I am going to try using ACR for capture sharpening with a 1dsMkIII after regularly using PS. Does anyone have a any recommendations for initial settings to start with for this camera? And, possibly for several shooting situations ie. landscape, fine detail, portrait.
Thanks for any suggestions
Jonathan

First thing I've done is to make my own DNG camera profiles using the Adobe DNG Profile Editor. It's the best path toward better, more accurate color from your camera when using ACR.

Are you able to provide examples Chris?

I appreciate that you are demonstrating the possible use of the profile editor, but my question does involve the suggestions for the use of the detail panel in CR for input sharpening using the 1dsMkIII. At the moment I am not trying to correct color etc. Please try to limit replies to that issue right now.
Thanks
Jonathan

It depends on the image. There is no "one-size-fits-all" recommendation.

However, as a starging point you may wish to be guided by the presets provided with Lightroom. Lightroom provides "Landscapes" and "Portrait" presets. The former is for higher-frequency images and the latter for lower frequency images. In the former, the default settings are Amount 40, Radius 0.8, Detail 50 and Masking 0. In the latter the default settings are Amount 35, Radius 1.2, Detail 20 and Masking 70.

The underlying approach is as follows: Amount determines the extent to which Radius and Detail are applied. Higher Amount applies them more strongly. For Radius, lower values provide more narrow edge detail. A general recommendation is to use below 1.0 for high frequency images and more than 1.0 for lower frequency images. Detail distinguishes edges from halos. Lower values suppress halos and allow higher Amount for more edge sharpening; higher values accentuate halos and shows more of the sharpening effect. Increasing Masking increasingly protects areas from sharpening which you don't want to be sharpened - such as skin in a portrait.

You can only see sharpening effects with the image magnified to 100%. It may also be helpful to view the effect in B&W by depressing Alt when clicking on the Sharpening control of interest.

You can see for yourself. Follow these instructions and when you're done simply compare the results with the existing camera profiles. The difference is huge. In every color sample, there was a significant shift in hue, saturation and/or lightness. (I used a new ColorChecker and dual illuminants).

http://www.photoshopforphotographers.com/p...uresharpen.html

First time I ever tried ACR's capture sharpening was after watching this, couldn't get my head round it beforehand.

Just tried this with a 'real' image, a paltry 55 megapixel architectural stitch. Sorry guys but the mixture of ACR capture and output creams PK IMO, far less mush in the fine details on problem subjects, I also found the output sharpening (normal) more to my taste than PK's. Faster to apply as well. Of course it's harder to use, no masks to work with, resizing for print, etc.

What I did was take my 2.2Gb file, flatten it, load it into ACR and apply Capture, open to PS for resizing (easier than trying with the clunky ACR method of resizing), save, reopen in ACR and apply Output based on the print size, open and print! Of course if you want masks for the capture all you do is open an unsharpened version then layer the sharpened on top. Bit of a pain but I'm so happy with the results I don't particularly care. What would be very very nice would be a proper size window in ACR where I can specify the size in inches and the relevant DPI. That would cut a whole stage out of the workflow but I know it's high on the wishlist for many with ACR...

For as far as the sharpening controls in ACR/LR can take you, you can get the same results sharpening in ACR or sharpening with PK in Photoshop. Perhaps the adjustment responds faster in ACR because it's working on a raw, un-rendered image, or for any image - applying meta-data (parametric editing), with the processing time being shifted to the Export function.

I've been using PK for 3 years and I used the ACR sharpening as per the above video for the first time today. I've never been able to get PK to look as good as the ACR.

I did look at Evening's CS4 video and it does give a very complete and understandable picture of the whole capture sharpening process. Can you please go into more detail on the explanation you gave as to switching back and forth from ACR to PS to achieve your sharpening workflow, and inadditon, I don't think I understand the point: "Of course if you want masks for the capture all you do is open an unsharpened version then layer the sharpened on top." I thought you could use the Masking slider in ACR for capture sharpening?
Thanks to all for the helpful ideas.
Jonathan

You know, this can be kept really simple - and you CAN get just as good results using EITHER PKS in PS OR the sharpening panel in ACR/LR. Starting with the latter, there are no layers in this workflow, so anything to do with layers doesn't apply. You use the Masking slider to hide area you don't want sharpened from being sharpened. That's all there is to it. Turning to Photoshop, ANY sharpening you do with PKS is automatically done on two layers covered with "pass-through" layer encompassing the Light Contour and Dark Contour Layers. If you wish to apply PKS selectively to Capture or Output sharpening all you need to do is add a Hide All layer mask to the pass through layer, grab your WHITE paint brush and paint in the sharpening where you want it to the extent you want it (with opacity adjustment on the brush). Another approach with PK is to let PK create a brush using the Creative sharpening algorythms. PKS is an infinitely flexible instrument.

