Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: 10 bit workflow  (Read 5859 times)
technoholic
Newbie
*
Offline Offline

Posts: 9


« on: March 11, 2013, 12:25:08 PM »
ReplyReply

After struggling for a long time to get acceptable ink jet prints, I resolved to get it right once and for all.  An Xrite Colormunki is in use to profile my monitor and Epson R3000 printer. Still, prints do not match the monitor and some prints are way off only to end up in the trash. All the research I have done lately points to the monitor as the culprit.

After a lot of research, I have decided to upgrade to a NEC PA271W-BK-SV monitor. Since this monitor is 10 bit capable, as is much of my workflow, it seemed to make sense to make it all 10 bit. Thats when I hit a wall, namely Lightroom 4 which is not 10 bit capable and cannot even read a 10 bit psd file. As such, Lightroom 4 is useless, even counterproductive in a 10 bit workflow. After investing heavily in Lightroom lately, I would hate to abandon it. Hopefully, Lightroom 5 will support 10 bit color. I doubt that I am the first person to encounter this dilemma so how do you structure your 10 bit workflow?
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #1 on: March 11, 2013, 12:31:43 PM »
ReplyReply

Bit depth is not the cause of your problem. There is tons of material on this website about colour-managing your display and your printer to get predictable results. I recommend you do some research and reading - the LULA Camera to Print tutorial is probably the best overall resource.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
technoholic
Newbie
*
Offline Offline

Posts: 9


« Reply #2 on: March 11, 2013, 01:50:35 PM »
ReplyReply

Guess I wasn't clear in that I didn't mean to imply that bit depth was the root cause. The intent is to get the best possible end result, the print. So if 10 bit color is more accurate that 8 bit, why would one not use it? Also considering that I shoot in RAW and want to work in at least Adobe RGB or ProPhotoRGB, 8 bit post processing is really not the way to go. Does 8bits not limit you to an sRGB color space? And if one wants to exploit 10 bit color, obviously the system needs to support that end to end. I could create a separate repository for 10 bit images and process them only with PS CS6, using Lightroom for less critical work but that adds another layer of complexity. Guess I'm hoping for a more elegant solution. 
Logged
new_haven
Jr. Member
**
Offline Offline

Posts: 65


« Reply #3 on: March 11, 2013, 01:56:33 PM »
ReplyReply

Take a look at this thread -

http://photo.net/digital-darkroom-forum/00bQ0n
Logged
Rory
Full Member
***
Offline Offline

Posts: 195


« Reply #4 on: March 11, 2013, 01:59:08 PM »
ReplyReply

While Lightroom does not directly support 10 bit color, it does dither the results.  Not perfect, but often hard to tell...
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #5 on: March 11, 2013, 02:10:13 PM »
ReplyReply

Guess I wasn't clear in that I didn't mean to imply that bit depth was the root cause. The intent is to get the best possible end result, the print. So if 10 bit color is more accurate that 8 bit, why would one not use it? Also considering that I shoot in RAW and want to work in at least Adobe RGB or ProPhotoRGB, 8 bit post processing is really not the way to go. Does 8bits not limit you to an sRGB color space? And if one wants to exploit 10 bit color, obviously the system needs to support that end to end. I could create a separate repository for 10 bit images and process them only with PS CS6, using Lightroom for less critical work but that adds another layer of complexity. Guess I'm hoping for a more elegant solution. 

Don't process in 8-bit depth for ProPhoto or similarly wide colour spaces. Apart from that, this is not something to be concerned about using Lightroom or Photoshop. Best practice for print is to use a wide colour space and 16-bit depth. The applications will look after bit-depth differences under the hood, and do a good job of it.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
technoholic
Newbie
*
Offline Offline

Posts: 9


« Reply #6 on: March 11, 2013, 02:44:38 PM »
ReplyReply


Thanks for the pointer, interesting conversation with some good news. It does bring up another question pertaining to bit depth nomenclature. While monitors usually are spec'd as 8 bit or 10 bit, the image mode in CS6 gives 8 bit, 16 bit and 32 bit. Can you help me understand how that maps to the monitor 8/10 bit spec? I understand the monitor spec is a per channel value which results in 24bit and 30 bit color, respectively.

Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #7 on: March 11, 2013, 02:50:01 PM »
ReplyReply

Thats when I hit a wall, namely Lightroom 4 which is not 10 bit capable and cannot even read a 10 bit psd file. As such, Lightroom 4 is useless, even counterproductive in a 10 bit workflow.

Uh...there is no way to support a "10-bit" PSD file...processed files are either 24-bit (3x 8-bit/channel) or 48-bit (3x 16-bit/channel).

You are confusing the image processing bit depth with the display pipeline bit depth. If you are using SpectraView with your NEC, the display is using 10-bits/channel in the internal calibration. That's as close as you are gonna get. The only app that I know that supports a 10-bit display pipeline is Windows Photoshop CS5/CS6 with an appropriate video card (not all do). And if you are on Mac, forget about 10-bit display pipeline because Apple hasn't seen fit to add it.

As for your prints not matching what you think they should be, that's a breakdown of your color management, not the display pipeline. You should be looking other areas in your workflow to fix, like the profile qualities and learning how to soft proof.
Logged
Rhossydd
Sr. Member
****
Offline Offline

Posts: 2001


WWW
« Reply #8 on: March 11, 2013, 03:20:36 PM »
ReplyReply

As for your prints not matching what you think they should be, that's a breakdown of your color management, not the display pipeline.
Not necessarily. A bad monitor is often a cause for people to think their prints are poor, sometimes when the prints are absolutely correct.
From the OP it appears that the intention is to upgrade the existing monitor, not that he's actually using a Spectraview now.

An in-depth understanding of colour management and access to software that helps understand gamut limitations can allow some people to work round poor hardware, but it doesn't seem this is the case here.
Logged
technoholic
Newbie
*
Offline Offline

Posts: 9


« Reply #9 on: March 11, 2013, 03:45:28 PM »
ReplyReply

Uh...there is no way to support a "10-bit" PSD file...processed files are either 24-bit (3x 8-bit/channel) or 48-bit (3x 16-bit/channel).

You are confusing the image processing bit depth with the display pipeline bit depth. If you are using SpectraView with your NEC, the display is using 10-bits/channel in the internal calibration. That's as close as you are gonna get. The only app that I know that supports a 10-bit display pipeline is Windows Photoshop CS5/CS6 with an appropriate video card (not all do). And if you are on Mac, forget about 10-bit display pipeline because Apple hasn't seen fit to add it.

As for your prints not matching what you think they should be, that's a breakdown of your color management, not the display pipeline. You should be looking other areas in your workflow to fix, like the profile qualities and learning how to soft proof.

Good info, thank you. It looks then like processing bit depth and display pipeline bit depth do not map 1 to 1. As to why, I will ponder that another day.

As for now, I agree there is a breakdown in my color management and I do believe it is with the monitor. Everything I have read states that if your monitor is not accurate, profiled or not, your prints wont be either. Soft proofing has been frustrating. I have been muddling along with an inexpensive 8bit, TN panel, non calibrated but profiled monitor. Profiling has helped but was not what was hoped for. It looks good but then I have no frame of reference to judge by. I have finally  decided to bite the bullet and upgrade to a professional grade monitor and video adapter and I want that upgrade to be top notch, hence the questions about bit depth, something I have not been concerned with until lately. Once I have the proper tools to work with, fine tuning should be a lot less frustrating.
Logged
Jim Kasson
Sr. Member
****
Offline Offline

Posts: 1078



WWW
« Reply #10 on: March 11, 2013, 03:52:42 PM »
ReplyReply

A couple of years ago, I went to 10-bit (or 30-bit, depending on whether you're in marketing or engineering) color. Here's what I said then.

I'm still using displays and graphics adapters with 10-bit color, but I almost never turn it on. The differences between a 10-bit and an 8-bit display pipeline are, for my eyes and images, somewhere between extremely subtle and non-existent. Certainly the differences are infinitesimal compared to the differences from soft proofing to the real print in even a perfectly calibrated color managed workflow.

Jim
Logged

technoholic
Newbie
*
Offline Offline

Posts: 9


« Reply #11 on: March 11, 2013, 03:54:22 PM »
ReplyReply

Not necessarily. A bad monitor is often a cause for people to think their prints are poor, sometimes when the prints are absolutely correct.
From the OP it appears that the intention is to upgrade the existing monitor, not that he's actually using a Spectraview now.

An in-depth understanding of colour management and access to software that helps understand gamut limitations can allow some people to work round poor hardware, but it doesn't seem this is the case here.

I agree, a bad monitor by any other name is still a bad monitor. After many hours of research, I am convinced that is the issue here. I thought I had a pretty good grasp on color management but delving deeper into the subject just opens my eyes wider.

I respect the expertise represented in this forum, that's why I'm here. I appreciate everyone's time, patience and input.
Logged
Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #12 on: March 11, 2013, 05:11:06 PM »
ReplyReply

The differences between a 10-bit and an 8-bit display pipeline are, for my eyes and images, somewhere between extremely subtle and non-existent.

Hum...not my experience...if you use a display that has an internal 10-bit DAC (like the NEC or EIZO) the calibration portion of the profile generation produces a more accurate, less banded image display. If you start with an-bit and apply white point and gamma correction, you end up with a profiled display of less than 8-bit/channel which can lead to banding in the shadows (in gradations particularly) and less accurate color display. The displays that use an internal 10-bit/channel calibration produce a superior display and in the case of NEC, if you use a wide gamut display you get about 98% of Adobe RGB. If you are doing digital imaging and printing, this is really useful.

