I agree with Andrew and Eric that no raw converter will be technically "perfect", whatever that means. Therefore one needs a raw converter with the best capability for translating the image data into a rendering we like best - judgment call.[{POST_SNAPBACK}][/a]
Oh, the problem as described by Bill et al. is very well known:
>> There is currently a fundamental tradeoff between the skin tone rendering and the rendering of deep saturated reds. For many individual images, this can be worked around using the calibration & HSL sliders. However, no single preset will tackle both issues generally (i.e., across several images). And for some images that contain both, there is no workaround at the moment.
Eric<<
[a href=\"http://www.adobeforums.com/webx?14@@.3bb6a85c.59b4cc97/8]http://www.adobeforums.com/webx?14@@.3bb6a85c.59b4cc97/8[/url]
Some of us like to contemplate about the root cause, however, I recognize that there was a kind of bug fix announced in the course of this thread. Let’s see.[a href=\"index.php?act=findpost&pid=189256\"][{POST_SNAPBACK}][/a]
Peter,
If you read what I said, I think we'll agree I didn't say there are no issues.
[a href=\"index.php?act=findpost&pid=189266\"][{POST_SNAPBACK}][/a]
Mark,
Referring to the compiled term “technically perfect Raw converter”, which I read in your post, this for sure has some relevance for the initial step of colorimetric interpretation, or what is called color reconstruction. Wouldn't give up this claim too early. "Accuracy" and some perfection are supposed to lay a better ground for all the creative rendering tools (including Clarity) which are set on the top. Or, to say it in Bruce Fraser’s words: >>"Accurate" is not necessarily the final goal, but it tends to make for a much better starting point than randomly inaccurate.... <<.
Mentioned antagonism of pure reds vs skin tones is a measurable effect with regard to hue accuracy. Get one right (via calibration) and the other one becomes worse. This “balance” was found to be surprisingly consistent across different camera models and brands.
While numbers can always fool you (me), it happens that many of us are able to see this issue because typically we don’t like orange reds and/or magenta skin tones. Of course, maybe depending on cultural background, one could see it differently. However, while this subject comes up every now and then, people obviously tend to recognize what the numbers confirm. So for the moment, let’s believe in this claim for initial (hue) accuracy with these colors.
The issue you refer to in the Adobe Forums can also be atttributed to the fact that ACR is not yet equipped for localized adjustments within a single colour group.
That would certainly be nice – however, at the end of the day, it can just remove the symptoms, but not the root cause.
As well, I haven't seen any reference to a bug or a bug fix in this thread. If you are alluding to Jeff Schewe's post, I wouldn't presume he is talking about a bug fix. I expect, without knowing, it is something much more fundamental.
No problem to agree and to withdraw “bug fix”. Something “fundamental” might be more appropriate.
Anyway, I’m glad to see that the threadopener seems to be happy with the discussion and I certainly don’t have insist on this off topic part.
Peter
--