OK, if you get halo's with PK then you have a mask that you can paint to reduce the effect. You can deal with halo's within the capture sharpening in ACR but if it a few specific halo's but the rest is OK then you have to open one version with zero sharpening then open a 2nd version with the sharpening and layer it on top of the first one. Then you can mask the sharpened one and adjust opacity or paint out areas using the mask. If you open the sharpened layer as a smart object then you have even more control of course. A neat trick is using the snapshots tab in ACR to save your sharpening so that you can open with or without easily.

Once I've opened the file with capture sharpening I can then work on the file as usual in PS with all my mutiple layers, etc. Once I come to the stage of resizing and sharpening for print I will flatten the image, resize it and then save it as a new TIFF. I then open up that TIFF in ACR and apply output sharpening which will be applied relative to the final print size that I set the image to earlier. I can open the file again as a smart object so that I can choose between the output settings or even open a new layer with no output sharpening so as to again be able to adjust the opacity.

At least that's how I see it. The capture and Output in ACR don't seem to have been designed to do the above, they seem to be made for a ACR one stop processing solution. Doesn't help me much when I have files that are going to have multiple layers applied!

Ben and Mark, thank you both for taking the time to point me in the right direction. I am going to try several of the suggested options in the next few days and I will post a follow up of any applicable results.
Jonathan

Ben, sorry, but I find much of this needlessly convoluted. The sharpening in ACR was designed to make one's sharpening life efficient and effective. You can even sharpen selectively using the adjustment brush. There is no output sharpening algorithm in ACR that I've seen - as much as I've seen so far, only the one in the Detail tab which is intended as a capture sharpener. The workflow is intended as a straight one-way pass through from ACR to PS. Lightroom is a different proposition because it has a print module with some layout features which may attract one to go back there for printing, so it has an Output sharpener.

Actually Mark if you look in the Workflow OPtions you'll see that Output Sharpening has indeed been added in Camera Raw 5.2 and above...The Adobe/Pixel Genius deal was always intended to have capture and output sharpening in the Camera Raw pipeline (but not in Photoshop–yet) and with 5.2+ it's there. Of course, Camera Raw resize/resample usability sucks at this point (meaning it's tough to know what size your print will be) so I would only use output sharpening from Camera Raw ONLY if no further resize/resampling will be done on an image. That should improve in the future...

That's what I've been using, I just resize in PS then reopen in ACR to apply the output sharpening as per the above.

I've never actually seen the adjustment brush in ACR do anything in sharpening mode, certainly not on the level of PK's creative brushes.

Oh yes, indeed - I stand corrected - what a "discrete" place to put it! I would have expected a more obvious placement like we have in LR.

Anyhow, regardless of it being there, I still would not be sending images back and forth between ACR and PS for reasons of sharpening. I don't think it's necessary or efficient.

You also have to wonder if opening in PS-resizing-correcting etc and then reopening a flattened tif in CR to sharpen might mess things up somehow. If so, and you're really sold on this technique, maybe it would be better to save it originally out of CR as a smart object that can just be reopened later in PS for the output sharpening. Just a thought.
Jonathan

Hi, firstly you save a lossless TIFF so there will be no image degredation. It's not a jpg. Secondly, the smart object idea won't work, as far as I know, yup, just confirmed it. Opened an image as a smart object, resized it, double clicked on the smart object and it brought up ACR showing the original size.

As far as why I'm bothering Mark? Firstly I vastly prefer the sharpening. Secondly I have to close down PS 64 and open 32 bit just to use PK so if I'm opening and closing anyway I might as well use the method I like which takes very little time anyway as I have Bridge and PS open at the same time anyway. Thirdly it's faster. I have paid for PK and had it for a long time, I just find the capture sharpening controls more to my taste in ACR and the output sharpening more natural. Everyone to their own....

BTW would there be any interest in ACR/LR offering types of Output Sharpening to match type of printing, i.e. Contone/inkjet/halftone as there is at present in PK?

Ben, yes of course everyone to their own - I'd be the last person to dispute that; but at the same time I'd be really curious to see the difference in outcomes between each of these methods properly applied. Pixelgenius has issued a 64-bit beta for PKS if you are running Windows Vista (not me - ever).

Now, when you flatten your TIFF, it is lossless as Jonathan says, but then you lose all your layers and your "non-destructive" workflow. Of course you can avoid that by saving duplicates with the workflow minus the output sharpening you do in ACR. Also, introducing Smart Objects into a Photoshop workflow also carries with it some limitations (I have gotten pop-ups that such-and-such operations can't be performed on a SO). Well, as you say, each to their own.

