Digital workflow questions: RGB/sRGB, sharpening, TIFF/jpg


Status
Not open for further replies.

David

Deregistered
Mar 21, 2002
1,204
0
0
Visit site
I'm finally getting all geared up for processing my digital images in bulk after realizing I've got a back log of literally hundreds of images! (Btw, I shoot mainly RAW.)

But I realize there are some tight issues for me to deal with, which I hope CS members could kindly advice and share:

1. We know that sRGB is for monitor display; Adobe RGB if want to print out the images. Problem is, if I want to BOTH display my images on screen (say for friends to see) and also print some of them out, how do I save the images? Do you all have 2 sets of images, one in sRGB, and for those to be printed, in Adobe RGB? :bigeyes:

2. After the RAW conversion, do you save the images in TIFF or jpg? TIFF sucks memory but is lossless even after many editings. But this takes up unnecessary disk space and print labs usually only require jpg to print.

3. Images that appear perfectly sharp on screen usually print soft. For printing, I notice I need to sharpen even more than what I get on screen. But this becomes bad on screen as they look over-sharpened. So again HOW?? Do you all again create 2 sets of images? One for viewing on screen another for printing? :bigeyes:

*headache* Digital photography is fun but post-processing is NOT! For me at least. :cry:

Would certainly appreciate any help! Thanks....
 

1) I think you got it mixed up, it's the other way round.

2) TIFF for VERY large and precise work prints, JPEG for normal prints up to S12R.

3) Yes, it happens. Just leave it to the lab to sharpen for you.
 

Thanks espn.

1. Huh.... Isn't sRGB used for web display and Adobe RGB is preferred for printing? :eek: AFAIK, I tend to get a reddish hue and over saturation if I save my images as Adobe RGB and display them using Windows browser.
 

you might want to try this new format with extention .PNG. smaller then tiff and still lossless.
 

1) For best results...shoot in RAW...edit in Adobe RGB...save as sRGB jpeg for both web and printing. Or, can shoot in large/fine jpeg....edit in Adobe RGB...save as sRGB jpeg for both web and printing. Also can, shoot in large/fine jpeg....edit in sRGB...save as sRGB jpeg for both web and printing.

Of all the standard color spaces, sRGB comes closest to that used by most minilab printers such as Fuji Frontiers, Noritsu QSS, and Agfa d-Labs printing on consumer grade paper. Use sRGB for home printing also.

2) Don't convert to tiff unless sending the photo to a magazine. Also, you loose your EXIF if you convert to tiff format. Suggest you convert to an uncompressed jpeg.

3) "Images that appear perfectly sharp on screen usually print soft." Well...what is your workflow like for printing? Should be very different than your 'save for web' workflow! Yes, you will have 2 sets of images because you need a high-res image for printing!

For example, with my Canon 20D (all parameters in camera set to zero)*...if I want to print to 4R size, workflow would be somthing like this: crop -> resize to 6MP (20% reduction in size) -> adjust brightness/contrast -> adjust levels -> adjust color balance -> USM 300/0.3/1 -> adjust curves -> boost saturation -> save as a jpeg (quality 10).

I apply the sharpening (USM) only to the Lightness Channel in the Lab Mode... don't forget to convert back to RGB color after the sharpening step.

*same USM for RAW converted jpeg using default Bibble Lite sharpening setting. See my post here.
 

r52lanc said:
2) Don't convert to tiff unless sending the photo to a magazine. Also, you loose your EXIF if you convert to tiff format.
Lose your EXIF? My TIFFs all have EXIF intact.
 

Royce said:
Lose your EXIF? My TIFFs all have EXIF intact.
Yikes! You are correct! Great...can export from Bibble to PS as 16-bit tiff, then edit, and save a jpeg! Great! Thanks! :)
 

1. shoot in RAW or jpeg(fine) with preset parameters in sRGB

2. transfer files from cf card to computer

3. correct for

(i)levels
(ii)curves
(iii)brightness/contrast
(iv)white balance
(v)hue/saturation
(vi)colour balance

4. sharpen

5. save as TIFF - these are the only files i will back up

for web, just make another copy of the TIFF file, resize, sharpen, save as jpeg, upload. - these files i will not back up

for printing, same thing just make another copy of it, save as jpeg. again, will not back up these

this is what i currently do. if theres a better workflow then i will follow ..hehe
 

Thanks all, especially r52lanc for your sharing...

Just more clarifications needed below:

1. So when you have converted to jpg, you are dead sure you won't do any more editing right? I assume your workflow would be RAW --> TIFF (editing using this) --> save final image as jpg?

2. When you convert to 6MP from your 8.2MP Canon 20D, is it part of resampling in Bicubic? I read from Photoshop books that up- or down-sampling deteriorates the image quality. Is there any other method to reduce the dimensions of the image without affecting the image?

3. Why do you use USM in Lightness mode and not in Adobe RGB directly?

4. Ok this ques is bugging me like hell for a long time!! Every standard textbook will say Adobe RGB has a larger color gamut than sRGB. So why do we use Adobe RGB for editing then convert it back to sRGB for printing? Isn't that deliberating throwing away possible colours? Or is Adobe RGB some idealistic colour that could never exist realistically as prints or web display?

