Is this best method to sharpen a pic in pshop?

Is this the best method to sharpen my pics?
I use Pshop6
Go to image> mode> lab colour> channel> lightness> sharpen/unsharpen mask
Then amount 100%. radius1.0.threshhold 0 levels
Then image> mode> RGB
The results look good BUT is there another way that may be better
Thanks for any help
Michellei

Michelle,

If there were just one way to apply USM, there wouldn't be all those variables. How you select variables will be determined by both the subject matter, resolution, and inherent sharpness of the image to begin with. Yes, USM sounds daunting. But the important thing to remember is to practice and experiment. You must learn through experience how to apply the sharpness you want to an image without destroying the image. Although I've certainly used the settings you described before, I've also used something like 300%/ 0.5/ 4, just as an example. When I shoot JPG files in my D1, I often go ahead and use the in-camera sharpening in Low Mode. Some use Normal, some use None. When I shoot JPGs, I'm generally shooting images that I intend will have little manipulation and processing later. I still ad just a little USM in PS to these files. When I shoot RAW NEFs, the image has no inherent sharpening, and my settings are quite different and considerably stronger.

You'll need to simply experiment and learn to see. As you do, here is an excellent discussion on sharpening:

http://www.bythom.com/sharpening.htm

Good luck

M

michelle said:

Is this the best method to sharpen my pics?
I use Pshop6
Go to image> mode> lab colour> channel> lightness> sharpen/unsharpen mask
Then amount 100%. radius1.0.threshhold 0 levels
Then image> mode> RGB
The results look good BUT is there another way that may be better
Thanks for any help
Michellei

Click to expand...

michelle said:

Is this the best method to sharpen my pics?
I use Pshop6

Click to expand...

Something to consider. Ultra-sharpen pro 5.0, a plugin for PS6 does a very nice job. It is available for download at http://www.ickybits.com/ .

This uses masking to sharpen what you want sharpened and leaves the rest of the pic alone. Works well on portraits where you want to avoid sharpening the blemishes....

Another thing you can do, is sharpen the RGB image, then fade with "luminosity". Gets you to the same place as working on the lightness channel.

One last thing. - take a look at this thread
http://www.photo.net/bboard/q-and-a-fetch-msg?msg_id=001DGz

The Bruce Fraser method Johnny mentions works exceptionally well.

Hope this helps

You might also try the following plugin:
http://www.theimagingfactory.com/website/download/11USM

It sharpens the brightness of RGB images which has two advantages over the method mentioned:
+ no conversions to and from Lab necessary,

+ preserves the color of the enhanced edges: sharpening the Lightness channel in Lab images will result in white edges where color was represented previously...

just another method though,
regards,
O

You might try this tool, Quantum Mechanics. Rob Galbraith thinks it is essential when using Photoshop, as a finishing up tool (per a private conversation with him). He uses it on all his final output from Photoshop!
Here is the URL to the manufacturer:
http://www.camerabits.com/
Steve

michelle said:

Is this the best method to sharpen my pics?
I use Pshop6
Go to image> mode> lab colour> channel> lightness> sharpen/unsharpen mask
Then amount 100%. radius1.0.threshhold 0 levels
Then image> mode> RGB
The results look good BUT is there another way that may be better
Thanks for any help
Michellei

Click to expand...

Hello Michelle,

Here is a method I learned from PEI magazine. PEI stands for Photo> electronic imaging. Their web site is http://www.peimag.com

The June 2001 issue is chocked full of special tutorials, one being sharpening techniques in PS using USM. Make sure you check it out before you buy alot of plug-ins for PS. All that you need is right there in PS, although it is a long process to figure them all out. The tutorial contains four methods for sharpening in different situations. All can be scripted as an action in PS.

Smart Sharpening- This method sharpens the image edges without sharpening the noise, film grain, pores, or out of focus areas. This may seem like to much trouble but once it is scripted it is a repeatable action that is easy to apply with one click of your mouse.

1. Duplicate the Background layer.
2. Duplicate the channel with the highest image contrast.
for example the red, green, or blue channel.
3. Select Filter> Stylize> Find Edges.
4. Invert the channel by slecting cmd/ctrl-I.
5. Select Filter> Noise> Median and use a value of 2 to accentuate the edge
lines.
6. Select Filter> Other> Maximum and use a value of 4 to spread the edge lines
even more.
7. Select Filter> Blur> Gaussian Blur and apply a value of 4.
8. Return to the composite view (cmd/ctrl- ) in the Channels palette.
9. Load the channel mask (cmd/ctrl-opt/alt-4).
10. Apply the unsharp Mask filter to sharpen just the edges of the subject,
without sharpening the noise and grain. Try Amount -300, Radius-1.0,

Threshold-0 to start with and adjust to your liking or output. For inkjet prints,
photo prints or monitor veiwing the amount of sharpening should look
natural on your monitor. For offset printing the amount of sharpening
should look a little, just a little, overdone.
11. Select select> deselect and your finished.

If buy chance you sharpen an edge you don't want to sharpen you can use the eraser tool to erase that part from your duplicate layer.

Please take the time to try this and record it as three different actions one for the red, green & blue channel. Then in the future when you go to sharpen an image, once you have decided which channel gives you the edge contrast of what you want to sharpen you can just run that action.

I thought this was to much trouble at first but when I figured it out and saved the actions, I am amazed at how well it works.

Happy USMing.

David B. Nicolay

Daivd B. Nicolay said:

Smart Sharpening- This method sharpens the image edges without
sharpening the noise, film grain, pores, or out of focus areas.
This may seem like to much trouble but once it is scripted it is a
repeatable action that is easy to apply with one click of your
mouse.

1. Duplicate the Background layer.
2. Duplicate the channel with the highest image contrast.
for example the red, green, or blue channel.
3. Select Filter> Stylize> Find Edges.
4. Invert the channel by slecting cmd/ctrl-I.
5. Select Filter> Noise> Median and use a value of 2 to accentuate
the edge
lines.
6. Select Filter> Other> Maximum and use a value of 4 to spread the
edge lines
even more.
7. Select Filter> Blur> Gaussian Blur and apply a value of 4.
8. Return to the composite view (cmd/ctrl- ) in the Channels palette.
9. Load the channel mask (cmd/ctrl-opt/alt-4).
10. Apply the unsharp Mask filter to sharpen just the edges of the
subject,
without sharpening the noise and grain. Try Amount -300, Radius-1.0,
Threshold-0 to start with and adjust to your liking or output. For
inkjet prints,
photo prints or monitor veiwing the amount of sharpening should look
natural on your monitor. For offset printing the amount of sharpening
should look a little, just a little, overdone.
11. Select select> deselect and your finished.

Click to expand...

Could you elaborate on steps 2 (duplicating a channel) . If I select a channel and duplicate that layer, the duplicated layer has all channels. Do I deselect the other 2 channels again?

What layers are active when I move to step 3?

At step 8, return to composite view in Channels palette, do I go to layers and select or deselect any of the layers created above?

Step 9, how do I load the channel mask as CMD-Opt-4 doesn't do anything (MAC)?

Step 10, which layers are active?

I think somehow the mask created by the specific color channel is supposed to be applied only to the background layer created in step 1.

Thanks for your help.

Kent

Kent said:

Daivd B. Nicolay said:

Smart Sharpening- This method sharpens the image edges without
sharpening the noise, film grain, pores, or out of focus areas.
This may seem like to much trouble but once it is scripted it is a
repeatable action that is easy to apply with one click of your
mouse.

1. Duplicate the Background layer.
2. Duplicate the channel with the highest image contrast.
for example the red, green, or blue channel.
3. Select Filter> Stylize> Find Edges.
4. Invert the channel by slecting cmd/ctrl-I.
5. Select Filter> Noise> Median and use a value of 2 to accentuate
the edge
lines.
6. Select Filter> Other> Maximum and use a value of 4 to spread the
edge lines
even more.
7. Select Filter> Blur> Gaussian Blur and apply a value of 4.
8. Return to the composite view (cmd/ctrl- ) in the Channels palette.
9. Load the channel mask (cmd/ctrl-opt/alt-4).
10. Apply the unsharp Mask filter to sharpen just the edges of the
subject,
without sharpening the noise and grain. Try Amount -300, Radius-1.0,
Threshold-0 to start with and adjust to your liking or output. For
inkjet prints,
photo prints or monitor veiwing the amount of sharpening should look
natural on your monitor. For offset printing the amount of sharpening
should look a little, just a little, overdone.
11. Select select> deselect and your finished.

Click to expand...

Could you elaborate on steps 2 (duplicating a channel) . If I
select a channel and duplicate that layer, the duplicated layer has
all channels. Do I deselect the other 2 channels again?

Click to expand...

When you have selected the Red,Green, or Blue channel of the background copy, to the right of the channels palette in the title bar is a circle with an arrow in it. If you click on it you will see duplicate channel. This will make a copy of the channel you selected at the bottom of the channels palette. Red copy, green copy, or blue copy. That channel is active. Once you have duplicated the channel in this way all the other steps will fall into place. All work is done in the channel copy of the background copy

Member said:

What layers are active when I move to step 3?

Click to expand...

The copied channel of the background copy.

Member said:

At step 8, return to composite view in Channels palette, do I go to
layers and select or deselect any of the layers created above?

Click to expand...

No, while still in the channels palette hit, cmd-~ then cmd-opt-4

Member said:

Step 9, how do I load the channel mask as CMD-Opt-4 doesn't do
anything (MAC)?

Step 10, which layers are active?

I think somehow the mask created by the specific color channel is
supposed to be applied only to the background layer created in step
1.

Click to expand...

this action is applied to the copy channel of the background copy. I had trouble getting it at first but I checked the directions and the above list is correct. I think once you copy the channel of the background copy properly by duplicating it in the title bar of the channels palette and not from Layer> Duplicate Layer all will work just fine. Remember once you get it to save an action for each channel. feel free to email me.

David

Member said:

Thanks for your help.

Kent

Click to expand...

Daivd B. Nicolay said:

When you have selected the Red,Green, or Blue channel of the
background copy, to the right of the channels palette in the title
bar is a circle with an arrow in it. If you click on it you will
see duplicate channel. This will make a copy of the channel you
selected at the bottom of the channels palette. Red copy, green
copy, or blue copy. That channel is active. Once you have
duplicated the channel in this way all the other steps will fall
into place. All work is done in the channel copy of the background
copy.

Click to expand...

When I clicked on the circle with the arrow, the "Duplicate channel" is grayed out and not selectable. I checked in Photoshop Bible and saw that if you drag the channel name (i.e., red) over the new channel icon at the bottom, it will duplicate the channel.

You were right - everthing else fell into place. I tested it with a D1x image from Phil's review and it looks great. I could not tell the difference from sharpening the same image with the Quantum Mechanic Pro demo.

Next, I'll try it on a resolution chart scanned with the Nikon CP 4000 to confirm there is no impact on noise and grain.

Thanks again,

Kent

You might want to also read Bruse Fraser's articles at creativepro.com.

They can be found at the following urls and they cover some of the territory in this thread.

http://www.creativepro.com/story/feature/12189-1.html
http://www.creativepro.com/story/feature/11242-1.html

Mario Georgiou
http://www.georgiou.com

All of the previous posts have good info on sharpening, but I usually try to stick with the simplest. First of all, yes, there are great third party tools for enhanced sharpening. But frankly, unless you are dealing with an image that has problems (noise, grain, damage) the unsharp mask tool can do it all. Yes, you can fine tune your way with lots of layers and masks, but let's focus on the result: is the image going to web, print, or display? If to print, is it inkjet, or offset, or other?

Sharpness is a factor a multitude of factors, the most important of which are not subtle issues like CCD size and the quality of one piece of glass or the other. The single most important aspect to sharpness is contrast. Sharpness is the perception of contrast between areas of different luminance and color. This contrast is enhanced by adequate lighting, reduced camera shake, smaller apetures, and of course, good focus. Sharpness is also a function of the viewing conditions. An image printed in newsprint will benefit from significant OVERsharpening to enhance contrast, whereas the same image printed offset at 175LPI will look artificially constrasty and almost posterized or solarized slightly.

I use the unshapr mask wetting with my output in mind. As an earlier poster said, experiment, and trust what you see, not what the numbers are telling you. I recently had some success with some images of lacross players I shot running an unsharp mask setting of 60%, 26 pixels, 2 threshold. Wow, you say, what's that look lie? Well, it does crush some of the fine detail, but it enhances the overall punch and contrast that I was looking for in the image. Try it! These kinds of extremem setting can produce great results from soft or out of focus shots that HAVE to be saved.

Otherwise, my basic D1 setting is soemthing like 100%, 1.2 pixels, 2 threshold for fine jpegs. This mostly works, but only when I am batch processing a bunsh of stuff I am not too concerned about.
Kevin