Jeff Schewe - "Real World Image Sharpening" Workflow question

[Anthony Hutchinson]

This is my first post in the forum and it is actually an email I sent to Jeff Schewe after reading his book, "Real World Image Sharpening". The email actually directed me here so hopefully Jeff, if you're around, you will see this post!

I’m a black & white photographer from Toronto, Canada looking to bring my work to the next level through printing. As mentioned, I have just finished reading Jeff's book, “Real World Image Sharpening” and I have learned so much from it. For those who haven't read it, it deals with the three important phases of sharpening: capture sharpening (for source and content), creative sharpening and output sharpening. I am in the process of re-reading it so I can go through all of the examples and processes on my computer while I read them with the hope of gaining a greater grasp of all the concepts and processes. But I do have a question.

My question is in regards to how these sharpening processes fit into the overall workflow. In order to try and answer this question myself, I was able to get a hold of another of Jeff's books, “Adobe Photoshop CS5 for Photographers: The Ultimate Workshop.” The only section I have found so far that addresses this question is the following line on page 27 that states,

“… but we find that 90% of the photographs we take can mostly be improved by applying the following editing steps: lens corrections, crop, color, tone, finesse, capture sharpen, noise reduction and spotting.”

Output sharpening (or creative sharpening) isn’t mentioned but I assume that output sharpening at least would basically come at the end. This is interesting because after reading the “Image Sharpening” book, I would have assumed that the capture sharpening, including sharpening for source and for content would have come at the very beginning.

If you could bear with me I would like to outline my current workflow and I would be grateful for advice from those who apply Jeff's processes and techniques as to where each of the three sharpening processes fit in (particularly capture and creative).

RAW FILE IN LIGHTROOM:
1. Lens correction tab.
2. Spot removal
3. Then I work my way through the Basic tab adjusting WB, exposure, contrast, etc. (Global adjustments)

TRANSFER TO PHOTOSHOP as a Smart Object (as a TIFF file)

OPEN SMART OBJECT in Silver Efex Pro for conversion to black & white.
In Silver Efex, I would do more global adjustments but mainly focus on local adjustments

BACK TO PHOTOSHOP:
Mainly use Curves for further local fine tuning with masks.
“Creative” sharpening where needed

BACK TO LIGHTROOM:
Final Noise reduction, sharpening and cropping

I haven’t done any printing of my images as yet so the output is currently for web display.

Can anyone advice me if and where Jeff's sharpening techniques and processes would fit into this workflow. Based on the line I quoted above it would seem that capture sharpening would be done in the “BACK TO PHOTOSHOP” phase after all tonal adjustments but before any creative sharpening, although it seems to me that these two sharpening processes would cancel each other out.

Or do I need to modify my workflow to fit better with these capture and creative sharpening processes? Intuitively, I would prefer to do the Capture (including noise reduction) in Lightroom immediately before sending to Photoshop for B&W conversion and further tonal adjustments, in that way I am seeing and dealing with a sharp image. And then the creative sharpening could stay where it is. My fear with this though, is that further tonal adjustments including the B&W conversion may add more noise to the image.

Anyway, hopefully I could get some advice on this.

[Mark D Segal]

Anthony,

This may sound a bit radical, but based on processing thousands of photographs including numerous B&W, I think you should try radically simplifying your workflow by doing absolutely everything feasible and needed in Lightroom, and reserve Photoshop for those things that the occasional photos need and Lightroom isn't equipped to do. This way you go from ingestion to print without leaving the application, and the sharpening tools in Lightroom are fine - in fact designed in line with the Pixelgenius approach that you are using in Photoshop.

[Mark D Segal]

I should add in respect of the B&W work, while Silver Efex Pro has many great presets and some nice local adjustment features, much if not all you need for creative B&W conversions is also possible in Lightroom. I would highly recommend Martin Evening's Lightroom 5 book, and the LULA video tutorials on Lightroom.

[PeterAit]

I should add in respect of the B&W work, while Silver Efex Pro has many great presets and some nice local adjustment features, much if not all you need for creative B&W conversions is also possible in Lightroom. I would highly recommend Martin Evening's Lightroom 5 book, and the LULA video tutorials on Lightroom.

