Bulk Sharpening question in Lightroom

jfriend

First off I should say that I realize sharpening can be a very advanced topic and that every image is technically different and can be optimized with different sharpening techniques. If one is producing fine art photographs for sale, that is certainly the philosophy to follow and I own several sharpening books that cover that level of detail.

But ... I'd like to ask for folks ideas on something simpler. I shoot a lot of events (softball games, soccer games, track meets, school events). It is not uncommon for me to publish (in web galleries) 100-300 photos per week. I am not going to be individually sharpening every single shot. Instead, I need to develop some "rules of thumbs" for how to do appropriate bulk sharpening in Lightroom so, right before I export to JPEG for upload to the web, I can apply whatever sharpening on the whole shoot. Since I know there are lots of other event shooters out there that are not going to tweak sharpening on each individual shot before putting them up into Smugmug, I'd love to hear what you all do for sharpening settings in Lightroom.

I should add that the photos I put on the web are uploaded as full res originals. Viewers will do one of three things with these images. 1) View the reduced web-sized versions of the images on the web or 2) Download the high res originals to either make their own prints or just add to their own digital album or 3) Order prints via Smugmug (which will be made from the high res original). So, because the images I'm preparing are multi-purpose, the sharpening has to be too (which necessarily means it's a compromise).

Here's where I get confused in Lightroom. Lightroom has the Sharpening controls in the Develop module (which default to amount=25, radius=1.0, detail=25, masking=0). It also has sharpening in the Export dialog called "Output Sharpening" and has options to sharpen for "screen" or "matte paper" or "glossy paper" and has amounts of "low", "standard" and "high". So, I'm trying to figure out what are some intelligent values to use for the typical sports shooting I do (sample images below). I shoot RAW with the Nikon D300 and either 70-200 or 200-400. I'm usually shooting at f/2.8 or f/4 for subject isolation. For workflow reasons, I'm looking for something that can be done purely via Lightroom.

What should I set the sharpening controls in the Develop module to?
What should I set the export sharpening controls to?

I've been getting OK results with 75,1.0,25,0 in develop and "Screen"/"Low" in export, but I feel like this is just a wild guess and there are probably better ways of doing this or figuring out what would work best. Any thoughts on bulk sharpening for these types of images would be appreciated.

Here are a couple sample images to see what kinds of images I'm talking about. If there are any other events/sports shooters who can share you normal sharpening settings in Lightroom, I'd appreciate it.

Shishu

I'm in the same boat with family snapshots. I really don't have time to treat them individually, but with the amount of time I have spent trying to find a one-size-fits-all sharpening scheme, maybe I should have .

What I do now (which will not answer your question), is to do USM in Photoshop at 500/1/0 on an inverted red channel. It's not ideal, but brings out good fine detail, and keeps sharpening away from skin.

I would be *really* interested to know how other folks batch sharpen.

Tiziano

Maybe you said this in your post (and then I missed it) but does the D300 in default settings add sharpening to RAW images? My D90 did (even though I always shoot in full manual) and I had to turn that function to 0.

jfriend

Tiziano wrote:

Maybe you said this in your post (and then I missed it) but does the D300 in default settings add sharpening to RAW images? My D90 did (even though I always shoot in full manual) and I had to turn that function to 0.

There is no sharpening done to the D300 RAW images. There is sharpening metadata in the RAW file so if you use a Nikon RAW processor, it will read and apply that metadata, but I'm using Lightroom which does not read the sharpening metadata from the RAW file.

CynthiaM

jfriend wrote:

First off I should say that I realize sharpening can be a very advanced topic and that every image is technically different and can be optimized with different sharpening techniques. If one is producing fine art photographs for sale, that is certainly the philosophy to follow and I own several sharpening books that cover that level of detail.

But ... I'd like to ask for folks ideas on something simpler. I shoot a lot of events (softball games, soccer games, track meets, school events). It is not uncommon for me to publish (in web galleries) 100-300 photos per week. I am not going to be individually sharpening every single shot. Instead, I need to develop some "rules of thumbs" for how to do appropriate bulk sharpening in Lightroom so, right before I export to JPEG for upload to the web, I can apply whatever sharpening on the whole shoot. Since I know there are lots of other event shooters out there that are not going to tweak sharpening on each individual shot before putting them up into Smugmug, I'd love to hear what you all do for sharpening settings in Lightroom.

