Image Sharpening Article (Sorry a bit of a Rant)

[TheBigE]

Well i have to admit that in reading this article I was a bit taken back by the general simplicity and premise overall.

To be honest, for whatever reason, it has kinda been stuck in my gut. I read it earlier this week and went back to review and frankly could not get it out of my mind. Perhaps my thoughts below will be sufficient to allow me to move forward on the topic.

I know, it is just another article on the internet and it is our responsibly to read with care. It is published on s tie with a fairly large following. Regardless, I feel that the topic is glossed over and really not discussed. Much of this can come from the "fast" approach to many things in life.

I am still a newbie in relative terms and generally do not feel that I have the pedigree to challenge many assumptions. However, in this case to have an article on Photoshop Sharpening and not really address or acknowledge the work of Bruce Fraser and Jeff Schewe in Real World Image Sharpening with Adobe Photoshop, Camera Raw and Lightroom is truly a disservice.

The article does not even talk about why sharpening is needed, nor make any reference to the three areas of sharpening - Capture, Creative, and Output. What value does it add?

I have contemplated replying to the article or even the post on FB that announced the article, but not sure of any value in that matter. The whole article just plays too fast and loose about sharpening and akin to telling someone to take a great photo set your aperture to f/8, ISO 200, and shutter speed 1/600s in the sun. There is no understanding or concept behind it....

Thanks for listening and I guess i need to spend more time behind the viewfinder and less time reading the internet....

[Drops the Mic]

[Shadowman]

I only scanned the article but it looks the author was only demonstrating how each of the programs can be used. Perhaps the author didn't quote any experts to eliminate the need for citing a reference; although technically every Adobe example has been written many times over. I think one statement should be in every sharpening tutorial; and Uwe mentions it in his text, "there are many other valid sharpening strategies that will suit your needs (p.148)" or stated another way "everyone has their own tastes as to what is sharp".

[DanK]

I just skimmed it. Personally, I don't see anything wrong with it. It's just a simple explanation of 4 ways to sharpen using the Adobe Suite. There is a lot omitted, even beyond what you mention; for example, the author doesn't describe how the results of the four methods differ or what attributes of the image might lead you to choose one over the other. But IMHO, that's fine; that's beyond what the author was trying to do.

This is another topic, but re capture, creative, and output sharpening: I personally don't find that three-way distinction very helpful. What I think is essential is the difference between sharpening before outputting and output sharpening. Yes, I know that the first is supposed to compensate for softening in the capture--e.g., from an AA filter--but in many cases, the distinction has no practical significance. Both "capture" and "creative" sharpening are just an effort to make the image appear as you want on the editing screen, and if you use Adobe's parametric editing--Lightroom or ACR--it in fact makes no difference whatever when you do that sharpening or how many sharpening steps you make. For example, you can zero out the default capture sharpening that LR applies at import and simply add that much more sharpening at the end. It won't matter. The end result is not affected by the sequence in which you do the edits. Of course, if you are working in a pixel editor, that isn't the case; then the adjustments are applied in the order you perform them. This is why people using pixel editors generally reduce noise (if they are going to at all) before sharpening--so that they don't sharpen the noise. In contrast, with LR and ACR, it makes no difference which order you use. There again, however, I personally don't find it helpful to classify sharpening as capture or creative.

Output sharpening is a whole different beast. What you will want depends on things separate from the image on the editing screen--the output medium (even the type of paper) and the output size.

[Mike Buckley]

Agreed with everything in Dan's post. I'll add that the three-way concept that includes the first two parts of sharpening (capture and creative sharpening) is now very different for anyone using one of the more recent camera models that have no anti-aliasing filter.

[Manfred M]

Eric - I've had a quick read of the article as well and can't find any real flaws in what the author has written. One can only say so much in an article and it should definitely not be viewed as anything else. Again, when you point to Fraser and Schewe work, they wrote a whole book on this subject, so their treatment of the subject is much more comprehensive than this article.

Let me add my own view on the part that you suggest is missing; the reason for the three stage sharpening recommended by Fraser and Schewe.

1. Import sharpening - this stage is all about correcting the inherent softness associated with the digital image capture process. There are two main components; Mike has mentioned one of these - the blurring brought on through the use of an anti-aliasing filter. More modern cameras with their high resolution sensors means that these filters are much weaker than they have been in the past or even non-existent in some camera models. This means less sharpening is required than with older and lower resolution cameras.

There is a second source of capture softening and that is related to the use of the Bayer array that is used in virtually all digital cameras (the Foveon sensor used by Sigma does not use a Bayer array). Raw convertors take the RGGB data from from a number adjacent photo diodes of and turn that into a single RGB colour value. This capture softening effect will be there even in cameras with no AA filter and needs to be taken care of at image import.

Import sharpening is "global" in nature, i.e. it affects the entire image. It should be evaluated at 100% image size.


2. In-process sharpening is local sharpening done to specific parts of an image. Areas of fine detail like eyebrows or eye lashes in a closeup portrait or vegetation or trees in a landscape will often benefit from local sharpening as well. Using the landscape example - if we sharpen the entire scene, including the sky and water will often look strange and harsh, so locally sharpening other parts of the image gives superior results. This is probably my most common application of in-process sharpening.

I tend to use the same tools for in-process sharpening as I use for import (and output) sharpening, but apply it locally through through a layer mask, in Photoshop.

3. Output sharpening - This is a step done when an image is resized to its final size. This is something I only do for when I do prints, as anything displayed on a screen is on an image I post to the internet will be downsized and I have no control over the screen the image will be displayed on.

When I print, often I will upsize the image (which adds softness) and the different papers all react a bit differently. A matte paper tends to have a bit more bleed than a glossy paper and the reflected light is more diffused, so will require more sharpening than a glossier surface. This type of sharpening is done at 100% of the final size and after resizing the image.

[Mike Buckley]

If we're gonna get into the specifics, which the article by design does not do...

Import sharpening is "global" in nature, i.e. it affects the entire image.

Not the way I do it. As just one example, if a plain blue sky is sharpened, there is a good chance that that area of the image will look bad when displayed on a large television. That explains why I never apply any sharpening to that area of the image. Similarly, if large portions of the image are out of focus by design, I never sharpen those areas. So, my first step of sharpening is the standard import sharpening but applied only locally if necessary before moving on to the creative stage of sharpening.

It should be evaluated at 100% image size.

I really hate that word, should, when it's applied to post-processing. I've very rarely sharpened at 100% because doing so very rarely works for me. I remember years ago Moose Peterson explaining that all of his sharpening was being done at the time at 600%. I rarely have to check my sharpening at high magnification because I've gotten so used to the results of my sharpening parameters. When I do check at a higher magnification, it's almost always at 67%. I wonder if my method would change if I bought a camera that produces larger files. All I know is that whatever works for someone is the only magnification that "should" be used by that person.

Output sharpening...is something I only do for when I do prints, as anything displayed on a screen is on an image I post to the internet will be downsized and I have no control over the screen the image will be displayed on.

My solution to that dilemma is to create an image file that is the largest that can be viewed in totality on my monitor and to sharpen to display it there attractively. At least I know the image looks good for anyone using the monitor resolution that I use and that it looks good on my monitor.

[xpatUSA]

Well i have to admit that in reading this article I was a bit taken back by the general simplicity and premise overall.

I went to the link and, as is now the mode, up pops a "subscribe to our site" dialog which blots out the home page. So, the hell with that, eh?

So I didn't read the article. If anyone would like a sane article on the subject, may I recommend:

http://www.falklumo.com/lumolabs/art...ess/index.html
.

[MrB]

I went to the link and, as is now the mode, up pops a "subscribe to our site" dialog which blots out the home page. So, the hell with that, eh?

So I didn't read the article....

One click closes the pop-up and you can read the article, if you want to.

[Manfred M]

Mike - I personally don't care one way or other how you sharpen and why you use the workflow that you do.

I was commenting on the Fraser & Schewe three step approach. I did say "Let me add my own view on the part that you suggest is missing; the reason for the three stage sharpening recommended by Fraser and Schewe." and should be read in that light.

1. I don't know what software you are currently using, but a number of products, including Lightroom and Adobe Camera Raw automatically do import sharpening, unless you disable this functionality. This is something I have done as I want 100% control to do import sharpening (or not). I usually do a touch, but not always.

2. The 100% size is because there is no "inherent" sharpening or softening of the image by the PP software resizing algorithms when it is viewed at 100%. That is the technical reason behind that rationale.

3. While I work my images at full size (usually 36MP), I tend to display them at 1600 or 2000 pixels on the long side. The downsampling algorithms do tend to inherently sharpen the images by maintaining the edge contrast. I have found that I have had little complaint about image sharpness, so I assume this is working for me. I do check them, just to make sure and at times I have gone back and softened them if they look too hard. This seems to be more of an issue with high contrast images.

Prints on the other hand tend to up upsampled (my printer's native resolution is 360 ppi). I tend to only use a couple of papers, so I have a good feel for the amount of sharpening I need, based on experience.

[Mike Buckley]

Manfred,

I was purely adding to the discussion. Not too long ago Dave publicly invited me to "move on" to another website. Now you have added the terse comment that you personally don't care one way or the other how I sharpen or why I use the workflow that I do. All of that in my mind is completely inappropriate, especially considering that both of you are moderators. Very, very disappointing. Not at all inviting or in keeping with the company's tag line that this is a learning community.

This is my last post in this thread.

[Manfred M]

Manfred,

I was purely adding to the discussion. Not too long ago Dave publicly invited me to "move on" to another website. Now you have added the terse comment that you personally don't care one way or the other how I sharpen or why I use the workflow that I do. All of that in my mind is completely inappropriate, especially considering that both of you are moderators. Very, very disappointing. Not at all inviting or in keeping with the company's tag line that this is a learning community.

This is my last post in this thread.

Sorry Mike if you read things into my response this way, it was not meant to be read this way. Your sharpening methodology is what you use, but being a "learning community", I was specifically responding to Eric's point":

The article does not even talk about why sharpening is needed, nor make any reference to the three areas of sharpening - Capture, Creative, and Output. What value does it add?

I felt that your response added confusion to what I had written and felt I needed to clarify.

[Mike Buckley]

I'll break my promise not to post in this thread again to thank you, Manfred, for the apology. Accepted. Enough said!

[xpatUSA]

I went to the link and, as is now the mode, up pops a "subscribe to our site" dialog which blots out the home page. So, the hell with that, eh?

So I didn't read the article....

One click closes the pop-up and you can read the article, if you want to.

I knew that. This modern trend irritates me beyond compare, to be honest.

[TheBigE]

Thanks everyone - a bit of level loading on my part - I would say a touch of an overreaction....it happens. Really the article is about the very basics and how to proceed in PS. Looking back now with a few days in between nothing all that disturbing...

I will blame the fact that I read the article in the middle of doing my taxes for the emotional response.....not good risk management on my part.

[Mike Buckley]

I will blame the fact that I read the article in the middle of doing my taxes for the emotional response.....not good risk management on my part.

Consider using my method of risk management: let someone else do your taxes. However, to be completely forthcoming, my wife wouldn't let me do our taxes if I wanted to do them. She does them because she wouldn't trust me to do them correctly.

[TheBigE]

Consider using my method of risk management: let someone else do your taxes. However, to be completely forthcoming, my wife wouldn't let me do our taxes if I wanted to do them. She does them because she wouldn't trust me to do them correctly.

Well for full disclosure I actually do not do my taxes, as an EXPAT a company does them for me as I have to file both German and US taxes. That being said, I still need to gather all the information and put them into the "organizer"....which includes updating a calendar with my physical location for every day of the year.....but then I digress.