ACR 4.4 and Clarity

[Mark D Segal]

It's pretty obvious that I can do a better job of calibrating *my* camera in my hand than Thomas Knoll can without seeing it, ever, if I have access to lab grade equipment to spectrally measure my camera's sensor response, and the ability to spectrally measure the lighting under which I take my shot.

 In the same way,  seven or eight years ago Andrew could certainly make a better profile of *his own* printer than the canned one supplied by a manufacturer. Of course, back in the days when he started making profiles the equipment (spectrophotometer) he needed was pretty rare and expensive, and could almost be considered lab grade. Nowadays,the printers are matched to close tolerances, and maybe even Andrew is the consultant who makes those canned profiles, so I guess then the  canned profiles match his own printers
 I'm not speaking off the top of my head here - I discuss this stuff with the camera manufacturer's techs and color scientists and Raw processor writers.

Edmund
[a href=\"index.php?act=findpost&pid=184764\"][{POST_SNAPBACK}][/a]

"It's pretty obvious" doesn't answer my points, because it isn't "pretty obvious". What makes you think the cameras aren't manufactured to close tolerances today? We aren't talking about what may have been the case with printers ten years ago - we're talking about high-end camera manufacturing processes today.  

I wouldn't suggest or assume you are talking off the top of your head, but you are making hypothetical points which may not reflect the reality.

[eronald]

"It's pretty obvious" doesn't answer my points, because it isn't "pretty obvious". What makes you think the cameras aren't manufactured to close tolerances today? We aren't talking about what may have been the case with printers ten years ago - we're talking about high-end camera manufacturing processes today. 

I wouldn't suggest or assume you are talking off the top of your head, but you are making hypothetical points which may not reflect the reality.
[a href=\"index.php?act=findpost&pid=184776\"][{POST_SNAPBACK}][/a]

The *top* cameras are indeed manufactured to very close tolerances these days. At least that's what the guys who make'em tell me -

Edmund

[digitaldog]

The *top* cameras are indeed manufactured to very close tolerances these days. At least that's what the guys who make'em tell me -

Edmund
[a href=\"index.php?act=findpost&pid=184784\"][{POST_SNAPBACK}][/a]

Like they'd say otherwise?

[eronald]

Like they'd say otherwise?
[a href=\"index.php?act=findpost&pid=184791\"][{POST_SNAPBACK}][/a]

Well, I've had indications that some of the non-pro cameras have been foobarred.



Edmund

[dwdallam]

I feel like I'm in a physics class discussing light and human interface devices.  

"Clarity," using it, and its pros and cons over other methods in ACR will be my next new thread.

[madmanchan]

Just go here and scroll down to the section Clarity:

http://www.photoshopnews.com/2007/05/31/about-camera-raw-41/

[Mark D Segal]

Just go here and scroll down to the section Clarity:

http://www.photoshopnews.com/2007/05/31/about-camera-raw-41/
[a href=\"index.php?act=findpost&pid=184934\"][{POST_SNAPBACK}][/a]

Eric, yes - good reference to what it does and how it works. Also well covered in Jeff's book.

The rest of Doug's query though is how it compares with other methods in Photoshop. For starters there's the general point that the more one can adjust up-stream of rendering the image the better in terms of preserving the integrity of the image file. There are several techniques for doing it in PS which can produce indistinguishable results. I mentioned these in a previous post discussing in tandem with Jeff the variations this technique has been through.

[mistybreeze]

Second only to KY, Clarity has become my favorite slider.  

[madmanchan]

The primary technical difference between doing sharpening in ACR/LR (and clarity falls into this category, because it is loosely a form of sharpening) and doing sharpening in PS is that in the former case it is performed in a linear space whereas in the latter case it is performed in a gamma-encoded (non-linear) space. This has subtle impacts on things like the brightness of light halos vs. dark halos, for instance. Of course one has more options in PS to adjust sharpening & clarity-like settings. In ACR/LR there is a single degree of freedom for clarity and four for sharpening.

