Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: Color in Digital Capture  (Read 9533 times)
Anders_HK
Sr. Member
****
Offline Offline

Posts: 1001



WWW
« on: December 31, 2007, 06:52:07 AM »
ReplyReply

Hi, I shot Fuji Slides before, Velvia was a magic and I honest still enjoy use it and I do not feel like giving it up. Film is sort of magic   , and with it to me one special aspect: the Colors. Perhaps you read http://luminous-landscape.com/forum/index....970&hl=challeng. There the digital captures look a little pale, or as one user posted there towards the end:

"Thanks for doing all this and sharing the images. The film scans have a real wonderful quality to them. While looking at these last 100% crops, I find that I'm really drawn to them and the ZD images that looked so great before look lifeless in comparison."

Please do not get me wrong here, I am not trying to compare film vs. digital or prove that one or other is better, I know they are different. I much like the medium format digital captures also. Although... I feel they lack in color. So, I am keen to improve my skills of color in processing. I currently use SilkyPix, Camera Raw & Photoshop CS3, and I do not think it is as simple as a saturation boost... I need much more advanced and appealing to my aesthetic eye.

Please point me to any advise or sources to improve on color in my digital captures to very pleasantly and appealing appoach... or to even exceed that of Fuji Slides

Thanks!  

Regards
Anders
« Last Edit: December 31, 2007, 06:54:33 AM by Anders_HK » Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #1 on: December 31, 2007, 07:20:14 AM »
ReplyReply

One approach would be to use the camera profiling function of Eye-One Match. You'll need to shoot the test chart with the film you wish to emulate, scan the film using the same output RGB space (ProPhoto, Adobe RGB, or whatever) as you use for your RAW output RGB space. Make the film profile from the scan in Eye-One, and save it as Velvia.icc or whatever.

Now all you need to do is convert your RAW captures to the film profile, then assign your working RGB space to the file, and you'll have exactly the same color "look" as the film.
Logged

Anders_HK
Sr. Member
****
Offline Offline

Posts: 1001



WWW
« Reply #2 on: December 31, 2007, 07:48:51 AM »
ReplyReply

Quote
One approach would be to use the camera profiling function of Eye-One Match. You'll need to shoot the test chart with the film you wish to emulate, scan the film using the same output RGB space (ProPhoto, Adobe RGB, or whatever) as you use for your RAW output RGB space. Make the film profile from the scan in Eye-One, and save it as Velvia.icc or whatever.

Now all you need to do is convert your RAW captures to the film profile, then assign your working RGB space to the file, and you'll have exactly the same color "look" as the film.
[a href=\"index.php?act=findpost&pid=164191\"][{POST_SNAPBACK}][/a]

Jonathan,

That is very truthful interesting, but... is it not that film is non linear while digital is linear? If we look at Velvia 50 as example, it captures one thing in bright light, but put it at a waterfall with dim lights in forest due overcast skies, it aborbs colors like nothing eles   ...

Perhaps the answer is rather some 'magic' applied in PP, or??? Any suggestions?

Regards
Anders
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #3 on: December 31, 2007, 04:20:50 PM »
ReplyReply

Quote
That is very truthful interesting, but... is it not that film is non linear while digital is linear?

Only in the RAW capture data. Once you demosaic the RAW or convert the scan data to a standard RGB space (all of which are non-linear), digital color is whatever the profile in the RAW converter or scanner software makes it. So if you follow the procedure I outlined above, you can make any digital camera output Velvia color within the measurement accuracy of your spectrophotometer as long as the digital exposure is not clipped.

Linear/non-linear only affects how you expose for optimal image quality.
Logged

Anders_HK
Sr. Member
****
Offline Offline

Posts: 1001



WWW
« Reply #4 on: December 31, 2007, 11:46:51 PM »
ReplyReply

Quote
Only in the RAW capture data. Once you demosaic the RAW or convert the scan data to a standard RGB space (all of which are non-linear), digital color is whatever the profile in the RAW converter or scanner software makes it. So if you follow the procedure I outlined above, you can make any digital camera output Velvia color within the measurement accuracy of your spectrophotometer as long as the digital exposure is not clipped.

