View Full Version : Unsharp Mask applying methods
I have answered a question posted by Glenn Kroeger in the past. Because of my interest on the matter, I post again to know some opinions about USM. Thanks.
Some points about USM,
a) Any digital image, both from scanners and digital cameras, will need some amount of USM in order to be sharply viewed at final step (screen, inkjet, photogrphic paper, etc.). There are a lot of particular characteristics of digital imaging process that induces lost of sharpness.
b) The amount of USM necessary for a given picture depends on, essentially, three questions: The first is the scanner or digital camera own performance. The second arises from the amount of present detail (image frequency) in a given picture. The third, finally, is related with the final step of visualization (screen, inkjet, photogrphic paper, etc.).
c) All the stated above concludes to the impossibility of the existence of any rule of thumb about wich is "the best" amount of USM or "the best" smart software to apply it.
d) There are at least two questions to have into account: It seems obvious that we need some test of our equipement (scanner and/or camera) to try the best amount of USM to apply to our pictures. In second place we need to analyze EACH picture in particular and the amount of texture and detail to decide the appropriate amount of USM to be applied. It is too important to realize if it is necessary or advisable to apply the USM or the same USM to the whole picture area.
e) Using the USM tool a gaussian blur is applied to a copy of the original image to produce a new image with no fine detail; the blurred image is then substracted (mathematicaly) from the original image to extract the fine detail. This fine detail is added to the original to highlight the detailed areas.
f) The RADIUS determines the amount of gaussian blur first applied (Probably large files need more radius than small files to be the same effect). The AMOUNT defines de quantity of detailed version added to the original. The THRESHOLD indicates how different in brightness (from 0 to 255) must be two adjacent pixels to enables the USM be applied.
g) The threshold permits to avoid some areas to be contrast enhanced by the USM, like the face skin in a portrait or the sky in a landscape. May be interesting to analyze previously the neighbor pixel differences in those areas with the INFO Palette in order to decide the thershold to be applied.
Disagreement or discussion will be welcome,
Sometimes an image looks a little "hazy" and even good curve corrections don't fully do the trick. In those situations, for overall contrast enhancement with no perceptible loss in image quality, try using USM with a huge radius and small amount sometime: Amount 10-20%, Radius 50, Threshold 0 is a good ballpark.
This does not replace the sharpening step, it's just an interesting thing that the USM filter can do.
This is my way of working with unsharp maskes. When looking at prints from digital files I noticed that you often see contrast enhancement in the mid fine detail range. This can be seen at bounderies between dark and light. At both sides of the dark and light the dark is darker and the light is lighter, thus enhancing the contrast. In real life this hardly ever happens (with the exception of butterfly wing patterns). In our retina there are connections which produce the same effect, but we also use this retina to look at photographs, so there should be no reason to put it in the photograph.
On the other hand I agree with Carles Mitja that a loss in contrast in for instance the scanner should be compensated for.
So my way of working is the following: using a given scanner and film format, I make a series with USM settings. From this series I select the image with the smallest radius and with the highest amount in which the contrast I measure (with photoshop) in the final image is not increased.
This compensates for the limitations of the scanner, but does not produce this brutal digital sharpness effect, completely pushing away subtle details
I usually apply different amounts of USM to different areas of the photograph depending on the amount and importance of detail in different areas. I use a selection tool to select large areas with little or no tonal range difference (e.g. the sky or large areas of skin in a portrait) and omit those areas entirely from USM. Sometimes I then apply USM uniformly to the remainder of the photograph or select other areas and apply some USM to them but less than to the remainder of the photograph depending on the tonal range and the amount of detail desired. I always set "threshold" to zero since if I want to apply USM at all to a given area I want to apply it to everything in that area. I usually set "radius" between 1 and 2. Anything more than 2 for me seems to always lead to an image in which the application of USM is very obvious. "Amount" is the biggest variable for me. I think of "Amount" as like the volume control on a radio - the bigger the number the more you get. I use the preview window in USM to determine the appropirate setting for "Amount" and it can range from as little as maybe 75 to as much as 300 or so depending on the image. I always apply USM with the image on the screen at the actual anticipated print size since USM is less obvious with smaller prints (i.e. what looks right with the image at 5x7 may be too much when printed at 16x20).
I don't know that this is the "best" way to work, I've just found that it works well for me. I can't stand that "overly sharp" look you get with a digitally printed photograph where it's obvious that too much USM has been applied. I'd much rather have slightly too little than too much USM. The way I do it is pretty much the way I was taught to do it in George DeWolfe's digital printing workshop.
There is a program for applying USM that can be downloaded from the Luminous Landscape web site, I think it's found under the "Tutorials" section. I used this method for a while and it seemed to work very well. However, there are a lot of steps and it's time-consuming to do manually. If I knew how to convert it into a usable Photoshop Action I'd probably use it all the time but I didn't find the improvement it made over my usual way of working to be big enough to justify the time it takes to do it manually.
I might add that I don't understand paragraph (e) in Carles' message and the first and second sentences of paragraph (d) seem to conflict with each other. Any explanations, especially of (e), would be welcome.
In the paragraph (d) of the previous post I'm trying to say: If there is stated that any digital image have a certain loss of sharpness dued to the equipement itself, it seems logic to begin applying some amount of USM that counteracts this fact. In addition, we can apply some different amounts of USM in differents parts of the image to enhance the sharpness in those areas. In any case, the amount of USM that we need is different in each image but always depending on the kind and brand of equipement used.
In the paragraph (e) of my previous post, I simply try to describe the procedure used by the software (Photoshop or any other Image Processing Software) under the tip "Unsharp Mask". From this procedure we can infer the consecuences of deal with Amount, Radius and Threshold settings. It may result a bit confused, but you can find complete explanations in classic books about Digital Image Processing like:
R. C. Gonzalez, R. E. Woods, Digital Image Processing, Addison Wesley, 1993, ISBN 0-201-50803-6
A. K. Jain, Fundamentals of Digital Image Processing, Prentice Hall, 1989, ISBN 0-13-336165-9
W. K. Pratt, Digital Image Processing, 2nd Edition, Wiley, 1991, ISBN 0-471-85766-1
I hope to clarify,
there is a very useful little (free) suite of sharpening tools here (courtesy of their maker - John Brownlow):
You just download (the first two files), unzip the .atn files and then in photoshop load them as actions, then use the actions in button mode.
No instructions, so it's a bit of a learning curve. Different ones work better on diferent types of images/files.
I find I use the Deadman's Custom Sharpen! most on 4x5 and 8x10 scans from my mid range umax powerlook (and an Epson 1680). It's an edge sharpener, selecting the edges which you can then sharpen up tp 200% (I usually find somewhere from 90-125 works on what I photograph most, which is buildings right now and FWIW I set the first dialogue box that comes up to 1 and the second to 1.8 and take it from there - but there is a lot of room to play around/experiement/learn with all 5 actions).
I rarely find a scan of this sort that doesn't benefit from using them and does away with some of the inherent softeness from this type of scanner (they work very well in 35mm - I've also used the HiPassHardLight sharpener on scans from somewhat soft. blurry arhives prints and it has actually had the effect of rudcing the apparent blurriness - up to a point).
These days I rarely use the (fairly crude) photoshop USM
Like Tim, I've used Deadman's for a while now. I like the idea of having such visual control in the process. Slightly more time consuming, but I'm not in the production process, so time is not an issue for me.
A very rational approach to the three areas where sharpening can/needs to be used (acquisition/creative/output) are embodied in a product called Photokit Sharpener ( http://www.pixelgenius.com/sharpener/index.html ). I'm not in any way affiliated with this product, but I bought it and like it. I found USM's flexibility overwhelming and this handles digital images from all sources in a logical workflow with a range of "pre-sets" and creates a series of layers in Photoshop which thereby provide great control over subjective issues and simplifies their implementation. I have a whole lot of 35 mm transparencies, digital images from several digital SLRs and (most recently, which is why I'm here!) scans of 4x5 transparencies, all of which need different treatments (aside from any image-specific needs) and I print to a digital printer. The drudgery of determining optimal USM settings for each of these combinations has been done by the makers of the product, and while it doesn't do anything PS can't do, it really makes it easy to be efficient.
Tim, I'll also be checking out the (free) actions you linked to - thanks - and PK Sharpener also has a free trial download - it may not be better, but the learning curve is easy and they have good documentation.
I haven't tried the specific sharpener (a bit high priced for me right now when the Deadman ones seesm to do what I want - though it does look pretty interesting)
BUT I do have the original photokit - while I don't use a lot of it it does have a great set of dodge and burn tools in it - especially the dodgeing top 2, bottom 2 or all four corners. Great for those wide angle blue skies etc that darken just too much at the corners - I use it quite often.
Tim, Fair enough on the price issue, if a free product does the job!
As I understand it, the new product is much more comprehensive in its capabilities. I like it for two main reasons - it has a logical workflow that is easy to master (as opposed to having to figure out one's own, a big time saver) and the output sharpening , a whole series of settings for different output dpis and different printer types. The onscreen version of the resulting output file looks so ugly I would never have thought of trying to print it, but the prints get "dithered" by the droplets and the sharpening takes that into account - result :- a sharp looking print that DOES NOT LOOK OVERSHARPENED - much to my surprise and delight - this is not something that can be adequately conveyed electronically! Its use of layers means it's RAM hungry but they come in the form of masks where it is easy to erase the effects in parts where it's not needed - if the sky looked a little noisy but the rest of the image looked improved by a certain set of parameters, just paint out the sky on the mask to restore its smoothness.. definitely worth checking out the free trial and you have to make a print to give the output sharpener a fair assessment!
Powered by vBulletin® Version 4.2.0 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.