Bit depth comes into play in four stages, the original capture input, the working space bit depth, the display pipeline and the print pipeline. Of these, really only the first two are user controllable...on Mac, there is no 10-bit display pipeline but there is a 16 bit print pipeline. On Windows, there's a 10-bit display pipeline (usable in Photoshop) but no 16-bit print pipeline. Of all of these, it's really the original capture and the processing in the working space that has the most impact on the image quality.

As far as the value of computer displays, there are really two camps, the low end stuff which is very price sensitive and generally poorer quality and higher ned component specifications like the NEC and EIZO. TO a large extent, you tend to get what you are willing to pay for and a really good display is important if you care about optimizing your images and getting the best final output.

Logged
Rhossydd
Sr. Member
****
Offline Offline

Posts: 2001


WWW
« Reply #13 on: March 11, 2013, 05:15:03 PM »
ReplyReply

After many hours of research, I am convinced that is the issue here
Your proposed choice of the PA271 is pretty much the 'gold standard' here.
As others have suggested, worrying about a 10bit video path is probably not worth the effort. Spend any money assigned to that idea on a decent and reliable print viewing solution instead, just something cheap and simple like a Graflite(http://shop.colourconfidence.com/section.php/10390/1/grafilite) will help provide a known viewing condition you can rely on.
Then you'll be better placed to pull everything together to a good working system.

Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #14 on: March 11, 2013, 05:21:51 PM »
ReplyReply

The product from that website is not usable in North America and after drilling down three screens I still can't find out what the colour temperature of the illumination is. For those of us in North America, Solux (Tailored Lighting, Rochester, NY) would be the preferred solution. But yes, you are right to point out the relevance of the viewing condition. In the case of the OP, however, I think Jeff is likely correct that the problem is lodged somewhere in Colour Management between display and printer, and soft-proofing.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
Jim Kasson
Sr. Member
****
Offline Offline

Posts: 1078



WWW
« Reply #15 on: March 11, 2013, 05:29:50 PM »
ReplyReply

Hum...not my experience...if you use a display that has an internal 10-bit DAC (like the NEC or EIZO) the calibration portion of the profile generation produces a more accurate, less banded image display.

<snip>

As far as the value of computer displays, there are really two camps, the low end stuff which is very price sensitive and generally poorer quality and higher ned component specifications like the NEC and EIZO. TO a large extent, you tend to get what you are willing to pay for and a really good display is important if you care about optimizing your images and getting the best final output.

Jeff,

I'm using the NEC PA301W and the ATI (Now AMD) FirePro 4800. Could be my gear, could be my eyes. Do either Photoshop or Lightroom dither the display?

Jim
Logged

Schewe
Sr. Member
****
Offline Offline

Posts: 5542


WWW
« Reply #16 on: March 11, 2013, 05:37:51 PM »
ReplyReply

For those of us in North America, Solux (Tailored Lighting, Rochester, NY) would be the preferred solution.

Yeah, Martin Evening has a GrafLite small viewing light I liked and I could never find one here–pretty sure they are all 220V. I agree with Solux for task lamps but I also use a GTI digitally dim-able viewing booth. However, I re-lamped my desktop booth with D65 bulbs instead of D50 bulbs (that it came with). I use the GTI plus a separate Solux as more of a warm daylight simulator. The GTI's are kinda expensive so if on a budget, a Solux bulb in an appropriate lamp with a neutral background would work...
Logged
jrsforums
Sr. Member
****
Offline Offline

Posts: 750


« Reply #17 on: March 11, 2013, 08:47:38 PM »
ReplyReply

The product from that website is not usable in North America and after drilling down three screens I still can't find out what the colour temperature of the illumination is. For those of us in North America, Solux (Tailored Lighting, Rochester, NY) would be the preferred solution. But yes, you are right to point out the relevance of the viewing condition. In the case of the OP, however, I think Jeff is likely correct that the problem is lodged somewhere in Colour Management between display and printer, and soft-proofing.

Looks as Graflite uses Ottlite technology
http://www.grafilite.com/product_info.html
Logged

John
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7121


WWW
« Reply #18 on: March 11, 2013, 09:06:52 PM »
ReplyReply

Yes, does look like that. Thanks.
Logged

Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....." http://www.luminous-landscape.com/reviews/film/scanning_workflows_with_silverfast_8.shtml
technoholic
Newbie
*
Offline Offline

Posts: 9


« Reply #19 on: March 11, 2013, 11:04:00 PM »
ReplyReply

I do have an Ott Light, it looks very much like the the GrafiLite. Is there a noticeable difference between the two?
Logged
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad