Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
Chromatic aberration and "post-processing"
[quote name='PuxaVida' timestamp='1308211140' post='9274']

I tried to correct the image you loaded Ian...

CS4: Image Adjustments --> Hue / Saturation. Then select the channels and reduce lightness and saturation. The channels selected were magenta, green and blue respectively.

Oh, and I have to correct myself regarding the cause of PF. After a quick research on the net, there's no proof that the microlenses over the sensor cause PF. In general, people are divided in two:

- people say that PF is caused by the lens (because it is effected by chaning the focus in / out).

- people say that PF is caused by the sensor (because PF is less seen on film compared to digital).

And a few words about the recent discussion: the content was fruitfull and it could be even better if you BC could be careful about the way you contribute. I personally believe that you have a decent knowledge but with this way of approach they become less valuable than they deserve. I read posts of Wim for a long time here, and I never saw him refusing to step back concerning his arguements in case they are wrong (and I don't know if he is right / wrong in this case, this is not my point). Digital photography might be consisting of 0s and 1s, but social environments are not. I hope we can see posts from both of you here in future...



Purple fringing is caused by the sensor and lens (one needs to establish first that LoCA's and PF are two different phenomena, else the discussion about what causes it gets murky and polluted). If it was the micro lenses, one would se it with every lens that lets through lots of light. And this is not the case, it is lens model dependent.

One does see LoCA on film too, but if one separates LoCA purple occurrences from PF purple occurrences, PF is seen with sensors, not with film photos (at least, I have not seen any PF in film photos yet).

The problem on internet that at times real knowledge is not easily found, or is just not there. One example of that is how phase detect AF works on SLRs. Another is what causes PF. It is not a mystery in the world, just on internet.. Camera and lens manufacturers know the exact reason (and have taken counter measures with new lens designs), we do not.

The most plausible cause, in my opinion, is that sensors reflect more light back than film. That light somehow wreaks havoc with the back element of the lens, which explains why with some lenses it occurs and with others it does not, and why with film it doesn't show up and with sensors it does.

What then remains to be explained is how the light affects areas next to the brightness, but why it does not get scattered much more. Maybe that has to do with the shape of the back element, but then that would make PF usually mainly a problem in the center of the image, which is not the case.

Another idea is that PF is caused by parts of the UV spectrum, where some lenses filter UV more successfully than others.

If you read my first post to Wim's post, my post was just a normal reaction with no negativity. Then read Wim's reaction to that post.
I clocked last night - after finishing work late - that there'd been a lot of feedback (far more than I'd anticipated) to this post. I knew it would take me some time to work my way through it, so didn't post a reply then.

Just logging in now, I haven't read most of the posts but it's very quickly apparent that this has sparked debate which has degenerated yet again into argument and antipathy...

I really don't understand why this has to be the case, and what people think it achieves, nor what they get out out of it. I don't mean to preach, but as I've said here before, direct argument rarely, if ever, persuades people to change their point of view - though perhaps I should read what people have said first

I'll read the posts later as I still have more work to get done this evening.

Meanwhile, thanks for all the positive contributions - I know there have been many...! <img src='<#EMO_DIR#>/smile.gif' class='bbc_emoticon' alt='Smile' /> - and a quick reply to Pinhole's question from yesterday.

[quote name='Pinhole' timestamp='1308139898' post='9257']

Ian, it looks like you are using a very small part of an image - do you need to crop it so much for real-life usage, or did you only do that to demonstrate the problem?


I don't need to crop that much, as you'll see from the image below. It was just to demonstrate the issue, wondering if (i) a quick fix (like automatic post-processing of CAs) was possible. I was also wondering (ii) why the lens had produced such strong aberrations and (iii) why the profile corrections hadn't sorted this in Lightroom. I think I know the answer to (iii) now, and am fairly confident the answer to (ii) will be in the thread once I've sorted through everyone's responses...

Compressed version of the full image attached.

Thanks folks.


Forum Jump:

Users browsing this thread:
1 Guest(s)