Elaray

Question and suggestions please but let me explain myself a bit first.

Work on my husband's digital images a lot so I know the basics on sharpening a picture, even learned the hard way to not over-sharpen. I play with the pictures, I don't snap them, that's the husband's hobby.

I've been taught to leave all the resolution in a picture based on the theory that more information in an image is good. Say I crop a 5x7 and it has a resolution of 495. I then run the unsharp mask at 500% at .2 radius with 0 threshhold. If that doesn't clean it up, I'll run it again at a .1 radius.

The images I shrink down so that I can place them on the web-- I reduce them to a resolution of 200 and generally the size of 600x400 pixels. When I sharpen them the same way with less resloution, seem to turn out much cleaner.

I understand it's probably because there are less pixels in the image, so is it better to reduce my resolution on the big images I send to print? Or are my eyes decieving me?

I know just enough about this stuff to get me in trouble. Any suggestions on using the unsharp mask and helping me understand the resolution mystery would be greatly appreciated. 8))

Bacardi

I've been taught to leave all the resolution in a picture based on the theory that more information in an image is good.

In fact, this is only true for print. Inside Photoshop or anything digital, this is not important at all. For Photoshop (or a webbrowser) a pixel has its place (second row, third from the left) and its colour information (so much red, so much green and so much blue).
As far as the web is concerned, you get as many pixels as your image is wide and high.
If your monitor is set to display, say, 800 wide, and your image is 810 wide, then you will have to scroll. But this does not change the image's quality.
If you open the same image in Photoshop, it can be displayed in its entirity, but only because Photoshop does not show you all pixels. Simply because PS cannot change the size of your pixels. That is a priviledge of your OS (win XP, OSX,...). To see them all, you need to set it to 100% (double click the magnifying glass icon the toolbar).

For print, this is different. Here Photoshop can tell how many pixels will become dots of the printer in each inch. Normally, 266 to 300 per inch is perfect. 150 is still ok, but less will become really visibly less good.
So, to print 5x7inch at 300 dots per inch, you will need 5 x 300=1500 , and 7 x 300=2100 pixels.

All Photoshop itself is interested in is 1500x2100 pixels that it has to work with and that it needs to reserve memory for.

Say I crop a 5x7 and it has a resolution of 495. I then run the unsharp mask at 500% at .2 radius with 0 threshhold. If that doesn't clean it up, I'll run it again at a .1 radius.

This method is not bad. Doing things is small increments is always a good idea. There are other methods though. I wrote my ideas down in a text that's called "a chautauqua on sharpening" and that you can find in the tutorials section. Yes, I admit: it is a bit chaotic, but I really wrote like as if I were experimenting. many methods are in there. You may want to read it, and you can ask questions here in this thread. The main thing is that "sharpening" is in fact enhancing the contrast of edges which makes images appear as "sharper".

The images I shrink down so that I can place them on the web-- I reduce them to a resolution of 200 and generally the size of 600x400 pixels. When I sharpen them the same way with less resloution, seem to turn out much cleaner.
I understand it's probably because there are less pixels in the image, so is it better to reduce my resolution on the big images I send to print? Or are my eyes decieving me?

In your example, this means that you have a 2475x3465 pixel image. I bet your monitor cannot show all of them without you having to scroll.. You have perhaps 1600x1200 so under no circumstance Photoshop can display all these. To see a complete image, it has to omit a certain number of pixels (for viewing only!!). The quality of this is, understandably, less good.

When you downsample (I guess you also do this in small icrements, and not from 300+ to 600 in one step, and that you omit sizes that are easy to divide like half, or one quarter) to 600, you do see the image at full size on your monitor (except perhaps on a 640x480 one, but I don't think you can run PS on that). Because all pixels are shown, you get a "real" idea. To compare quality wize with the image of 3000+ pixels, you should look at this at full size, meaning you cannot see the full pic, but only a detail of it.

Dodo

I'm glad Erik explained that - I would've become confused halfway through it [confused].

Another thing to note on is that when you create new images you can also select a resolution (pixels per inch). Again this is only for printing purposes (as far as I know) - and 72 dpi is a traditional selection - some folks say 75 dpi is better due to some mathematical considerations - mainly because scanning at that gives you roughly the same dimensions on a screen.

Changing this figure does not affect your file size - that's just dependent on the screen or rather, pixel information. What it does do is affect the size it will be output at. If you go to "Image --> Image Size" it gives you a "Document size" in either centimetres or inches (or whatever). This makes it more quantifiable in real-world terms. Note that there are some slight variances with different printer types that you should also consider.

So to "prove" it to yourself you can create two images with the same pixel dimensions - but different resolutions. Then print them and compare to the size given by the "Image Size" information. It should line up. You will see how the same pixel information is spread out - the dpi is effectively the density of the information on paper.

Just one other thing I found strange - Sheba - you indicate that you work on your husband's digital images (as well as nerves? - just kidding). Most digital cameras take their pictures at 72 dpi - it's the megapixel resolution that defines their "quality". Or does he use a film camera and then scan the images? That would explain how you start out at such a high dpi resolution.

Rantin Al

Hi Sheba.

Erik seems to have covered the topic on resolution well enough, so
I won't overload you with more details.

You might want to try this other method for sharpening, using the High Pass Filter.
Details in this thread:

http://www.photoshopgurus.info/forum...gh+pass+filter

Cheers, Al.

Elaray

Oh thank you all! I have much to read and digest now. I'm sure I'll have some questions in a day or two, but I'll let this sink in first.

To dodo: I probably *do* get on his nerves, this is why I need a fresh angle from you fine folks. [bustagut] My husband's camera is a Nikon D1X and he also has a D2H, digital SLRs. I wish I understood more about the camera end of it, but my *job* is finishing the pictures, (and packing camera equipment around lol) I edit them while they are a nef, and convert them to a jpeg to send to the printer.

He's good at taking sharp pictures but I have heard before that all digital images need to go through the sharpening process. (?)

I get dpi and ppi messed up. dpi = print quality then. *taps forehead*

And my main concern is with getting the best clean image on print. I didn't express that well enough in the beginning of my first post, I see. :B

I'll go read now... Thanks for the thread referrals that have already discussed this. I thought I had scanned through the board better than that! [excited] Thanks again!

Rantin Al

Sheba, another tip particularly if you want the best quality is not to use jpg for your printer format. Keep the images as tif or native psd files. The jpg is a 'lossy' compression, you will lose some pixel information even at high quality.

Cheers, Al.