Home · Register · Join Upload & Sell


  Previous versions of gdanmitchell's message #16110364 « R5 mkII rumors, questions »


Upload & Sell: Off
Re: R5 mkII rumors, questions

Scott Stoness wrote:
gdanmitchell wrote:
Regarding AA filtering and image detail: Back with Canon introduced the 5Ds and 5DsR, folks were still very concerned about the prospects for aliasing if AA filtering were to be eliminated. For this reason, as Nikon did with their D800, Canon did with the 5Ds bodies providing the 5Ds with the familiar AA filtering (to assuage concerns about aliasing) and the 5DsR that cancelled the AA filtering (to assuage concerns of less-than-maximum sharpness).

At the time I was trying to decide whether to get the 5DsR or the 5Ds. I got my hands on some carefully shot raw test images from both camera and put them to the test. It was true that if I began with raw files at the same default settings and looked really, really closely... the 5DsR files appeared to be a tiny bit sharper.

However, that is a pretty meaningless test, since that's not how we use post-processing workflow. Instead we optimize our workflow for whatever camera we us, working to get the best result possible from its files. So I did just that, ending up with somewhat different optimized sharpening settings for the 5Ds and the 5DsR. (Basically, I had to sharpen the 5Ds files just a little bit "more.")

It isn't any "easier" or harder to sharpen the files one way or the other. I have to set the same parameters either way — just using slightly different values.

Once I did this, the difference was so tiny as to be meaningless. If I looked at 100% and larger files side by side on a large computer screen I could sometimes convince myself that the 5DsR files were maybe, possibly, arguably, sort of a tiny, tiny be sharper on the screen... but completely indistinguishable in real world prints.

In the end, I decided it didn't really matter. Basically both were excellent, and any differences in the finished images were meaningless. I went with the 5DsR since, well, I had to pick something.

So, as much as I feel that AA filtering is pretty much unnecessary on high resolution cameras, if Canon decides to apply minimal AA filtering that improves video... I don't really care. :-)

I would rather sharpen less and have to sharpen less. It's just easier. And most measurements I have seen indicate no aa is achieves more resolution - but this is on straight lines so as you say its hard to notice.

Unless you don't sharpen on once case and do in the other (not likely with raw files, right?) how "much" you sharpen in immaterial if the end result is the same. I don't recall the specific settings I used with the two cameras, but my recollection is that it was just a matter of slightly different settings for amount, radius, and perhaps details. This is analogous to the way I use different settings with different types and brands of cameras — for example, I used to use different settings back in the 5D era, and I use different settings now when I shoot my Fujifilm system.

In any case, my point was that should Canon apply some small amount of AA filtering to a hypothetical R5II, in the end I don't think it would make a consequential difference.

By the way, back when we had a choice between versions of cameras that did and did not apply AA filtering, after I realized that it didn't make any significant difference in the photographic end result I ended up getting the camera...

... that doesn't apply AA filtering. Given a choice, I didn't think it made sense to apply a process to the images that wasn't adding anything of value. That's also in line with my belief that, at least for stills photography, there's hardly any good reason to apply AA filtering on today's high resolution sensor cameras.

Dec 03, 2022 at 04:37 PM

  Previous versions of gdanmitchell's message #16110364 « R5 mkII rumors, questions »


This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.