Hasselblad 100C and 35XL

[tenmangu81]

I've briefly read this thread, and can't understand why the 100C could give rise to banding when used with a tech cam, while the 50C won't. One of the differences between the two DB is PDAF, and I've read that it could be one of the reasons. Could somebody explain me why ?
I don't know much about tech cams

 

[Alkibiades]

For reference, until today I did not spot any banding with my SK60xl.

did you make bigger movements ?

 

[MGrayson]

I've briefly read this thread, and can't understand why the 100C could give rise to banding when used with a tech cam, while the 50C won't. One of the differences between the two DB is PDAF, and I've read that it could be one of the reasons. Could somebody explain me why ?
I don't know much about tech cams

I'll give the high-level stuff. PDAF is an automated software split-prism. One short strip of pixels looks only at the left side of the lens, another - located in the same place on the sensor - looks only at the right. If that spot is in focus, the two strips will line up. If one is shifted with respect to the other - just the way a split prism image is when not in focus - the software figures out the best way to move the lens so that they will match.

With an SLR (the key letter being R for Reflex), this happens away from the image sensor on separate focus sensors. On mirrorless, the focus sensors are built into the image sensor. That means that some pixels are different, and that difference needs correction in the imaging pipeline. When that correction fails, lines appear. It must be shift-dependent.

That's all the detail I can give.

Matt

(The name Phase Detect comes from the correlation between the two strips under different shifts. When they match most closely and the correlation is high, they are "in phase". This is exactly how you would determine the phase of a signal. You shift your test sine wave until it best matches the signal sine wave. That offset is the phase. On a camera, that offset is how much the focus motor has to move.)

 

[anyone]

did you make bigger movements ?

Yes, -25 rear fall. Please note, I did not try to get the stripes visible through a lot of editing. It's more a real-world observation, based on one picture.

It was this one:

https://www.getdpi.com/forum/index.php?attachments/b_0274_web-jpg.211579/

 

[MGrayson]

Oh, a data point. I just took this - posted in Fun with MF Images 2024 - with a Mamiya 645 50mm lens shifted 12mm on an X2D (same sensor) with even more extreme processing than shown here - 100% Dehaze, 100% Clarity, and contrast boost - it shows no lines in the upper corners. Now the rear element of that lens is quite far from the sensor, as it was designed for a 645 reflex camera. I note, also, that the shift here is parallel to the lines, so that may help mask them.



Matt

 

[akaru]

Beautifully explained, @MGrayson. It’s sort of like a rangefinder

 

[tenmangu81]

Thanks a lot Matt for your very clear explanation !

 

[Bill_Evans]

Has anyone who moved up to the 100C from the 50c started using this new back in their production tech cam workflow? I'd be interested to hear how the trade-offs measure up. Specifically the reduction in color casts, and maybe tiling of the wide Schneider lenses, vs this new "feature" of PDAF lines. The 50c doesn't seem to be without its workflow needs, ie LCC & shift limits. Curious if the new back is DOA, or just a new set of workflow hurdles.

 

[Alkibiades]

any news to the stripe problem?

 

[tenmangu81]

Looks like the last version of Phocus has just been released. Maybe they address the issue of the stripes with T/S, in the same way they did with the magenta cast observed with the XCD28P. Dreaming is always possible....

 

[chriswebb]

New features in 3.8

HEIF support

• With X2D/CFV100C FW 3.1.0, it is now possible to capture to 10-bit HEIF format.

Phocus 3.8 adds comprehensive import and editing support for Hasselblad HEIF​

images. With few exceptions, Phocus editing tools work for HEIF as they do for RAW.​

Editing HEIF is non destructive as changes are stored in sidecar files.​

Vignette correction and scene calibration is not supported for HEIF files.​

​

Performance enhancements

• Viewer performance has been improved resulting in smoother and more precise editing, particularly when using the exposure tool.
• Performance of thumbnail browsing has been improved, keeping Phocus responsive even when working with large numbers of images.
• Multi-shot images also enjoy faster render at 100% zoom and export.

Support for new hardware

• This release introduces support for the new 907X CFV 100C camera.

General improvements

• A large number of fixes and stability improvements have also been included in this release.

No mention of any specific fixes for this issue. Also the new release appears to only be for Macs so far.