OK – a simple anonymous request…
“Sharpening – when to do it. In camera, in software?”
- tell us your hints and tips for the best methods and also when not to sharpen.
Prick by Archidave. All Rights Reserved
About Exposure Leeds
Exposure Leeds is a new type of photographic society - created with an open, sharing and friendly ethos our focus is on developing skills and experience and in presenting the results to a wider audience.
All are welcome and anyone can become a member - come along to one of our events or join in one of our initiatives to see the benefits affiliating yourself with us can bring.
How I sharpen tends to depend on the workflow I am following for that photograph.
If I got the photo I wanted out of the camera then I will adjust the sharpening using the RAW conversion software ( I use the Canon software that came with the camera but I am trying other options) and do nothing else.
If I intend to do any Photoshop work on it, or process it as an HDR, I remove as much sharpening as I can when I convert from RAW then use Unsharp Mask in Photoshop as the last step. I do this because I read that sharpening more than once simply adds to the pixelation and the ‘artifacts’ in the photograph, and RAW conversion sharpening is still software sharpening.
With the unsharp mask I try to find the place where the sharpening clarifies the content, but doesn’t make it stand out. I usually try and reduce the sharpening after I have decided on the right level, in the same way as they say that a woman should look in the mirror on her way out the door and remove the piece of jewellery that catches her eye, I like to take a last look at the photo and try and see whether any area of particular sharpness catches my eye, and tone it down a little.
For anyone who struggles with sharpening, I can highly recommend a set of actions by Photokit called PK Sharpener. They’re available at the Pixel Genius website and whilst not cheap, they do a perfect job without having to think too much.
For those of you who can’t afford or don’t want to pay that much, the general theory is that you first apply some ‘capture sharpening’ This is just enough sharpening to get rid of the slight blurriness that you have in most digital and scanned shots. Keep a close eye on high contrast sharp edges and make sure you don’t end up with a bright line edge (halo). Using unsharp mask with a 1 pixel radius and very little threshold.
Once you’ve done this, you can sharpen your picture locally by creating a second layer and ‘oversharpen’ it. Then mask this layer and paint in the bits that may need a little extra sharpening – go easy on this step as you can easily make a picture look wierd. Same 1px radius on unsharp mask.
Finally, when it comes to printing, view your picture at 50% and apply just enough sharpening to make things crisp without creating halos using a 1-2px sharpening.
– or —
If you have lightroom, just use it’s standard sharpening for capture and print. The sharpening algorithm was designed by the guy that created Photokit Sharpener..
The guy is called Bruce Fraser (A genius who recently passed away sadly) and the book is reviewed and some of the concepts discussed here
Tim
Tim
I tend to use a slightly different technique for sharpening. Instead of using the sharpening tools already in photoshop you can create a duplicate layer and go to filters>other>high pass (in CS4, the high pass filter is in a different place in other versions). Usually you need to set the high pass filter somewhere between 2.5 and 8, depending on the shot. You then use Soft Light as a blending mode. You are adding an extra layer, so it can add more noise to the picture, but it tends to give you a lot of control over the amount of sharpening you get.
For those that shoot with digital cameras let me give you a hint or two to get sharp images.
1. Shoot raw and process with your camera makers software. In my case that’s both ViewNX 2 and NX2. Every setting in the camera is honoured and you can treat a RAW like a processed jpeg but with all the after the capture adjustment benefits.
2. Leave in camera sharpening ON. On my Nikons I leave a middle of the road sharpening setting – number 4. For what I primarily shoot, people, this counteracts the AA or low pass filter which largely softens the image.
3. In post and dealing with my NEF files I can apply local sharpening to eyes etc.
4. Final step is to apply sharpening at the output size you intend to print or display on the web.
Whilst I do use Lightroom and yes V3 is very good, I still maintain that utilising ViewNX2 and NX2 I get better raw file conversions with much less sharpening anomalies once printed. Furthermore, make some prints at various sharpening settings – what you se on your screen is not what you’ll see in your print.
As a general and final step in NX2 and for both my old D200 and my newer cameras I finish with a sharpening step of 36 / 12 / 4 and a final High Pass step anywhere between 1 and 3 pixels.
On screen you’d run screaming from my office saying I was mad but printed on a Epson3800…beautiful.
And finally, have a look at my flickr account, every photograph goes through this process.
On the high pass suggested by Alex. You can reduce the increase in noise by applying a slight gaussian blur to the high pass layer. Too much blur and you get “halos” around your edges so don’t overdo it.
A good read are these sharpening articles on the Ron Bigelow site. More techniques than you will ever use, but you learn a lot about how it works by reading and trying.