I agree with Mark - or another way to put it. KISS (keep it simple, stupid). The "stupid" was me a few years ago when I figured that the more programs I tried, the more papers I tried, the more workflows I tried, the better by prints would be. Nope. Get the smallest number of tools you can, good ones, and put in a lot of time with them. For me that was LR, Photoshop (needed back then because LR did not have some of the tools it does now), an Epson 4880 printer, and one kind of Epson luster paper. I worked for months until I could get prints I really liked. I asked questions on this forum until I swear they were about to ban me! Then and only then did I feel I was ready to start exploring the subtle differences that different papers and similar changes could provide.

If you were taking up the guitar, would you buy a rock guitar, a jazz guitar, a folk guitar, and a classical guitar at the start?

LR has everything you need to make truly excellent prints (but, only from excellent photos, which are up to you!). Capture sharpening is done automatically on import (unless you turn it off). Creative sharpening, the brush tool. Output sharpening, semi-automatic based on print size and output resolution. A lot of thought and skill went into LR, you might as well take advantage of it!

[sniper]

I'm sure Jeff will turn up with proper help but I'd suggest you check out his newer book the digital negative.
Regards Wayne

[Mac Mahon]

Anthony

I too work mostly in B&W — although I do print and seldom publish to the web.

I do my input sharpening (and noise reduction) as the first step in my Lightroom workflow.

And I leave output sharpening to Lightroom.  LR does a superb job of output sharpening, which it applies to the image being sent to export or print, and automatically sets parameters depending on resizing amount and output medium.

In his book "The Digital Print" Jeff Schewe points out that the creative sharpening step is the one that's weakest in Lightroom and ACR.  While you can locally increase (or decrease) sharpening using the adjustment brush you have no ability to adjust the sharpening parameters there.

If you need to apply creative sharpening beyond what's available in the adjustment brush then a trip to Photoshop or a plug-in like Topaz Detail is probably necessary.

I've not found curves adjustment in PS any more useful than curves in Lightroom - but that may be my lack of knowledge or skill.

You don't need to go to PS to use Siver Efex Pro:  it works as a plug in to LR. 

If you're on a Mac, I've found a new tool - Tonality Pro - has become my B&W convertor of choice:  it uses a layers-and-masks-type interface which gives, to my mind anyway, more control than Silver Efex.

Bottom line:  I'm with Mark and Peter.  Unless it's really necessary the trip to Photoshop adds complexity.

Cheers

Tim

[Tony Jay]

Just to add my voice to the chorus:
It should not be necessary to go into Ps routinely if one is using Lr as the basis for one's workflow.

You really do need to buy Jeff's excellent recent publications - The Digital Negative - and - The Digital Print.

There is plenty in there about sharpening - just predominantly using a Lightroom Paradigm.
BTW the Pixelgenius output sharpening algorithms are incorporated into Lightroom.

Tony Jay

[Anthony Hutchinson]

Many thanks for all the comments. I agree that I could simplify the workflow and that is something I will look at once I can figure out where the Capture sharpening fits in (as part of pre-tonal adjustments, post-tonal adjustments, pre-B&W conversion, post-B&W conversion?) While I am quite happy with Silver Efex, I could see myself moving to Lightroom for the conversion. I'm also very happy using PS for much of the post-conversion work because I tend to do a lot of B&W tonal adjustments using gradients within selections, something I can't do in Lightroom. As well, as has been mentioned by Tim, the creative sharpening really can't be done in Lightroom.

So once I get more advice on where exactly to place the capture sharpening/noise reduction within the workflow, I will definitely look at simplifying.

Thanks again.

[Anthony Hutchinson]

I would highly recommend Martin Evening's Lightroom 5 book, and the LULA video tutorials on Lightroom.

Thanks for that Mark. I actually finished reading Martin's book in its entirety a few months ago and I would also highly recommend it! I will have a look at the LULA videos. Before I go search for them, do you have a direct link?

[Mark D Segal]

I don't, but you'll find them easily in the Luminous-Landscape Shop.

