Capture one and pseudo chromatic aberration

[jaapv]

Those new lenses... The 24/1.4 seems to have a bit of CA. Even Sean Reid comments on it. I found it is not CA, it is the RAW Converter.

4 100% crops.

First C4 an old version

Then C4.8 (the latest update)

Then ACR ( CS4 and Lightroom)

Then again C4.8

The newest version of Capture one eliminates nearly all purple fringing. I think it is not CA, but such a high microcontrast that it induces sensor blooming, and C4. deals with that more efficiently

Capture one C4 old

Capture One C4.8

ACR

Capture one C4.8

 

[D&A]

Hi Jaap,

That's quite interesting and certainly warrents additional investigation. Is it possible to repeat above test with the RAW converters but using the just a single (one) selected image (strongly backlit or an image that shows a substantial amount of CA?). Thanks!

 

[jaapv]

Well, I think that both images, and especially the 2nd are in that category, but I would indeed be happy if anybody else could investigate this as well.

 

[willie_901]

Well, how do you know which converter is portraying the true, but unknown, amount of CA in the optics?

Capture might be correcting for the lens artifacts (CA) with proprietary algorithms while ACR is not. After all Leica and Phase One have a business agreement ( www.letsgodigital.org/en/10414/leica_m8_capture_one/ ) .

So far your experiment shows Capture One is clearly better than ACR for M8 RAW, but that doesn't mean that the lens does not have significant CA artifacts.

And, there must be a good film-based test for CA. I suspect that your suggestion that ACR doesn't handle Leica RAW well may be correct. But analog CA tests would confirm your speculation.

 

[splitimageview]

It's more likely the lens, as opposed to sensor bloom, as sensor bloom is not particular to a given color. One way to eliminate sensor bloom as a variable is to defocus the lens slightly and see if the fringing changes colors.

 

[johannielscom]

It's more likely the lens, as opposed to sensor bloom, as sensor bloom is not particular to a given color. One way to eliminate sensor bloom as a variable is to defocus the lens slightly and see if the fringing changes colors.

Seems to me Jaapv has used identical shots on both ACR and Capture One, so it must be the software.

I am very accustomed to ACR and LightRoom (well, LightRoom mostly) and would be sad if this would prove a problem for Adobe products.

Jaapv, can you add LightRoom results as well? Or, send the original DNG so I can post the LightRoom results of the lamp shot myself? Would be most obliged!

 

[splitimageview]

One could surmise that the issue inherent to the lens, however, as opposed to a sensor issue. One software is correcting for it, the other not. It can be corrected manually pretty easily.