So the natural follow-up question is whether any of these differences have any practical benefit or impact, one way or the other.

My opinion is that the primary advantage of having these operations performed in Camera Raw vs. PS is that CR will automatically apply the operations (e.g., clarity, sharpening, noise reduction) in an ordered sequence that minimizes artifacts (problems like unwanted hue shifts). For example, in CR you can decide to fiddle with the sliders in any order you want, go back and forth between them any number of times, but CR internally will apply the ops in a manner that makes sense.

In PS, doing operations on sequential adjustment layers -- even though they are non-destructive because you can always go back or turn off the -- is generally non-commutative. Doing a hue/sat adjustment layer prior to performing a RGB curve adj layer is not the same thing as the other way around ...

Eric

[Mark D Segal]

The primary technical difference between doing sharpening in ACR/LR (and clarity falls into this category, because it is loosely a form of sharpening) and doing sharpening in PS is that in the former case it is performed in a linear space whereas in the latter case it is performed in a gamma-encoded (non-linear) space. This has subtle impacts on things like the brightness of light halos vs. dark halos, for instance. Of course one has more options in PS to adjust sharpening & clarity-like settings. In ACR/LR there is a single degree of freedom for clarity and four for sharpening.

So the natural follow-up question is whether any of these differences have any practical benefit or impact, one way or the other.

My opinion is that the primary advantage of having these operations performed in Camera Raw vs. PS is that CR will automatically apply the operations (e.g., clarity, sharpening, noise reduction) in an ordered sequence that minimizes artifacts (problems like unwanted hue shifts). For example, in CR you can decide to fiddle with the sliders in any order you want, go back and forth between them any number of times, but CR internally will apply the ops in a manner that makes sense.

In PS, doing operations on sequential adjustment layers -- even though they are non-destructive because you can always go back or turn off the -- is generally non-commutative. Doing a hue/sat adjustment layer prior to performing a RGB curve adj layer is not the same thing as the other way around ...

Eric
[a href=\"index.php?act=findpost&pid=185018\"][{POST_SNAPBACK}][/a]

Eric, I'm with you till the last para. This is something I hadn't heard before. I thought these Adj Layers would do the same thing regardless of the order in which they are implemented - but what does matter often is the order in which they are stacked - not necessarily the same thing of course.

[digitaldog]

Eric, I'm with you till the last para. This is something I hadn't heard before. I thought these Adj Layers would do the same thing regardless of the order in which they are implemented - but what does matter often is the order in which they are stacked - not necessarily the same thing of course.
[a href=\"index.php?act=findpost&pid=185071\"][{POST_SNAPBACK}][/a]

Just move the order around and I think you'll see, depending on the adjustment, there's a difference. For example, make a curves adjustment layer, then an Exposure adjustment layer and switch the order. Or cruves and Photo filter etc.

[madmanchan]

Sorry, I meant the stacking order, not the creation order, of the adjustment layers.

[Mark D Segal]

Sorry, I meant the stacking order, not the creation order, of the adjustment layers.
[a href=\"index.php?act=findpost&pid=185089\"][{POST_SNAPBACK}][/a]

Ah - for sure! Makes total sense now.

Mark

[Philmar]

Well, just to get back on topic of the actual clarity slider, and how to use it......

As a digital noob I am confused by the Clarity slider in ACR. I have read Mr. Schewe's article about the changes to ACR and how to use the Clarity slider. It advocates judging it's application while viewing the RAW file at 100% view. But I never see any halos, even when applied liberaly (at least not to the extent that I see halos when I use the sharpening sliders at 100% viewing)
As a result, I usually back off the Clarity slider at around 10 (except with shots with lots of haze) out of fear of over-processing a shot.

Do people apply Clarity to most of their shots?
How much? Yes I know it depends on the shot but are there any general rules of thumb?
Are there types of photos (i.e. like hazy day shots, shot from a long focal length lens) that benefit more from Clarity than others?
Are there types of photos where use of Clarity would be of no benefit?