Linear/non-linear only affects how you expose for optimal image quality.
[{POST_SNAPBACK}][/a]

Jonathan, thanks.    


Jonathan and All,

Thus camera calibration and profiling per [a href=\"http://luminous-landscape.com/reviews/accessories/dig-calib-profil.shtml]http://luminous-landscape.com/reviews/acce...ib-profil.shtml[/url], but using Velvia and drum scan to digitally accurately reproduce the Gretag Macbeth Colorchecker (x-rite).

Alternative (and perhaps in combination) can be to use this http://www.rawfilmstyles.com/ and not only for Velvia but other quick clicks to get pleasing colors and other aspects as an initial departure point for further PP.

Next step is to further tweak in postprocessing. Any good books or other readings to recommend, specifically on color, constrast and all other paramters? By all means not just a manual for Camera Raw or Photoshop but something that gets deeper. Film was developed over generations, and I guess I seek such expert advises on colors and other image rendering adjustments in digital captures.

Regards
Anders
« Last Edit: December 31, 2007, 11:50:29 PM by Anders_HK » Logged
marcmccalmont
Sr. Member
****
Offline Offline

Posts: 1724



« Reply #5 on: January 01, 2008, 11:23:34 AM »
ReplyReply

A little off the subject, when I downloaded DxO's film pack I tried the looks of many films but I kept coming back to the 5D's "as shot" as the most appealing to my eye. I think you like whatever you are conditioned to like. Try Fuji Astia (much more accurate, I never cared for velvia) and make the comparion.
Marc
Logged

Marc McCalmont
jjj
Sr. Member
****
Offline Offline

Posts: 3197



WWW
« Reply #6 on: January 01, 2008, 03:15:22 PM »
ReplyReply

Quote
A little off the subject, when I downloaded DxO's film pack I tried the looks of many films but I kept coming back to the 5D's "as shot" as the most appealing to my eye.
Are you talking about the RAW files or the Canon's JPEGs, as with adjustable picture styles you can get very different looks with one's JPEG images out of camera? And those looks are usually how you first see the RAW file or the only way depending on how you are viewing the RAW file, as many programmes use the embedded JPEG for previews.


Quote
I think you like whatever you are conditioned to like.[a href=\"index.php?act=findpost&pid=164385\"][{POST_SNAPBACK}][/a]
I'd disagree with how you like, what you like, up to a point.
But that is in a sense how fashion works as you become accustomed to new fashions and what was once avant garde, beomes quotidian with repeated exposure and other prior styles suddenly look dated. But it doesn't mean you like it.
« Last Edit: January 01, 2008, 03:22:04 PM by jjj » Logged

Tradition is the Backbone of the Spineless.   Futt Futt Futt Photography
TomJB
Newbie
*
Offline Offline

Posts: 5


« Reply #7 on: January 01, 2008, 08:51:13 PM »
ReplyReply

"Next step is to further tweak in postprocessing. Any good books or other readings to recommend, specifically on color, constrast and all other paramters? By all means not just a manual for Camera Raw or Photoshop but something that gets deeper."

On Amazon, you might investigate "Real World Color Management 2nd Edition" by Fraser, Murphy, and Bunting.

Although it is about color management, the first 100 pages is about what color is and the science of the perception of color.  I've learned an amount about color that amazes me.

Another book that is very much about color is "Photoshop LAB Color" by Dan Margulis.  Although I highly commend it, this book is extremely challenging that it may take multiple readings to get a handle on...  Do a search (google, yahoo, et el) on the title of this book and you'll find a wealth of discussion about it.

 I hope that helps.  - Tom
Logged
jjj
Sr. Member
****
Offline Offline

Posts: 3197



WWW
« Reply #8 on: January 02, 2008, 08:07:11 AM »
ReplyReply

Quote
Another book that is very much about color is "Photoshop LAB Color" by Dan Margulis.  Although I highly commend it, this book is extremely challenging that it may take multiple readings to get a handle on...  Do a search (google, yahoo, et el) on the title of this book and you'll find a wealth of discussion about it.[a href=\"index.php?act=findpost&pid=164494\"][{POST_SNAPBACK}][/a]
There's been a lot of discussion here about Dan M by some regulars who seen to loathe him! But then some of them cannot countenance an alternative viewpoint/way of working, so hard to tell if they are correct about Dan M or simply raving.
Logged

Tradition is the Backbone of the Spineless.   Futt Futt Futt Photography
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #9 on: January 02, 2008, 08:56:15 AM »
ReplyReply

Dan has some very strange notions about color management that are at odds with common sense and the consensus of color management experts. He doesn't beleive in the value of editing in 16-bit mode. And some of the sharpening methods he's devised recently are demonstrably inferior to many older, more established techniques. Dan got his reputation as a PS expert doing retouching, but his attempts to parlay that reputation into credibility in other areas have not fared well. Dan's other primary issue is that he tends to respond to criticism ad hominem instead of dealing with the actual issue being discussed.
Logged

sojournerphoto
Sr. Member
****
Offline Offline

Posts: 473


« Reply #10 on: January 02, 2008, 09:31:03 AM »
ReplyReply

Quote
A little off the subject, when I downloaded DxO's film pack I tried the looks of many films but I kept coming back to the 5D's "as shot" as the most appealing to my eye. I think you like whatever you are conditioned to like. Try Fuji Astia (much more accurate, I never cared for velvia) and make the comparion.
Marc
[a href=\"index.php?act=findpost&pid=164385\"][{POST_SNAPBACK}][/a]

I had a very similar experience, that I preferred the 'as shot' look with adjustments to the film looks (by and large). But, I do like the DXo tri X look.

Mike
Logged
Anders_HK
Sr. Member
****
Offline Offline

Posts: 1001



WWW
« Reply #11 on: January 02, 2008, 09:56:46 AM »
ReplyReply

Quote
Dan has some very strange notions about color management that are at odds with common sense and the consensus of color management experts. He doesn't beleive in the value of editing in 16-bit mode. And some of the sharpening methods he's devised recently are demonstrably inferior to many older, more established techniques. Dan got his reputation as a PS expert doing retouching, but his attempts to parlay that reputation into credibility in other areas have not fared well. Dan's other primary issue is that he tends to respond to criticism ad hominem instead of dealing with the actual issue being discussed.
[a href=\"index.php?act=findpost&pid=164563\"][{POST_SNAPBACK}][/a]

Who Dan?? Me?? - No single person is perfect, nor should it be said that following the flow of most is correct, right?

I read Dan's book on Lab early last year, very good book and indeed I should go back to it. The method of moving end points of A and B channels inwards do give a punch. He also has Professional Photoshop Ed 5, dec 2006. Is that one worthwhile for what I seek?

My arrival today was Adobe Photoshop CS3 for Photographers by Martin Evening; I have alredy started read and it seems like excellent book to thouroughly teach me details on Photoshop. Yet it seems vaguely to touch on what I seek in this post.

Is there someone who has written specific on the color sensations similar to photographic impressions of film in digital? Granted it is not only color; the CS3 command for Shadow/Highlight is jaw dropping in what it can do. Yet... I still feel captures from digital appear pale in comparison to Fuji Slides.

Much thanks for further advises!  

Regards
Anders
Logged
Anders_HK
Sr. Member
****
Offline Offline

Posts: 1001



WWW
« Reply #12 on: January 02, 2008, 10:40:29 AM »
ReplyReply

This is too simple; http://www.photonhead.com/photo-editing/

This one sounds very interesting (sort of Zone system Shadow/Highlight command on steroids???): http://www.digitalfilmtools.com/ozone/  , http://www.digitalfilmtools.com/cs/csozone.htm  

Anything more???  
« Last Edit: January 02, 2008, 10:41:39 AM by Anders_HK » Logged
stamper
Sr. Member
****
Offline Offline

Posts: 2515


« Reply #13 on: January 02, 2008, 12:59:50 PM »
ReplyReply

Quote
Who Dan?? Me?? - No single person is perfect, nor should it be said that following the flow of most is correct, right?

