Print Sharpening in Lightroom

[Mike Buckley]

There are good reasons why Adobe has not chosen to include a capture sharpening filter in Photoshop is (1) it would have too many parameters and would be too complicated for the average user (2) it is already incorporated in Lightroom.

Another reason might be that capture sharpening is available in Adobe Camera Raw, which is a plug-in to Photoshop.

[jisner]

Another reason might be that capture sharpening is available in Adobe Camera Raw, which is a plug-in to Photoshop.

Of course! I should have thought of that. I think ACR and Lightroom capture sharpening ("Detail") evolved somewhat independently, but by now they're probably identical.

[dje]

If I were to speculate as to how the print sharpening algorithm works, it would go something like this: Look for existing halos (oops, I need a better term for "tonal gradients across edges") and widen them just enough that they'll subtend one minute of arc when viewed at the standard viewing distance.

John I doubt that it works like that. After all, the software has no way of knowing just how sharp any particular edge should be.

Generally, most sharpening filters involve mathematically convolving the image data matrix with a kernel either globally or selectively if some form of masking has been used. The sharpening method determines the values used in the kernel (which might be say a 5x5 matrix). This has the effect of increasing contrast either side of the edge. The bottom line is however that the user must decide the amount of sharpening applied. In the case of the LR print module, the amount is determined by the paper type, choice of sharpening level and other factors determined by LR based on resolution and re-sizing etc.

By the way, one technical term for the "shape' of an edge is the "Edge Spread Function".

Dave

[pnodrog]

There seems to be an assumption that because it is a parametric application of the adjustments that the order is irrelevant. However I have always understood these parametric adjustments are done prior to the image having the conversions (DPI/PPI, colour space etc) made for output or printing and that the print/output sharpening would be made after this scaling etc was done. If this is the case and results tend to indicate it is then sharpening should be treated as at least a two step process even with lightroom.

I thought I had sharpening reasonably sussed until I jumped from a 10 megapixel camera to a 36 megapixel camera and had to completely review my default sharpening settings. Two years ago one of my photographs was used for a 1.5 meter by 4 meter billboard and not only did I have to once again review the sharpening but also had to significantly reduce the clarity settings because of the visible artifacts it produced on such a large scale print.

[DanK]

John I doubt that it works like that. After all, the software has no way of knowing just how sharp any particular edge should be.

Exactly. As I understand it, print sharpening is designed to offset the change in sharpness when moving the image to a printer, not to reach some 'ideal' level of sharpness.

There seems to be an assumption that because it is a parametric application of the adjustments that the order is irrelevant. However I have always understood these parametric adjustments are done prior to the image having the conversions (DPI/PPI, colour space etc) made for output or printing and that the print/output sharpening would be made after this scaling etc was done. If this is the case and results tend to indicate it is then sharpening should be treated as at least a two step process even with lightroom.

Yes, the OUTPUT sharpening is distinct in Lightroom. It can't be accessed from the Develop module, and it is applied to the final result of the edits. Capture and creative (if global) are not distinct, and the order doesn't matter. I can't locate the earlier threads about this. However, if I recall, there is documentation somewhere that the order of edits imposed by the user does not affect the order of the processes undertaken by the software to create the image. I also did a test of this that I posted. I created one edited image and exported to jpeg, then went back to the base image and applied the edits in the reverse order. Made no visible difference at all.

[jisner]

Thanks to everyone for all your replies. Getting back to my original problem, I need to explain in my document why capture sharpening is essential if print sharpening is to work as expected. The stated goal of capture sharpening is to get the image to look sharp at 100%. So if you forgot to do capture sharpening and you realize it when you're about to print, why not just return to the Develop module, sharpen, then return to the Print module and print? I'm sure you can do that when you're printing a raw file (because raw adjustments are parametric), but what if the file has gone to Photoshop and come back as a rasterized image? In that case, would sharpening in the Detail panel no longer qualify as capture sharpening?