[Mark D Segal]

Well, just to get back on topic of the actual clarity slider, and how to use it......

As a digital noob I am confused by the Clarity slider in ACR. I have read Mr. Schewe's article about the changes to ACR and how to use the Clarity slider. It advocates judging it's application while viewing the RAW file at 100% view. But I never see any halos, even when applied liberaly (at least not to the extent that I see halos when I use the sharpening sliders at 100% viewing)
As a result, I usually back off the Clarity slider at around 10 (except with shots with lots of haze) out of fear of over-processing a shot.

Do people apply Clarity to most of their shots?
How much? Yes I know it depends on the shot but are there any general rules of thumb?
Are there types of photos (i.e. like hazy day shots, shot from a long focal length lens) that benefit more from Clarity than others?
Are there types of photos where use of Clarity would be of no benefit?
[a href=\"index.php?act=findpost&pid=188523\"][{POST_SNAPBACK}][/a]

The 100% view is useful for evaluating the appearance of halos. Otherwise it is best to use it with the image in full screen mode allowing you to assess its aesthetic impact on the photograph - which does vary from image to image and according to the taste of the photographer. When the image starts looking artificially "clarified", back it off a bit. I also leave a bit of headroom for capture sharpening because the combined effect of the two can be a bit much, but this as well is largely a matter of taste.

[ejmartin]

Furthermore, to improve upon the canned math in ACR you would need two necessary conditions: (a) you can do a better job profiling the camera than Thomas Knoll can, and ( there is noticeable copy-to-copy variability in the capture characterestics of the camera. My guess is that you would be very hard pressed on (a), and for ( - it doesn't apply - at least to the higher-end pro DSLRs, such as 1Ds series, 5D, D3, etc. I think that's why the majority of users, me includeed, ignore the calibrate tab.

Some prima facia evidence to the contrary:

http://luminous-landscape.com/forum/index....opic=17064&st=0

Every camera I have owned has inaccurate reds coming out of ACR, and calibration greatly improves the accuracy of the color rendering relative to the ACR defaults.  Moreover, the shifts on the red calibration sliders are *ALWAYS* in the same direction, and it's not just my random bad luck of camera samples (including my 1D3), as the above referenced thread indicates.  It seems a lot of people can do a better job than Thomas Knoll   since the median values of the calibrations reported is much different than zero.

It also seems to me that the ACR calibration tab does not allow the full latitude of a 3x3 matrix, which would be the most general transformation of the color data for each pixel (after Bayer interpolation) preserving the linearity of response that is a property of raw capture.  Granted, one degree of freedom is an overall rescaling which can be absorbed in the exposure slider of the main tab; but eight remain.  The calibrate tab only has six sliders (ignoring shadow tint, which I presume makes a shift based on luminance level and not a linear transform of the data) and so two parameters cannot be adjusted.  Optimization of the calibration could be improved if those two additional parameters were available on the calibrate tab.

[digitaldog]

Every camera I have owned has inaccurate reds coming out of ACR, and calibration greatly improves the accuracy of the color rendering relative to the ACR defaults. 

OK, that problem is solved.

Moreover, the shifts on the red calibration sliders are *ALWAYS* in the same direction, and it's not just my random bad luck of camera samples (including my 1D3), as the above referenced thread indicates.  It seems a lot of people can do a better job than Thomas Knoll   since the median values of the calibrations reported is much different than zero.

Well we don't know the sample of users who fall into this camp. We don't know the exact conditions Thomas used to build the profiles or shoot the targets (pretty sure he's using a Macbeth Judge booth to capture his Macbeth). But Thomas was fully aware that his conditions and yours would likely not sync up, that's why he built the calibrate tab.

The question is, does it not provide an adjustment, along with the other sliders to produce a desired color appearance?