I read Dan's book on Lab early last year, very good book and indeed I should go back to it. The method of moving end points of A and B channels inwards do give a punch. He also has Professional Photoshop Ed 5, dec 2006. Is that one worthwhile for what I seek?

My arrival today was Adobe Photoshop CS3 for Photographers by Martin Evening; I have alredy started read and it seems like excellent book to thouroughly teach me details on Photoshop. Yet it seems vaguely to touch on what I seek in this post.

Is there someone who has written specific on the color sensations similar to photographic impressions of film in digital? Granted it is not only color; the CS3 command for Shadow/Highlight is jaw dropping in what it can do. Yet... I still feel captures from digital appear pale in comparison to Fuji Slides.

Much thanks for further advises!  

Regards
Anders
[a href=\"index.php?act=findpost&pid=164572\"][{POST_SNAPBACK}][/a]

Photoshop color correction  by Michael Kieran is good and easier to follow than Dan's books Don't forget layer blend modes The key to good color is contrast and saturation together On the flip side highly saturated images aren't always the best
« Last Edit: January 02, 2008, 01:01:01 PM by stamper » Logged

Anders_HK
Sr. Member
****
Offline Offline

Posts: 1001



WWW
« Reply #14 on: January 02, 2008, 04:03:31 PM »
ReplyReply

Quote
Photoshop color correction  by Michael Kieran is good and easier to follow than Dan's books Don't forget layer blend modes The key to good color is contrast and saturation together On the flip side highly saturated images aren't always the best
[a href=\"index.php?act=findpost&pid=164619\"][{POST_SNAPBACK}][/a]

Hi Stamper

You may have said the magic words: "Color Correction".

That book sounds very good but from 2002 and for Photoshop 7.0. Although basics perhaps are same, is it still very applicable to CS3?

I also found following titles:

Digital Color Correction
by Pete Rivard March 2006

Color Correction For Digital Photographers Only
by Ted Padova, Don Mason June 2006

Thanks for advise.

Regards
Anders
Logged
Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1139



WWW
« Reply #15 on: January 02, 2008, 11:47:12 PM »
ReplyReply

What you consider the look of film from what you've posted seems to not only involve increasing saturation but also injecting a color crossover effect between highlite white balance and shadow neutrality where whites are warmish and shadows are coolish. Pick complementary colors like yellowish whites and purplish shadows or orangish whites and bluish shadows.

This is an old renaissance master painter eyebrain trick to make an image "Pop". Study Maxfield Parish paintings to get an idea where film manufacturers came up with their film color palette.

Have you tried just using curves to do this in RGB. Working in Lab would be quite difficult adding apposing highlite/shadow color temps because the center of the curves controls neutrality for both of these regions.
« Last Edit: January 02, 2008, 11:55:40 PM by tlooknbill » Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #16 on: January 03, 2008, 09:25:33 AM »
ReplyReply

Using the profiling method I described earlier will very closely duplicate any film look (including contrast changes, color crossovers, etc) with only 2 menu clicks in Photoshop, easily incorporated into a batched action. All of the other methods are going to be much more tedious and time-consuming.
Logged

Tim Lookingbill
Sr. Member
****
Offline Offline

Posts: 1139



WWW
« Reply #17 on: January 03, 2008, 03:51:28 PM »
ReplyReply

Quote
Using the profiling method I described earlier will very closely duplicate any film look (including contrast changes, color crossovers, etc) with only 2 menu clicks in Photoshop, easily incorporated into a batched action. All of the other methods are going to be much more tedious and time-consuming.
[a href=\"index.php?act=findpost&pid=164788\"][{POST_SNAPBACK}][/a]

Not doubting your method would be the most efficient but correct me if I'm wrong but every color calibration target I've come across has been optimized for accuracy at the factory to NOT embue the color crossover inaccuracies many seem to love about film. I mean I have to wonder where this look of film originally comes from since its history shows that it can go through so many processing variables from the age, stock, brand of the film to whatever crackerjack operation chemically processed it.