[Mike Buckley]

what if the file has gone to Photoshop and come back as a rasterized image? In that case, would sharpening in the Detail panel no longer qualify as capture sharpening?

My question has to do with the audience intended to read your document: Considering that anyone who is so advanced in their post-processing skills that they make their adjustments in Lightroom and then make other adjustments in Photoshop that can't be made in Lightroom, why would you even take into account the situation when that person makes such a rookie mistake as to forget to use capture sharpening?

[george013]

Thanks to everyone for all your replies. Getting back to my original problem, I need to explain in my document why capture sharpening is essential if print sharpening is to work as expected. The stated goal of capture sharpening is to get the image to look sharp at 100%. So if you forgot to do capture sharpening and you realize it when you're about to print, why not just return to the Develop module, sharpen, then return to the Print module and print? I'm sure you can do that when you're printing a raw file (because raw adjustments are parametric), but what if the file has gone to Photoshop and come back as a rasterized image? In that case, would sharpening in the Detail panel no longer qualify as capture sharpening?

As soon it is visible in the converter it's a RGB raster image. That's the only format that's visible. All editing is done on that. What's coming up is more a question of the sequence.
Why do you need capture sharpening when you're satisfied?

George

[jisner]

As soon it is visible in the converter it's a RGB raster image. That's the only format that's visible. All editing is done on that. What's coming up is more a question of the sequence. Why do you need capture sharpening when you're satisfied?

The problem is that a lot of people skip capture sharpening. They don't inspect their image at 100% so they don't realize how soft their image is until they print it. My printing instructions warn that print sharpening requires that your image already be sharp when viewed at 100%. So for the person who hasn't sharpened and reaches this step, what are they supposed to do? I'm tempted to say "just go back to Develop and do your sharpening and noise reduction, then come back to print. And I wouldn't make a distinction between raw or raster files. That makes it sound incredibly easy. But is it correct? I suspect not.

[george013]

The problem is that a lot of people skip capture sharpening. They don't inspect their image at 100% so they don't realize how soft their image is until they print it. My printing instructions warn that print sharpening requires that your image already be sharp when viewed at 100%. So for the person who hasn't sharpened and reaches this step, what are they supposed to do? I'm tempted to say "just go back to Develop and do your sharpening and noise reduction, then come back to print. And I wouldn't make a distinction between raw or raster files. That makes it sound incredibly easy. But is it correct? I suspect not.

I'm not an editor, i just shoot. CaptureNx is enough for me.
I always learned that sharpening has to be done in the last place. If that's in Photoshop, so be it. I still don't know what is the essential difference between sharpening in Lightroom or Photoshop. Maybe you've some presets in Lightroom, that sells better. But what is the essential difference? And why first sharpening, making artificial lines first and then mixing them up again in further editing?

I can't help you further. But I keep reading.


George

[dje]

The problem is that a lot of people skip capture sharpening. They don't inspect their image at 100% so they don't realize how soft their image is until they print it. My printing instructions warn that print sharpening requires that your image already be sharp when viewed at 100%. So for the person who hasn't sharpened and reaches this step, what are they supposed to do? I'm tempted to say "just go back to Develop and do your sharpening and noise reduction, then come back to print. And I wouldn't make a distinction between raw or raster files. That makes it sound incredibly easy. But is it correct? I suspect not.

John I don't see a problem with that. Sharpening is sharpening, whether you have the word "capture" in front of it or not. Sometimes I use Topaz Detail for capture sharpening (as a plug-in for Photoshop). This is after raw conversion in ACR.

Dave

[jisner]

John I don't see a problem with that. Sharpening is sharpening, whether you have the word "capture" in front of it or not. Sometimes I use Topaz Detail for capture sharpening (as a plug-in for Photoshop). This is after raw conversion in ACR.

Dave

OK, I'm going to go with the consensus here, which seems to be that before you print, your image should be reasonably sharp when viewed at 100%, and it doesn't matter how you do it. You can sharpen in the Lightroom Detail panel, or in Photoshop, or in a plugin. And it doesn't matter if you do it on a raw file or a rasterized file.

The unique thing about ACR/Lightroom capture sharpening (in my opinion) is how it is paired with noise reduction. That makes it convenient. But maybe "noise reduction is noise reduction" so you can ignore the Detail panel in ACR/Lightroom altogether and do noise reduction and sharpening anywhere you like (just do noise reduction before sharpening).

[jisner]

Considering that anyone who is so advanced in their post-processing skills that they make their adjustments in Lightroom and then make other adjustments in Photoshop that can't be made in Lightroom, why would you even take into account the situation when that person makes such a rookie mistake as to forget to use capture sharpening?

It's a school, so everyone's a rookie!

[dje]

The unique thing about ACR/Lightroom capture sharpening (in my opinion) is how it is paired with noise reduction. That makes it convenient.

Yes I like that aspect of it too.

Dave

[DanK]

OK, I'm going to go with the consensus here, which seems to be that before you print, your image should be reasonably sharp when viewed at 100%, and it doesn't matter how you do it. You can sharpen in the Lightroom Detail panel, or in Photoshop, or in a plugin. And it doesn't matter if you do it on a raw file or a rasterized file.

The unique thing about ACR/Lightroom capture sharpening (in my opinion) is how it is paired with noise reduction. That makes it convenient. But maybe "reduction is noise reduction" so you can ignore the Detail panel in ACR/Lightroom altogether and do noise reduction and sharpening anywhere you like (just do noise reduction before sharpening).

I think you can simplify this. The bottom line, IMHO, is simply this:

Make sure the image is as a sharp as you want it to be before you deal with output sharpening. The purpose of output sharpening is to maintain that level of sharpness, whatever it happens to be, when you move the image to an output device.

You want it not to be too sharp, and you prefer it without any capture sharpening? That's fine. You like it even less sharp, and you apply negative clarity in LR or even Gaussian blur? That's fine too. Just make sure it looks the way you want to look before you deal with printing.

Re the order: it matters in pixel editors. If you do sharpening first and noise reduction last, you end up sharpening noise. It doesn't matter in LR because the order in which it performs calculations is not determine by the order in which you make the edits. I frequently move up and down the development panel, revising one edit based on another. It makes zero difference. That's one reason I do my basic editing in LR.

Does it matter how you do it? Yes, it can. LR has a fairly powerful sharpening function, but only one. Often, it is fine for me, but not always. Photoshop has several, and they sometimes can differ quite a bit. Sometimes I will create multiple sharpening layers at the top of the stack--e.g., one with a high-pass filter, one with smart sharpen, maybe even one with USM--and toggle them off and on to see which I like best.

None of this, however, has any necessary relationship to output sharpening. When I have what I like--using any of those sharpening methods--I put the image back into LR and print from there, using the standard output sharpening. It seems to do a good job of maintaining the appearance of sharpness I have ended up with, regardless of how I got there.

[jisner]

I think you can simplify this. The bottom line, IMHO, is simply this:

Make sure the image is as a sharp as you want it to be before you deal with output sharpening. The purpose of output sharpening is to maintain that level of sharpness, whatever it happens to be, when you move the image to an output device.

Thanks, DanK. I'll take that as the definitive answer!

[Mike Buckley]

Thanks, DanK. I'll take that as the definitive answer!

Excellent choice!

[george013]

....
Re the order: it matters in pixel editors. If you do sharpening first and noise reduction last, you end up sharpening noise. It doesn't matter in LR because the order in which it performs calculations is not determine by the order in which you make the edits. I frequently move up and down the development panel, revising one edit based on another. It makes zero difference. That's one reason I do my basic editing in LR.
......