It also seems to me that the ACR calibration tab does not allow the full latitude of a 3x3 matrix,

That is the case.

[ejmartin]

Every camera I have owned has inaccurate reds coming out of ACR, and calibration greatly improves the accuracy of the color rendering relative to the ACR defaults.

OK, that problem is solved.

Yes, it is solved for me because I have taken the time and effort to do the calibration.  I am suggesting that if Adobe wanted to improve their product for the vast majority of Photoshop users who do not go to the trouble, they might want to investigate why their calibration procedure is so inaccurate according to the end results obtained by comparing GMCC conversions in ACR to reference values, and why there is a systematic bias to calibration efforts by the end users who have gone to the trouble to calibrate their cameras.  I would hope that Adobe would be concerned about such a systematic calibration bias reported by their customer base.  

It also seems to me that the ACR calibration tab does not allow the full latitude of a 3x3 matrix

That is the case.

Well, why not?  If one is going to offer the possibility to do user calibration, why not offer the full latitude of possible adjustments?

[digitaldog]

Yes, it is solved for me because I have taken the time and effort to do the calibration.  I am suggesting that if Adobe wanted to improve their product for the vast majority of Photoshop users who do not go to the trouble, they might want to investigate why their calibration procedure is so inaccurate according to the end results obtained by comparing GMCC conversions in ACR to reference values, and why there is a systematic bias to calibration efforts by the end users who have gone to the trouble to calibrate their cameras.  I would hope that Adobe would be concerned about such a systematic calibration bias reported by their customer base. 

We need to do a lot more sampling of users and camera types before we go out and say for a fact there's an issue here. Lets say Adobe (Thomas) tweaks the existing profiles. Do we know for a fact that a larger audience would have other color issues? The tool exists so any user can tweak the calibration. I'd be pretty shocked if everyone needed to do this BUT Thomas. He's a pretty bright guy, knows a thing or two about image processing, Raw rendering and color management (he wrote the application that builds the Adobe ICC profiles installed, the DNG converter and originally Photoshop).

You seem pretty sure of yourself that this is a systematic issue with the calibration, with not much to back it up other than personal experience. If you have more solid numbers, we're all ears.

[ejmartin]

We need to do a lot more sampling of users and camera types before we go out and say for a fact there's an issue here. Lets say Adobe (Thomas) tweaks the existing profiles. Do we know for a fact that a larger audience would have other color issues? The tool exists so any user can tweak the calibration. I'd be pretty shocked if everyone needed to do this BUT Thomas. He's a pretty bright guy, knows a thing or two about image processing, Raw rendering and color management (he wrote the application that builds the Adobe ICC profiles installed, the DNG converter and originally Photoshop).

You seem pretty sure of yourself that this is a systematic issue with the calibration, with not much to back it up other than personal experience. If you have more solid numbers, we're all ears.
[a href=\"index.php?act=findpost&pid=188762\"][{POST_SNAPBACK}][/a]

Yes, more sampling needs to be done.  But there are close to thirty calibrations reported in the thread I linked to, all but one point in the same direction -- negative red hue correction, positive red saturation correction.  A quick google of

camera raw calibration "red hue"

yields quite a few more reported values

Red Hue -16 Red Saturation +35
red hue -21 red sat 29
Red Hue: +8 Red Saturation: +20
Red hue: -22 Red saturation: +39
Red Hue: -13 Red Saturation: +20
Red Hue = -16, Red Saturation = +20
Red Hue -13 Red Sat 8
Red Hue: -14 Red Sat.: 22

and this is just from the first 20 entries that Google returned (out of 5350, though I presume only a few hundred are on point).  So I don't think it's just my personal experience.  Different camera models to be sure; lots of spread in the data, to be sure; but the mean/median shows systematic bias which is increasingly hard to discount.

I'm not saying it can't be corrected for (that's what the calibration tab does, even if it has only six out of the eight controls it ought to have), I am simply pointing out that the software ships with a built in bias, according to a large number of user reports.