Ad
Ad
Ad
Pages: [1]   Bottom of Page
Print
Author Topic: profile conversion  (Read 2677 times)
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« on: September 07, 2005, 11:09:56 AM »
ReplyReply

No. Converting to CMYK is a complete waste unless the final image needs to be CMYK, and then it is better to wait until all image editing is complete and convert to CMYK as a final step. Never convert from one color space to another unless you have a specific reason for doing so, and always do all editing and color space conversion in 16-bit mode to minimize quantization and rounding errors. Doing a single color space conversion in 16-bit mode will always yield better results than any multi-stop color space conversion in 8-bit mode.

Dan Margulis knows a lot about Photoshop, but his opinions on color management and the advantages of 16-bit editing are not based on practical reality or a solid understanding of the subject matter. Andrew Rodney, Bruce Fraser, and Ethan Hansen are far more knowledgeable and credible in those areas than Dan Margulis or Will Crockett.
Logged

pfigen
Sr. Member
****
Offline Offline

Posts: 469


WWW
« Reply #1 on: September 10, 2005, 11:18:50 AM »
ReplyReply

"I had looked at Professional Photoshop but decided not to buy it, since it emphasized CYMK, which I do not use."

You should reconsider. You'll learn more about color theory in general than all the other books put together. The principles that Dan lays out apply to any color space, and the new Lab book will completely blow you away. With that, you'll rethink practically everything you thought you knew in the first place. It's a book that is going to take at least a dozen readings to get into. And, there's no reason you can't incorporate Dan's techniques into a fully color managed workflow. He shows you where they fall short, and teaches you to decide how to deal with it.
Logged
bobrobert
Guest
« Reply #2 on: September 11, 2005, 06:55:55 AM »
ReplyReply

Jonathan could you explain how you mask of the highlights and shadows? TIA
Logged
bobrobert
Guest
« Reply #3 on: September 06, 2005, 06:10:09 AM »
ReplyReply

There seems to differences of opinion as to some of Dan Margulis's theories In his book Professional Photoshop he recommends a profile conversion from Adobe RGB 1998 to CMYK using RGBK and another users profile which I believe belongs to the Curvemeister site to minimise the loss of colour due to conversion and back again to Adobe RGB 1998 Does this work in reality compared to the "normal" conversion theories? TIA
Logged
bjanes
Guest
« Reply #4 on: September 10, 2005, 08:15:21 AM »
ReplyReply

Quote
Dan Margulis knows a lot about Photoshop, but his opinions on color management and the advantages of 16-bit editing are not based on practical reality or a solid understanding of the subject matter. Andrew Rodney, Bruce Fraser, and Ethan Hansen are far more knowledgeable and credible in those areas than Dan Margulis

I had looked at Professional Photoshop but decided not to buy it, since it emphasized CYMK, which I do not use. I see Dan has a new book on editing with LAB and this is on the bestseller lists.

Andrew and Bruce have stated that this space (LAB) is not the most intuitive space to work in and I'm not sure if I want to suffer the learning curve needed to give it a try. What do you think about LAB editing?
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #5 on: September 10, 2005, 11:24:22 AM »
ReplyReply

I actually do most of my sharpening In LAB mode. Here's my basic workflow:

1. Convert RAW, doing as much CA correction, exposure and color adjustment as possible to minimize tweaking these later. Output to 16-bit ProPhoto.

2. Neat Image if necessary, custom noise profiling the image if possible.

3. Edge sharpening with Focus Magic, usually Radius 2 and amount 75-100%.

4. Convert to LAB mode and select the L channel only.

5. Do 5 iterations of USM sharpening on the L channel with the highlights and shadows masked off to prevent clipping, starting with a large radius and small amount, and working to a small radius and large amount. This allows extremely agressive sharpening without causing halo artifacts; you can sharpen as much as the image wants without having to worry about sharpening artifacts being a problem.

6. Shadow/Highlights on the L channel with no masking, tonal width and radius 15, amount 5-15% depending on the image.

7. Convert to RGB.

8. Any remaining miscellaneous stuff like cloning dust/acne, minor color tweaks, etc.

One handy way to get rid of CA is to run the dust/scratches filter on the A & B channels in LAB mode. Then select that as the history brush state, go back one step, and use the history brush to paint away the CA. This also gets rid of the type of purple fringing that is the same color on all sides of the object, which standard CA correction will not fix.
Logged

Pages: [1]   Top of Page
Print
Jump to:  

Ad
Ad
Ad