[Torbjörn Tapani]

Anthony, I would rethink using gradients within selections. Selections in PS are 8-bit only. Let's say you have a selection that in itself is a gradient, it could potentially cause banding or posterization if you push the adjustment too much. To benefit from the full 16 bit (rather 15) use masks and channels where you would use selections if possible.

[Tony Jay]

...I agree that I could simplify the workflow and that is something I will look at once I can figure out where the Capture sharpening fits in (as part of pre-tonal adjustments, post-tonal adjustments, pre-B&W conversion, post-B&W conversion?)...As well, as has been mentioned by Tim, the creative sharpening really can't be done in Lightroom.

So once I get more advice on where exactly to place the capture sharpening/noise reduction within the workflow, I will definitely look at simplifying.

Again a small bit of insight in to the differences in how Lightroom works compared to (particularly early editions of) Photoshop.
Using a pixel editor means that any changes made are baked in immediately and so when to do things in the context of a full workflow becomes critical.
Lightroom, on the other hand, is a parametric editor where no edits are actually applied at a pixel level until an image is exported or otherwise instantiated, such as with a print out of Lightroom.
The advantages of this are legion:
1. One can go backwards and forwards through various edits over and over again making small or large changes without the process itself degrading the image. This is potentially disastrous in a pixel editor. However, in Lightroom only the net effect will actually be applied and that only on export etc. For editing purposes in the Develop module any edits are simulated so the effect of the edits can be easily visualised.

2. The order of editing becomes much less important. In general, the layout of the Develop module right sub-panel suggests an order of editing.
However, especially when it comes to things like applying lens profiles, sharpening, and noise reduction I like to apply these early (sometimes as presets on import) in order to evaluate an image for quality.

3. Altering things like global contrast, clarity (really a local mid-tone contrast adjustment), and luminance noise all affect how one might sharpen an image. Using Lightroom I can round-trip several times making small adjustments to each in turn to optimise image quality.

4. The edge masking slider in the detail sub-panel is worth the price of admission into Lightroom almost on its own. I take slight issue with those who say that regional sharpening cannot be done well in Lightroom. There will always be those images where layers and masks will be needed to precisely direct the effect of creative sharpening. Nonetheless, the edge masking slider is brilliant at directing the sharpening effect to areas that are important and away from areas that should not be sharpened. Used in conjunction with the radius slider amazing results can be achieved.

5. The luminance noise slider should be regarded as the fifth sharpening slider and adjustments here should be done in conjunction with other sharpening adjustments. If the edge masking slider has been used appropriately then it often minimises the amount of luminance noise reduction required. This in turn minimises the amount of sharpening required overall to restore detail in areas of high detail. Not requiring high levels of sharpening then, in turn, reduces the amount of that annoying haloing that occurs with aggressive sharpening.

6. Round-tripping to optimise image quality becomes a cinch. I have formalised the adjustments for certain lens/ISO/Camera combinations with regard to lens profiles, sharpening and noise reduction. Sometimes no further adjustment to noise and sharpening are required but even if there are they are normally minor.

7. Some third-party noise reduction plug-ins are available for Lightroom (as well as other editing utility).

8. Although the whole process is conceptually much simpler using a parametric editor doing this well will still need a learning curve that will probably take a year or more of effort. Another brilliant thing is that at any time one can revisit images edited early in the process and re-edit them using new-found knowledge and expertise born of the experience.

I bet I have not fully described the advantages of the Develop module but this will do for now!

Tony Jay

[Schewe]

BACK TO LIGHTROOM:
Final Noise reduction, sharpening and cropping

This is the biggest issue I have with your workflow...you should be doing noise reduction and capture sharpening in LR before taking it to Photoshop, or, in the very least, do the Noise reduction and sharpening in the ACR Smart Object.

Why? Well, once you have gamma encoded your image, the sharpening and noise reduction are no longer working on a linear image...it has a gamma based on your color space. LR's noise reduction and sharpening were designed to work optimally on linear raw images, and less optimal on gamma encoded images. Yes, the noise reduction and sharpening works with gamma encoded images, but anything other than a linear (gamma 1.0) image and the ACR/LR pipeline must deal with the gamma...and that can limit some of the capability of the processing.

The other break is doing creative sharpening in Photoshop before going back to LR for noise reduction and sharpening. That's a pretty bad break in the workflow...you don't really know how much creative sharpening is needed because you haven't done noise reduction and capture sharpening before going to Photoshop. I would fix that...

So, why crop back in LR instead of in LR in the first place? And since you are in Photoshop, why not there? If there was one thing that I think LR still limits the usability is the inability to crop zoomed in. However, you can crop in ACR in the Smart Object...not a big thing, but a small nit.

[Bart_van_der_Wolf]

So once I get more advice on where exactly to place the capture sharpening/noise reduction within the workflow, I will definitely look at simplifying.

Hi Anthony,

Capture sharpening and Noise reduction are typically some of the first things to do in a Raw conversion workflow. Lightroom has a drawback that it doesn't separate Capture sharpening from Creative sharpening in a better way, but you could always do Creative sharpening outside Lightroom if you want.

The usually proper sequence of events is; Capture sharpening (which is linked with noise reduction), Creative 'sharpening', Output sharpening.

There is an IMHO incorrect/confusing advice being promoted by some authors, that states that Capture sharpening is to be based on subject feature sizes in the image, it should not! Capture sharpening should instead be solely oriented at the hardware induced blur (=optical/mechanical) of the capture process, which is why you need to address it as one of the first things, before focusing (pun intended) on image content. That's why it is confusing for some to see when to do what, and in which order, until you remember; Capture sharpening is to correct Capture process induced blur (as if to improve your 'focusing').

Cheers,
Bart

[Anthony Hutchinson]

This is the biggest issue I have with your workflow...you should be doing noise reduction and capture sharpening in LR before taking it to Photoshop, or, in the very least, do the Noise reduction and sharpening in the ACR Smart Object.

Thanks for your response Jeff. This is why I read your book! After reading Martin's Lightroom book I knew I was handling it wrong.

Now if I understand this correctly, I'm basically left with two options (bearing in mind that I prefer to get the color image "right" before converting to B&W). Either,

Do all tonal adjustments, the B&W conversion, and the sharpening and noise reduction in LR, and only sending to PS if I have further creative adjustments including gradient masks and creative sharpening,

OR,

Do all color tonal adjustments in LR, then the B&W conversion in SE, then back to PS for gradient masks, then capture sharpening followed by creative sharpening while still in PS.

Do either of these options sound about right?

Again, ultimately this is all being done in order to eventually maximize the quality of the final print.

(And sorry guys. I know some of you have suggested I do everything in LR, but gradient masks are my bread and butter).

[texshooter]

There is an IMHO incorrect/confusing advice being promoted by some authors, that states that Capture sharpening is to be based on subject feature sizes in the image, it should not! Capture sharpening should instead be solely oriented at the hardware induced blur (=optical/mechanical) of the capture process, which is why you need to address it as one of the first things, before focusing (pun intended) on image content. That's why it is confusing for some to see when to do what, and in which order, until you remember; Capture sharpening is to correct Capture process induced blur (as if to improve your 'focusing').

You hit the nail on the head. I wish someone told me this years ago. I can't tell you how many hours I wasted watching Youtube videos about sharpening in Camera Raw all the while believing that I was learning true "capture sharpening." Juggling the Amount, Radius, Detail and Masking sliders was so confusing to me. I felt like one of those acrobats spinning chinaware on sticks.  If I had from the outset treated capture sharpening as strictly a means to repair "hardware induced blur," I would have saved myself frustration over those sliders. I now treat the sharpening tools in Camera Raw as "all purpose" sharpening rather than as specialized capture sharpening.

[john beardsworth]

Do all tonal adjustments, the B&W conversion, and the sharpening and noise reduction in LR, and only sending to PS if I have further creative adjustments including gradient masks and creative sharpening,
OR,
Do all color tonal adjustments in LR, then the B&W conversion in SE, then back to PS for gradient masks, then capture sharpening followed by creative sharpening while still in PS.