I should add that the photos I put on the web are uploaded as full res originals. Viewers will do one of three things with these images. 1) View the reduced web-sized versions of the images on the web or 2) Download the high res originals to either make their own prints or just add to their own digital album or 3) Order prints via Smugmug (which will be made from the high res original). So, because the images I'm preparing are multi-purpose, the sharpening has to be too (which necessarily means it's a compromise).

Here's where I get confused in Lightroom. Lightroom has the Sharpening controls in the Develop module (which default to amount=25, radius=1.0, detail=25, masking=0). It also has sharpening in the Export dialog called "Output Sharpening" and has options to sharpen for "screen" or "matte paper" or "glossy paper" and has amounts of "low", "standard" and "high". So, I'm trying to figure out what are some intelligent values to use for the typical sports shooting I do (sample images below). I shoot RAW with the Nikon D300 and either 70-200 or 200-400. I'm usually shooting at f/2.8 or f/4 for subject isolation. For workflow reasons, I'm looking for something that can be done purely via Lightroom.

What should I set the sharpening controls in the Develop module to?
What should I set the export sharpening controls to?

I've been getting OK results with 75,1.0,25,0 in develop and "Screen"/"Low" in export, but I feel like this is just a wild guess and there are probably better ways of doing this or figuring out what would work best. Any thoughts on bulk sharpening for these types of images would be appreciated.

Hey John:

I recently came across this post and since you have been such a great help here to me, I figured that I would give this a stab.

First off, the sharpening that is applied throughout Lightroom, both in the develop and the print modules and on export(and I think also the web module? but not so sure there) is based on the work done by the Pixelgenius group http://pixelgenius.com/products.html). In the develop module, the sharpening there is what they call capture sharpening. The print module uses their inkjet output sharpening and the export dialog will apply either their inkjet sharpening (for choosing glossy or matte) or screen which is meant for images that have been reduced to 72ppi.

I also upload full size jpegs to smugmug. What I do if they are in large batches like you describe is to apply the glossy sharpening to the full size image. They look fine to me once they got on line even if they are displayed as a smaller image on smugmug.

The issue of printing becomes another problem. When you print through smugmug, these labs are not using inkjet printers which is what the glossy or matte sharpening is meant for in the export dialog. These labs use contone printers, which is a different algorithm or a different sharpening output in the Photokit Sharpener program. Ifyour viewers are downloading to do their own printing, most likely if they send the file out to a lab, most labs will also use contone printers. For the sake of simplicity, what you need to decide is what type of printers are your viewers most likely to be using; my guess is contone. But then you say, there is no contone output in Lightroom. What you can do, is to create an export action in Photoshop using the Photokit output for Contone printing, then when you upload to smugmug from lighroom, get your batch together and in the export dialog, I think the last box at the bottom lets you apply an export action, so it is here where you can apply the sharpening for Contone output. It's not ideal because folks are viewing these on screen, but the truth of the matter is the images will probably look fine on smugmug. The other option is to use whatever sharpening you thinks looks good once the images make it to smugmug, use a print delay, then re-upload the image with the contone output sharpening for printing to one of the smugmug labs.

As for your question about settings for sharpening in the Devolop module, what I suggest you do is take one of your typical photos and sharpen it, make a preset to the sharpening settings and then apply that preset to your images. I find that I am often turning the sharpening amount up to about 50-65, leave the radius close to the default, detail is often between 25-35 and crank up the mask to somewhere between 50-70 ( to see what the mask is doing, when in detail, go into a 1:1 view, then hold down the alt key and move the mask slider; image will go grayscale and what is white is sharpened, what is black is not, kind of like an edge sharpening).

Hope that helps,
Cynthia

arodney

CynthiaM wrote:

First off, the sharpening that is applied throughout Lightroom, both in the develop and the print modules and on export(and I think also the web module? but not so sure there) is based on the work done by the Pixelgenius group http://pixelgenius.com/products.html). In the develop module, the sharpening there is what they call capture sharpening. The print module uses their inkjet output sharpening and the export dialog will apply either their inkjet sharpening (for choosing glossy or matte) or screen which is meant for images that have been reduced to 72ppi.

Correct. And yes, there is some sharpening applied if so set in the web module (output sharpening). The only area where NO output sharpening is applied is using the Export to Photoshop command. Otherwise, in Print, Web etc, capture sharpening (done in Develop) is applied to output sharpening if set to be on.

The issue of printing becomes another problem. When you print through smugmug, these labs are not using inkjet printers which is what the glossy or matte sharpening is meant for in the export dialog. These labs use contone printers, which is a different algorithm or a different sharpening output in the Photokit Sharpener program.

Correct although I suspect that one of the export settings (Standard, Low, High) would work pretty well. You should test this of course. Then you’d export out of the Print module to a JPEG or Export Preset using one of the above settings. I suspect that depending on the type of “contone” printer, low or high would work.

As for your question about settings for sharpening in the Devolop module, what I suggest you do is take one of your typical photos and sharpen it, make a preset to the sharpening settings and then apply that preset to your images. I find that I am often turning the sharpening amount up to about 50-65, leave the radius close to the default, detail is often between 25-35 and crank up the mask to somewhere between 50-70 ( to see what the mask is doing, when in detail, go into a 1:1 view, then hold down the alt key and move the mask slider; image will go grayscale and what is white is sharpened, what is black is not, kind of like an edge sharpening).

The Alt/Option slider shows you the mask which is very powerful. But you are limited to doing the capture sharpening “by eye” here unlike capture sharpening in PhotoKit Sharpener. That said, with the same kind of subject/lens, if you find one sweet setting, make a preset and apply it to the others. Capture sharpening is subtle and the controls in LR were made so you can’t be too aggressive (well you can but not anything like you could break moving the slider in USM in Photoshop). You should view at 100% or greater and more is less! Better to under sharpen than over sharpen.

While the ideal way to do capture sharpening is one image at a time, viewing the masks, doing this in bulk can certainly work and save a ton of time. And its totally non destructive. You can always go to an image, scale back or increase the slider and no harm done. This all applies to raw data. With JPEGs, the camera may be applying sharpening and its not non destructive.

CynthiaM

Andrew:

Any thoughts as to what kind of sharpening to apply for those of us on smugmug who choose to upload full size images (we can tell smugmug to display the image in a smaller size which when downloaded is only at 72ppi, but we upload a full size jpeg)? Like John, the original poster, I'm never quite sure what to do with these. Somehow screen doesn't seem right because it is being applied to a full size image. But neither do inkjet or contone seem right as that is not what it is being output for. I do it anyway for inkjet, usually at standard and it usually looks fine.

Appreciate the input.

Regards,
Cynthia

arodney

CynthiaM wrote:

Andrew:

Any thoughts as to what kind of sharpening to apply for those of us on smugmug who choose to upload full size images (we can tell smugmug to display the image in a smaller size which when downloaded is only at 72ppi, but we upload a full size jpeg)?

Why do you upload full size images (what’s the output)? If web, why would you upload a file that large for output to the web?

CynthiaM

arodney wrote:

Why do you upload full size images (what’s the output)? If web, why would you upload a file that large for output to the web?

I upload full sized jpegs for several reason:

1. It becomes another form of backup. Although I have backups for my backups, if all should fail, I at least have full sized jpegs on smugmug.
2. If someone orders a print, smugmug needs a full sized image for printing.

When you have a smugmug account, as the owner, you have the control/ability to tell smugmug that while you are uploading a full sized image, to only display a reduced size, presumably to lesson the impact of image stealing. You can also disable right clicking. So I upload full sized images and tell smugmug to display a reduced size. That way, if someone were able to undo the right click disabling the best they could download would be something like 800x533 at 72ppi. But the original file that was uploaded is something like 3504 x 2336 at 240ppi but you don't see that. Nonetheless, that is what gets exported from a program like Lighroom, ergo the qaundry, how do you sharpen these files?

arodney

Output sharpening is done to a fixed size and a fixed device so this isn’t going to fly.

Isn’t the raw the original that you’d want to backup (remotely), rather than a rendered raw?

CynthiaM

arodney wrote:

Output sharpening is done to a fixed size and a fixed device so this isn’t going to fly.

I realize that, which is why I go by eye, by the aethetic, for smugmug purposes. If it looks good by the time it gets displayed on smugmug, I'm happy.

Isn’t the raw the original that you’d want to backup (remotely), rather than a rendered raw?

Yes, but I can't send raw files to smugmug. So I send them to mozy.com. In the meantime, if my house goes up in smoke, or gets flattened/drowned in a hurricane AND the mozy server goes under AND all of my backup DVDs get corrupted, at least I have rendered full sized jpegs.

jfriend

arodney wrote:

Why do you upload full size images (what’s the output)? If web, why would you upload a file that large for output to the web?

Andrew, it might help if you realize that you are at dgrin which is Smugmug's forum where nearly all of the people here are Smugmug customers. Smugmug offers (and many customers use) the ability to both view images on the web (in many different sizes) and to let friends, family or clients order prints that are prepared by one of two labs that Smugmug works with. If you want to put images up for display that people can also order prints from, it is desirable to upload the full resolution that you have (as it makes better prints than a downsized version). Smugmug will automatically make web-sized versions from that for web viewing, while retaining your full resolution image for use in generating prints.

For purposes of the convenience of one-stop web and print ordering, it is a compromise. One image is prepared that must serve the needs of both web and print. Yes, it is a compromise, but that is the context here on dgrin and most of Smugmug's customers have selected this convenience in exchange for that compromise. So, that is the context of my original question at the start of this thread. I was looking for helpful advice on how to best prepare the sharpening of such a full resolution image to serve this compromise.

As you can see by rereading the first post in this thread, I fully understand that if every image was hand sharpened for a particular purpose and a particular output device, one could optimize things better, but when I'm posting 800 images from a soccer season into 30 galleries for parents to either download, share with friends or order prints from, I'm not going to be doing hand sharpening of multiple sizes of each image. I'm going to look for some batch process that I can apply to all of them - upload the full resolution images to Smugmug and let the parents decide what kind of output they want.

The original point of this thread was to solicit Lightroom sharpening advice for this scenario.

arodney

jfriend wrote:

Andrew, it might help if you realize that you are at dgrin which is Smugmug's forum where nearly all of the people here are Smugmug customers. Smugmug offers (and many customers use) the ability to both view images on the web (in many different sizes) and to let friends, family or clients order prints that are prepared by one of two labs that Smugmug works with.

Great. So just realize its a far from ideal workflow in terms of ideal output sharpening since that is resolution and output device specific.

For purposes of the convenience of one-stop web and print ordering, it is a compromise.

Agreed!

The original point of this thread was to solicit Lightroom sharpening advice for this scenario.

Throw a dart at the barn and be done. There is no correct sharpening advise in this context, LR like PhotoKit Sharpener of which it was based uses capture and output sharpening for specific sizes to apply the sharpening. IF you sharpen for the full rez image, the smaller images (for web or print) may or may not be acceptable. And we as yet have no idea how the system resamples (there are three Bicubic options in Photoshop alone, in LR its an adaptive sizing), and we don’t know if the system applies any sharpening or blurring when sizing. So again, if you insist on using this workflow and service, YMMV (a lot).

jfriend

arodney wrote:

Great. So just realize its a far from ideal workflow in terms of ideal output sharpening since that is resolution and output device specific.



Agreed!



Throw a dart at the barn and be done. There is no correct sharpening advise in this context, LR like PhotoKit Sharpener of which it was based uses capture and output sharpening for specific sizes to apply the sharpening. IF you sharpen for the full rez image, the smaller images (for web or print) may or may not be acceptable. And we as yet have no idea how the system resamples (there are three Bicubic options in Photoshop alone, in LR its an adaptive sizing), and we don’t know if the system applies any sharpening or blurring when sizing. So again, if you insist on using this workflow and service, YMMV (a lot).

You may say throw a dart and be done, but the entire point of this thread was to figure out how to throw an intelligent dart. If you don't have anything to add on that point, we can be done.

FYI, we actually do know that Smugmug uses the Lanczos algorithm for creating the display copies after reduction and they give you the default parameters that are used for that and we can even change them in our own gallery settings. So, we do know roughly how display copies are done.

arodney

jfriend wrote:

You may say throw a dart and be done, but the entire point of this thread was to figure out how to throw an intelligent dart. If you don't have anything to add on that point, we can be done.

Well again, based on the fact that size and output plays a role here, its something you’ll have to thoroughly test a number of ways, getting output to various sizes and comparing them to the screen rendering, again at various sizes. This is exactly what Bruce Fraser did initially to produce PhotoKit Sharpener. He used many differing kinds of images (images with both high and low frequency), from different capture devices and at multiple sizes to multiple devices. Recipes had to be developed for optimal results. You can use on such recipe for suboptimal results to all but on of the above conditions.

FYI, we actually do know that Smugmug uses the Lanczos algorithm for creating the display copies after reduction and they give you the default parameters that are used for that and we can even change them in our own gallery settings. So, we do know roughly how display copies are done.

Knowing the algorithm is like one ingredient in a complex recipe. Its like saying “use Unsharp Mask” then not defining the settings for amount, threshold and Radius. Not real helpful.

If you insist on using a one size fits all workflow here, fine. But to then ask what is the optimal way to handling sharpening, well it goes back to throw the dart a few times, hope for the best and move on. You can’t throw it intelligently when so many factors are not provided can you?