Ben, that would be reasonably straightforward to add (from a technical perspective), but to be honest we have not heard many requests for the additional modes so far.

The current output sharpening in ACR/LR already do Inkjet and Contone–since in our testing, there was a close relationship to both. We decided that there didn't "need" to be two separate sharpenings. As for halftone, as Eric says, not hard to do if there is interest but at the moment one still needs to actually go into Photoshop for CMYK conversions and more accurate image sizing. If either of those situations change, then it may indeed make sense to put halftone output in ACR.

Its in V5, in the workflow options.

If they are close enough not to make a difference then fair enuf!

[quote name='pom' date='Mar 18 2009, 05:24 AM' post='268737']
Hi, firstly you save a lossless TIFF so there will be no image degredation. It's not a jpg. Secondly, the smart object idea won't work, as far as I know, yup, just confirmed it. Opened an image as a smart object, resized it, double clicked on the smart object and it brought up ACR showing the original size.

Ben.
I just tried several version of using the smart object:
1. Open image in CR-correct and capture sharpen-open it as smart object in PS
2. Resize using image size command (keeping the pixel counts the same) and add adjustment layers-double click smart object so it opens in CR
3. Add output shapening click done
Back in photoshop the image is output sharpened, no flattening of layers etc.
When resizing (actually cropping so that the pixel counts change) is when things get problematic and the sizes for the sharpening don't match back properly in PS.
You probably know all this already, but I had to try it out myself so as to get a better understanding of it.

In addition, I printed an image with some fine detail first using CR for capture and output sharpening and another version using CR capture and Martin Evening's (Bruce Fraser's High Pass) output sharpening in PS. The CR in and out version is somewhat soft and less defined compared to the CR in and PS out version (which to my eye is preferable).
Jonathan

Bit confused, your first test confirms what I said above, you can't open a smart object, resize it and then add output to the original smart object which is a) unresized does not have the changes made by the layers. What you are essentially doing is applying both capture and output as the very first stage of your workflow. Ouch! You need to flatten your layers, resize, save and open up this new file in ACR to apply output sharpening. You can not do it with smart objects.

As for your print test, there are 3 versions of output sharpening, it could be that you need the higher level rather than the standard. As with PK, the level of output sharpening efficiency is in direct correlation to the capture sharpening accuracy. It could be that you need to tune how you use capture sharpening with the output in mind though if you have a method that works then stick with it! My method as Mark has been pointing out is far from simple or even efficient in comparison to PK. I just happen to prefer the control, speed and look of it.

So, if I understand correctly, you are saying that even if I capture sharpen in CR and then open it as a smart object in PS where I add adjustment layers, resize using image size etc. , when I double click on the smart object in PS to open it up again CR, what I am actually adding the output sharpening to is the original raw image and not the layered resized image in PS, and as a result, I am capture and output sharpening to the original raw file. The tif stage is what changes that scenario to capture sharpening a new altered image. Correct?

I do understand that CR output sharpening has three levels of intensity available. So, I will have to try out the others, especially the stronger, and see the results.
All my output so far is destined for an Epson 3800 printer at 360ppi. Can you please add some specifics to your statement: It could be that you need to tune how you use capture sharpening with the output in mind

Thanks again
Jonathan

Exactly correct.



Try them and see, what I meant was that you might have to do a slightly higher level of capture sharpening than you would have otherwise. Admittedly by that point, if you don't like the output sharpening, either buy PK or use your own method rather than faffing around with my method which is rather counter intuitive!

Ben
If I want to futz around with your method, what settings n the 'save as' dialogue for tif, under image compression would be recommended: LZW or ZIP and for layer compression: ZIP or RLE?
Thanks again
Jonathan

Honest truth? I have no idea whatsoever. Have to admit that I use 'none'!

Had a funny (in a crying kind of way) incident on Friday. Went into the top pro camera shop in the city to get a couple of prints done on their wide inkjet. Guy takes my disk then says that my files are so big it's going to take 15 mins to load them into PS. There were two 300 meg images on the disk! Open on my simple laptop in well under a minute. Eventually he comes downstairs and with a look of disgust asks me if I've seen my pics at 100% as they are the worst he's ever seen. I asked for my disk back and walked out. Someone who has never seen or heard of output sharpening and insists on valuing a file at 100% on a flat screen to determine print quality is not someone I'm going to trust with printing.

I'll go back to sending these files to my pro lab and getting the great results I'm used to...