I never saw any color crossover effects in an it8 target. Maybe there's other targets I'm not aware of that have this. They all have grayramps that are quite neutral.

After looking at restored motion pictures shot on negatives from the '60's and '70's ("Tommy" and "Blowup" on TCM comes to mind) I was surprised to find film can be made to capture a scene quite accurately but still look pleasing which I thought was impossible since I originally saw these films on crappy prints in reruns on TV. These restored films did not show these types of color characteristics posted here.

Do you have or can you link to any sample images where this type of calibration was performed on a DSLR to induce these kinds of color characteristics?
Logged
pfigen
Sr. Member
****
Offline Offline

Posts: 450


WWW
« Reply #18 on: January 03, 2008, 07:52:14 PM »
ReplyReply

"One approach would be to use the camera profiling function of Eye-One Match. You'll need to shoot the test chart with the film you wish to emulate, scan the film using the same output RGB space (ProPhoto, Adobe RGB, or whatever) as you use for your RAW output RGB space. Make the film profile from the scan in Eye-One, and save it as Velvia.icc or whatever.

Now all you need to do is convert your RAW captures to the film profile, then assign your working RGB space to the file, and you'll have exactly the same color "look" as the film."

Hmmm... All you have to do is convert to the velvia profile. Scanner and Camera profiles from Gretag, which are essentially the same, are one way profiles. You can only convert from them, not to them. Have you actually done this or is just theory because I can only assign camera or scanner profiles to an existing image. Now, assigning my real Velvia profile, generated from a custom Hutchcolor target on my Howtek drum scanner, and made in PM5, does indeed change the character of the image, but not always for the better. Even if this would actually work, it's a one size fits all approach for a multi-faceted world. Curves in RGB and Lab color work more effectively on a image by image basis, but everyone has their own way of doing things and that keeps it interesting.
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #19 on: January 03, 2008, 09:28:12 PM »
ReplyReply

Quote
Hmmm... All you have to do is convert to the velvia profile. Scanner and Camera profiles from Gretag, which are essentially the same, are one way profiles. You can only convert from them, not to them. Have you actually done this or is just theory because I can only assign camera or scanner profiles to an existing image.

Eye-One camera camera profiles are basically a color transform from film color to calibrated accurate color, using the scanner's output space as a baseline. If your scanner outputs ProPhoto and you used a Velvia scan to build the profile, when you assign the Velvia profile to a Velvia scan and then convert to ProPhoto, you have converted Velvia's color palette to calibrated, accurate color. This is of course the opposite of what the OP wants.

Quote
Scanner and Camera profiles from Gretag, which are essentially the same, are one way profiles. You can only convert from them, not to them.

False. Just because that's the standard, "intended" use for such a profile doesn't mean it's the only way to use it. If you take a calibrated-color ProPhoto image, convert it to the Velvia profile, and then assign ProPhoto to the image (instead of assign, then convert), you have reversed the effect of the profile so that calibrated, accurate colors are converted to Velvia's color palette and tone curve, instead of the other way around.

You can expand this technique to give a shot made with one film the look of a different film (except for the grain pattern, of course). Let's say you have made profiles for your scanner and Reala and Velvia films, using ProPhoto as your scanner output space. You have an Reala shot but you wished you had used Velvia instead. Here's how you convert to the color look of the other film:

1: Assign the Reala profile to the Reala shot, which should be tagged with ProPhoto. Without changing the RGB values, this step redefines the meaning of the RGB color values so that they display as accurate, calibrated color.

2: Convert the image to the Velvia profile. This will alter the RGB values so that the color channel values are exaggerated, but simultaneously redefine the meanings of those RGB values so that accurate colors are still displayed.

3: Assign ProPhoto to the image. It will now have Velvia's color palette and contrast curve.
« Last Edit: January 03, 2008, 09:42:56 PM by Jonathan Wienke » Logged

Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad