What order to do things?

[scatts]

Hi,

I am gradually coming to terms with most of the basic adjustments available in Photoshop CS. However, I am not yet perfoming them in the same order each time and I can't help thinking this must make a difference in either quality of end result or time taken.

Most used are: sharpening, saturation, adjust levels/curves, cropping, resizing, rotating. Sometimes conversion to greyscale. I mostly use the basic equipment but also have a couple of plug-ins for Velvia Vision and sharpening.

Does anyone know or has seen an article on what is the best order to do these tasks?

Thanks.

[ckimmerle]

1. Rotate
2. levels/curves
3. Hue/Saturation
4. Crop
5. Resize
6. Sharpen (always last)

You can crop anytime in the process, but I like to wait until the end of the toning steps so that I can see what effect burning/dodging have on the overall image.

Chuck

[dougj]

1. Convert from RAW (Photoshop CS)
2. Crop
3. Resize
4. Adjust levels
5. Burn/dodge if required
6. Shadows/highlights
7. Brightness/contrast
8. Color balance
9. Hue/saturation

10. Sharpen: convert to Lab mode, select lightness channel, USM 200-300/0.3-0.4/0, convert to RGB mode
11. Hue saturation
12. For web use, final sharpening: USM 50/5/0 (also good for removing haze)
13. Convert from 16 bit to 8 bit
14. Add border & drop shadow


As Chuck indicated, sharpening should be last. I like lightness channel sharpening (step 10) as there is less pixelation, and prefer to crop & resize early so I can see the image as it develops in processing. For JPG files skip steps 1 & 13.

The easiest way to do the post processing is to record an "action" in Photoshop, this automates the process with stops at each step, keeps them in sequence and you don't forget to do a step.

Doug
http://www.pbase.com/dougj

[scatts]

Thanks guys,

I have never recorded an action but will give it a try. Do you do your post processing on adjustment layers or on the background layer directly?

Why is sharpening for the web (presumably = display on PBase) different to simply sharpening until you like the way it looks?

I'm using a plug-in for sharpening now, which seems to go through a lot of stages and so is presumably doing it 'intelligently' but it does have a 'volume control' and so I could sharpen a few times with differing strengths.

I have, until now, sharpened quite early in the process, I'll change that now.

For your pictures on PBase what are the characteristics of the final image you upload - size, resolution and so on. I just work on each one until I thinks it's okay and then upload but I think I might be missing something. They all end up as different sizes, resolutions and so on. I'd like to tidy things up.

BTW, I always convert the 8 bit to 16 bit before doing any processing, I read in a book it reduces the damage done and if you look at the difference in the levels histogram (far fewer white gaps of missing data) it does seem to help.

[dougj]

Scatts,

I usually do the processing directly on image unless there is a need for a layer. Sharpen the image so it is appealing on the monitor when your intention is post it on the web, printed images generally will take a little more sharpening.

Concerning image sizes for pbase, many people prefer to post images that are no more than 800 pixels wide or 600 pixels high, I use moderately larger, and when saving the image I select a compression that results in a final file size of about 200 KB.

Doug