Please please forgive me if my questions sound simplistic.... I find this topic really confusing which I hope to iron out asap so I can get on with my editing. :cry: Thank you so much!!!
 

Thanks all, especially r52lanc for your sharing...

Just more clarifications needed below:

1. So when you have converted to jpg, you are dead sure you won't do any more editing right? I assume your workflow would be RAW --> TIFF (editing using this) --> save final image as jpg? My workflow is a bit more complex than I make it sound. I do save my original editing work in a *.psd file. This file has a layer stack that includes adjustment layers plus a frame. Then, I save a jpeg version in a separate file for web or printing. Sometimes, I go back and tweak one of the adjustment layers.

2. When you convert to 6MP from your 8.2MP Canon 20D, is it part of resampling in Bicubic? I read from Photoshop books that up- or down-sampling deteriorates the image quality. Is there any other method to reduce the dimensions of the image without affecting the image? Yes, I use the PS default which is Bicubic. I have read that reducing size in several steps is preferable to a single resize step...say from 8MP to 0.24MP (600x400). Genuine Fractals software has their own algorithm for reducing the size of an image. No one has been able to prove that it is any better than Bicubic. BTW, reducing size actually reduces noise but degrades sharpness. That is why, you need to sharpen after size reduction.

3. Why do you use USM in Lightness mode and not in Adobe RGB directly? Sharpening the lightness channel only... eliminates color haloing.

4. Ok this ques is bugging me like hell for a long time!! Every standard textbook will say Adobe RGB has a larger color gamut than sRGB. So why do we use Adobe RGB for editing then convert it back to sRGB for printing? Isn't that deliberating throwing away possible colours? Or is Adobe RGB some idealistic colour that could never exist realistically as prints or web display? Well, I edit in Adobe RGB because I can ...also, I edit in 16-bit mode because I can! Theory is that you will loose less information during editing. You can print in Adobe RGB if your printer has an ICC profile for that colorspace. See here for example. Be careful... a commercial lab might assume you have used sRGB.

Please please forgive me if my questions sound simplistic.... I find this topic really confusing which I hope to iron out asap so I can get on with my editing. :cry: Thank you so much!!!
No problemo!
 

if i have the profile of the printer (my printer). what do i do with it?

Set it as my working space and save it as that?

Then print using that profile or using sRGB?
 

Actually I have the same question as David regarding Colour Space. Why does printers use sRGB when Adobe RGB is touted as a better gamut?
 

I read somewhere that it's because the ink/dye system cannot reach that kind of gamut. I may be wrong.
 

Aiyoh... I really have given up. I conclude all this color space thingy is really for the colour scientists to debate. Different people I meet and hear say different things.

I've just read from another Photoshop book where the author totally put down sRGB as an inferior color space to use. Maybe for Web display/ emails only as he said but even then, he highly recommended Adobe RGB for everything you do, from editing to printing and yes, even for computer monitors!

Incidentally, the Canon 20D manual will tell you they highly recommend you set it to sRGB instead of Adobe RGB without giving any useful reasons.

Now I just see what is nice, I use. Adobe RGB seems the way to go. Don't want to waste my time reading any more or fretting over this issue. I just want to enjoy photography!
 

yanyewkay said:
if i have the profile of the printer (my printer). what do i do with it?

Set it as my working space and save it as that?

Then print using that profile or using sRGB?
No, lah! You can select a profile for your print space. Put the printer profile on your harddrive where the other profiles live. Then, select it from your printer menu.

My 2 cents.... edit in Adobe RGB.... save in sRGB for web... print in whatever printer profile make you happy!
 

not sure if i should post here
along the same lines, if i SCAN SLIDES, should i select colour matching to be in AbodeRGB or sRGB?
(should i follow r52lanc's workflow ie scan in AbodeRGB thus edit in it and later save it in sRGB for web/print)

Also what does color depth: 8bit and 16bit and 16bitlinear mean? Which one should i choose to scan slides?
 

One word: Calibrate ur Monitor.

If u dun calibrate ur monitor properly, den wadever colour space u set at oso useless, cos the tinge of colour that the monitor will lean towards will always be there. And u'll either get red rosy faces, yellow sickly malaria struck people, green incredible hulks or blue horror Ju-On faces.

My photos used to lean towards a heavy red on-screen and printed... till I calibrated my monitor, yesterday tried a print with those regular graphics and was rather impressed by the printout. Tonight going try it again with a photo I took at the pro shutterbug event. :D
 

DTan said:
not sure if i should post here
along the same lines, if i SCAN SLIDES, should i select colour matching to be in AbodeRGB or sRGB?
(should i follow r52lanc's workflow ie scan in AbodeRGB thus edit in it and later save it in sRGB for web/print)

Also what does color depth: 8bit and 16bit and 16bitlinear mean? Which one should i choose to scan slides?
Dun touch linear :thumbsd: unless you know what you are doing. If you are editing in Adobe Photoshop CS then use 16bit/AdobeRGB....must convert to 8bit/sRGB before saving for web. If using older Photoshop version, then scan in 8bit/AdobeRGB. Have fun!
 

Status
Not open for further replies.