any tweaks (advice) to Sharpening for 50MP sensor?

[SimonsPictures]

Is there any tweaks to the advice on sharpening for 50MP sensors?

i.e. when printing or saving for web, using a sensor with a small pixel pitch iike the 5DsR and the 7D2, is there any change (up or down as a general rule) needed to the Radius, Detail (in LR) etc?

In Lightroom 6 there is a 'detail' slider and I'm not sure whether this had an identical partner in Photoshop (CS6 perpetual licence)

I have read somewhere that ramping it up to 100% changes the algorithm and so if so, is it useful for small pixel pitch sensor images sharpening?

Im guessing that for web it is just a case of doing it by eye as usual

How about prints though? Does the advice hold as per the viewing distance and ppi?

Also any advice specifically for using Nik sharpener pro?

thanks

[NorthernFocus]

If you're asking about final sharpening, i.e. output sharpening, then pixel pitch of the sensor is not relevant. Output sharpening settings should be chosen for the final image size/resolution, not for the sensor resolution.

[dje]

Hello Squire

I think with this sort of question, you need to separate the sharpening process into two - Capture sharpening and Output sharpening.

With capture sharpening, you are compensating for a loss of sharpness caused by the lens and sensor (including the AA filter if the camera has one). With cameras with fine pixel pitch, the lens tends to be the major contributor to loss of sharpness, particularly if there is no AA filter as in the 5Ds r. The pixel pitch in both cameras you mention is about 4um, as it is in my Sony a6000. I find with this camera that default settings in ACR (which is similar to LR) of Amount 40-50, Radius 0.7, Detail 50-70 and Masking about 25-30 work quite well. I would suggest these as a starting point. With the 5Ds R and a good lens, you may have to back off the amount to avoid over-sharpening. As a general rule, full frame sensors should need less sharpening than a crop sensor (because the image size is larger).

From what I have read, when the Detail slder is set to 100, a form of deconvolution sharpening is used. With it set to 0, a form of Unsharp mask is used. In between there is a mixture of both. I find that with Detail set to 100, it is too easy to get nasty grain like artefacts if you over-sharpen, particularly with human faces.
.
I look at Capture sharpening as part of the process of producing a nice clean full res "master image". After that, you may use the image in several different ways and this where Output sharpening comes in.

If you are displaying for the web, you will be downsizing and this is a process that inherently cause loss of sharpness. Some down-sizing algorithms include in-built sharpening eg Bicubic Sharper in PS but I prefer to retain control over this form of Output Sharpening myself and use straight Bicubic with sharpening added in PS using Smart Sharpen. (typical settings amount 40 and radius 0.5.)

With printing, you are compensating for loss of sharpness in the printing process and I really thinnk this requires trial and error and experience to get it right. Both LR and ACR have some form of Output Sharpening options but with ACR at least ( I can't speak for LR), there is no real control. If you Output sharpen in PS, you can use Smart Sharpen or Unsharp mask to get whatever degree of sharpening you want. Or you can use a plug-in such as Nik but I'm afraid I can't offer any advice on this.

With printing, I would think that the main requirement is to get the required dots per inch that the printer requires. Having a 50MP image will make this easier for very large prints.

Dave

[Manfred M]

+1 to what both Dan and Dave have written.

Unfortunately, when it comes to sharpening, there are no "general rules" and each image (or series of images) need to be assessed individually at each stage of the process. If you read the Jeff Schewe and Bruce Fraser book "Real World Image Sharpening with Adobe Photoshop, Camera Raw and Lightroom (this is considered by many to be the definitive reference for image sharpening), they make that point over and over again and clearly demonstrate why this is so.

http://www.adobepress.com/store/real...-9780321637550

1. Capture sharpening is used to compensate for issues from the AA filter (not an issue in the case on the 5D) but other factors including the lens you used, shooting conditions, ISO settings on the camera will all impact how you approach this step. With high ISO images I will not sharpen until after I have done some noise reduction,

2. In process sharpening - I don't do this 100% of the time, but some images need a bit of local sharpening, applied to specific areas of the image. I find that I will use this to enhance eyes, lips, eye brows and eye lashes in some of my portraiture. I've also used this in some landscape work to enhance some details in the image.

I will also do some global sharpening, depending on the image. I probably do 99% of my sharpening with the Unsharp Mask in Photoshop, but from time to time use Smart Sharpen and the High Pass Filter.

3. Output sharpening - This is a bit of a tricky processes as you will be upsampling and downsampling your original image and this process can cause unwanted softening or sometime too much sharpness in the image. I tend do a fairly global approach on anything I post to the internet as I can't control the size of the screen or the quality of the screen anyone is viewing the image at.

When I print, I will adjust output sharpening based on the type of paper I am printing on. A glossy print will take less than a matte paper. Of course print size is also a major factor here.

[TheBigE]

+1 to what both Dan and Dave have written.

Unfortunately, when it comes to sharpening, there are no "general rules" and each image (or series of images) need to be assessed individually at each stage of the process. If you read the Jeff Schewe and Bruce Fraser book "Real World Image Sharpening with Adobe Photoshop, Camera Raw and Lightroom (this is considered by many to be the definitive reference for image sharpening), they make that point over and over again and clearly demonstrate why this is so.

http://www.adobepress.com/store/real...-9780321637550

I am looking seriously at purchasing this book, my one concern is that it was published nearly 6 years ago. I am not sure if that is the first or second edition, since then there have been many updates to PS/LR. I presume the book is much more conceptual based and not deep into the specifics of these programs. In which the concepts COULD BE outdated.

My goal would be to better understand Sharpening overall, the techniques available in LR/PS, and a bit of a dive into how to look at areas of an image that could be improved with Sharpening. It seems like the book would be a good resource.

I am really ready to pull the trigger on this book, but just wanted to check that the relative age of the publishing has no impact on the information.

Thanks

[James G]

Erik,
I have had the book for some years now. (I use Photoshop CC and Lightroom.) I know of no other text which covers sharpening issues and techniques in as much depth. I still find it relevant and continue to use it today.

Jeff Schewe's 'The Digital Negative' is also worth considering, particularly since sharpening and its different approaches/techniques are considered as part of a coherent and integrated workflow for developing an image from capture through to final presentation.

Ultimately though, I have to agree with the previous comments already made earlier in this thread that sharpening is dictated by the specific characteristics if the image being worked on. As such a flexible approach to the use of the available tools needs to be adopted.

[Manfred M]

I presume the book is much more conceptual based and not deep into the specifics of these programs. In which the concepts COULD BE outdated.

The book is still as relevant today as when it was written. While DSLR sensor resolution has improved (I remember retouching a 50MP image out of a Hasselblad digital back in 2010 or so), but the issues related to de- mosaicing an image is just as relevant today as it was then. The AA filters on the modern high resolution cameras have either disappeared or become much weaker, but frankly this only impacts one aspect of sharpening; import sharpening.

Sharpening for compositional purposes (making certain features stand out more) are somewhat independent of camera resolution as what we really do in sharpening is to enhance edge contrast. That's something a camera can't do if it isn't there or isn't pronounced enough in the original image (so the need for in-process sharpening has not changed).

Likewise when we prepare a final image for output, especially if we print, that is another aspect of sharpening that is not affected by the sensor resolution.

In my mind, read the book and get to understand sharpening. It is (to a certain extent) independent of sensor resolution. The tools, from cameras to Photoshop, have changed. The underlying principles have not.