They're all pixel editors. I'll be very surprised if sequence doesn't matter. If it's visible is another matter. That depends on the tools and amount you use.
As far as I can think of it's pixel vs vector and parametric vs non parametric.
About the parametric editors, the question is how they save their final result: as a new image or as a list of commands to be done on a base image.
Another way to put it simply: can I stop editing and continue the next day with exactly the same screen with the done editing as I stopped the day before. Or a year.

George

[DanK]

They're all pixel editors. I'll be very surprised if sequence doesn't matter. If it's visible is another matter. That depends on the tools and amount you use.
As far as I can think of it's pixel vs vector and parametric vs non parametric.
About the parametric editors, the question is how they save their final result: as a new image or as a list of commands to be done on a base image.
Another way to put it simply: can I stop editing and continue the next day with exactly the same screen with the done editing as I stopped the day before. Or a year.

George

George,

Have you ever used Lightroom?

It is not a pixel editor. It is a parametric editor. It stores a history of your edits (in an XML sidecar file if you choose) and the raw file, nothing else. It does not create another file until you ask for one, for example, by exporting a jpeg. It renders the edited raw file for display while you edit, of course, but it does not create and store a file.

Yes, you can come back to the same image at any time--if and only if your edits have been saved, either in the catalog file or in a separate sidecar file. For a number of reasons, I use the latter, and if I were to lose the tiny XML file, all of my edits would vanish, because then all LR would have would be the raw file.

From what I have read (and tried), order mostly doesn't matter because the software determines, based on the total set of edits present at any stage, the order in which to perform the operations. The main exception lies in the phrase "total set of edits." Some edits are more math-intensive than others (e.g., the lens profile corrections), and some people advise leaving them out of the stack of edits until the end, not because that affects the final result--it doesn't--but because having it in the stack can slow down editing if you are using a slow computer. On my computer, it makes no appreciable difference, and I move up and down the editing panel freely, as I see the need for additional edits.

As I noted earlier, this was discussed in an earlier thread. I believe that thread included both documentation of this and my test images. However, I haven't been able to find that thread.

Dan

[george013]

George,

Have you ever used Lightroom?

It is not a pixel editor. It is a parametric editor. It stores a history of your edits (in an XML sidecar file if you choose) and the raw file, nothing else. It does not create another file until you ask for one, for example, by exporting a jpeg. It renders the edited raw file for display while you edit, of course, but it does not create and store a file.

Yes, you can come back to the same image at any time--if and only if your edits have been saved, either in the catalog file or in a separate sidecar file. For a number of reasons, I use the latter, and if I were to lose the tiny XML file, all of my edits would vanish, because then all LR would have would be the raw file.

From what I have read (and tried), order mostly doesn't matter because the software determines, based on the total set of edits present at any stage, the order in which to perform the operations. The main exception lies in the phrase "total set of edits." Some edits are more math-intensive than others (e.g., the lens profile corrections), and some people advise leaving them out of the stack of edits until the end, not because that affects the final result--it doesn't--but because having it in the stack can slow down editing if you are using a slow computer. On my computer, it makes no appreciable difference, and I move up and down the editing panel freely, as I see the need for additional edits.

As I noted earlier, this was discussed in an earlier thread. I believe that thread included both documentation of this and my test images. However, I haven't been able to find that thread.

Dan

You didn't understand me. It's pixel based vs vector based. Lightroom is pixel based. It changes the pixels in that RGB rasterfile that I draw in my diagram. That file where all the editing is done at.

Another division is how the results are stored/saved. A parametric editor stores the result as a list of commands above the base image. It's irrelevant if this is done in a sidecar file or as part of the basic file like CaptureNx2. A non parametric editor stores the result as a new image. That RGB rasterfile in my diagram is saved.

So Lightroom is beside a raw converter also a pixel based parametric editor. Or contains one, however you want to see it.

Say it in some other way.
Pixel or vector based deals with how the image is edited.
Parametric or non parametric how that result is saved.

George