Don't overlook some of the strengths of your LR>PSSmartObject>SilverEfex workflow. Applying your SilverEfex edits to a PS smart object means you can fine tune ACR / SilverEfex adjustments in subsequent sessions, even years later. The LR>SilverEfex route bakes those edits into a flat TIF file - fine tuning your adjustments means starting all over again.

I agree strongly that you can do excellent B&W in LR alone, and with other comments too, but you can think of SilverEfex fitting into your practice as if it's a different guitar. Play the same tune on a country guitar and you'll perform it differently from the one you'd deliver on a Rickenbacker (sometimes better, sometimes not).

So maybe you can greatly increase the proportion of B&Ws you do solely in LR. But when you do go to SilverEfex, follow that smart object route.

[tommm]

Bart has it right.

Capture sharpening should be purely to correct for hardware blurring, and once you have appropriate settings for a specific camera / lens combo can be applied to all images with that camera / lens without further adjustment for subject.

This should be one of your first steps in Lightroom (though as it's not a pixel editor it's not critical exactly when you do it, so long as it's before exporting to photoshop or other).

Creative / content sharpening is the next step in the sharpening workflow, unfortunately this is very limited in Lightroom, limited solely to using a brush to apply more or less of the same sharpening settings used in your capture sharpening. For this reason I would suggest doing creative sharpening in Photoshop (as you're going there anyway). Do this after you've done everything else you want to do in photoshop before returning to Lightroom.

Output sharpening, as you rightly said, is only applied when creating some sort of output to correct for the softening that happens when changing viewing size and media (most dramatically in prints). Lightroom applies some very sophisticated automatic sharpening to your image when using its output sharpening but unfortunately allows very limited control over it. I find I have to slightly over sharpen to my taste the previous sharpening in order for Lightrooms output sharpening to produce prints sharp to my taste (hope that makes sense - basically I find the output sharpening isn't strong enough).

In an ideal world:

Capture sharpening corrects for hardwear blurring bringing an image back into proper focus.

Creative sharpening emphasises certain elements and de-emphasises others by selective sharpening to get the image to look as good as you could hope on a good screen.

Output sharpening compensates for the output process to make, for example, a print look as sharp as the image you see on screen.

Hope that helps,

Tom

[damoison]

Dear Jeff Schewe,
I live in Paris, my english is not so good. And new to The Luminous Landscape
I read your book "Real World Image Sharpening" and I am working to understand it.
I know already your work will help me improve the quality of my photography.

One thing escape my mind, and I would like to have enlightment from you.
It is about the relationship between  Displays viewing zoom percent and printed result.
I understand your explanation to calculate our true displays resolutions:
"Using a tape measure, measure the width, in inches, of the image area on your display.
Divide the number of horizontal pixels that your display can show by the number of inches from Step 1.
The resulting number is the true resolution of your display in pixels per inch…"

- I have a display with  : 24", it shows: 1920 x 1200 pixels, with your calculation I obtain its true resolution at 92,80

It's the second part, the calculation you have made to obtain the more accurate Zoom percentage, that I don't understand.
"…If your display is around 100 ppi and you’re printing at around 240 ppi, viewing the image at 50% will give you a truer picture of the final sharpening. Similarly, if your display is around 75 ppi and you’re printing at 300 ppi, the 25% view will give you a closer idea of final sharpness than any of the higher zoom percentages…."

With my 92,80 true display pixels, how do I calculate the zoom percentage, to have the best viewing distance if I print 300 pip, 360, 150, etc.

Thankyou very much

[Bart_van_der_Wolf]

With my 92,80 true display pixels, how do I calculate the zoom percentage, to have the best viewing distance if I print 300 pip, 360, 150, etc.

Hi,

Welcome to Luminous Landscape. I don't know if Jeff is around, but it's actually a quite simple calculation.

Zoom = Display PPI / Output PPI

However, since Photoshop uses a rather crude display resampling (and fixed steps), the resulting zoomed display image will only roughly approximate the real output when viewed from the same distance.

Viewing distance changes things proportionally, if you view the print at 2x the distance that you view your display, you divide the zoom percentage by 2, because more distant things look smaller.

So then the complete formula becomes:
Zoom = (Display PPI / Output PPI) x (Display viewing distance / Print viewing distance)

Cheers,
Bart