
15_CCW90-left_right-C1.jpg

My question is: Are these stripes maybe from a not centered lens?
Or is doing Capture One just a better LCC job?

The problem with C1 and Hasselblad files is the lost of the better color from Phocus.

I can provide FFF files if interested?
All images are made with the electronic shutter.

Greetings.
 

usm

Member
It is hard to see through the forum image viewer. When saved the image or opened in a new tab it’s easier.
 

PSS

Member
have you brought this to the attention of hasselblad? they should be looking into this for several reasons, especially if C1 does a better job at corrections.
 

usm

Member
Well, I did. First answer was “This issue will be fixed if you enable "remove checkerboard patterns" in the noise filter”
This worked perfectly.
And I applied that on all files with the result that it creates a new pattern on a different file in which the sensor was in landscape orientation.
The answer was that this is a Phocus bug and they’re working on that.
 
Top