Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: Calibration devices + software  (Read 11943 times)
Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« on: April 08, 2006, 05:25:56 PM »
ReplyReply

I've just bought my first LCD monitor for photo editing, it is the Apple Cinema Display 20" which I'm running on a PC using the WINACD driver on a Nvidia 6200 card. So far so good.

My problem is that it is too bright. With my CRT I would profile using the Spyder 1 or 2 then turn the brightness down to match print brighness using a test print. This worked fine and I was getting a perfect screen to print match both in color and brightness. I use ACR/bridge for the vast majority of my work rarely having to go into PS and need the screen to be accurate for print which it was with the CRT.

After calibrating with the spyder 2 on my ACD, I find that I cannot turn the brightness down far enough and even using the video card software to lower the brightness (to 90%) so that my test print matches on screen, although the facial tones are OK the shadows are still far too open and show color shifts (red/green). It is possible that this is color inherent in the shadows from color photographs but should be too dark to notice but the point is that I want to get a decent result.

To be able to view pictures on screen at print brighness is crucial for my workflow when working 700 pics at a time for print. With my CRT I had no problem, I would calibrate, wind down and get perfect print/screen matches. The LCD seems to be nowhere near, not in the shadows, not in the contrast and to be perfectly frank, not in much else either.

Is there a device and software which allows calibration based on a set brighness level so that I can achieve my goal? i.e. I tell the software how dark the screen must be after calibration and the calibration works according to that? If such a thing exists it should give me good results throughout. I cannot be the only one who wants to process based on print brightness without having to work inside CS2 with 'proof viewing' enabled, what do the ACR/lightroom/aperture etc users do? The calibrated brightness of the screen without tweaking is just not accurate relative to the prints for brightness, never was even with my CRT or anyone's screen I ever saw.

I bought the ACD as it was supposed to be the best mid priced screen available for photo editing. So far my aging non descript Compaq 17" CRT is beating it hands down for color and brightness control. I can't deny that this is not making me happy and am praying that it is a calibration issue and not that what I would like to achieve is impossible.

Help?
« Last Edit: April 08, 2006, 05:26:40 PM by pom » Logged

Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« Reply #1 on: April 08, 2006, 05:43:20 PM »
ReplyReply

Having looked at B&W images, the color shifts are inherent to the photos, the blacks are being pushed too hard and showing up the noise and color that was there but shouldn't be visible at this level of brightness. Can't tell you how much this is getting on my nerves!
Logged

61Dynamic
Sr. Member
****
Offline Offline

Posts: 1442


WWW
« Reply #2 on: April 08, 2006, 10:33:32 PM »
ReplyReply

Quote
It is possible that this is color inherent in the shadows from color photographs but should be too dark to notice but the point is that I want to get a decent result.
The color shifting is probably the result of your adjustments to the video card. You are altering the 24-bit image before it reaches the display. It is the equivalent of altering the RGB controls on any other LCD display. The LUTs get changed, and the image on screen degrades.

Does the color shifting occur with the video card at defaults?

I'd say your main problem is that of practice. Previously with a CRT you would match the CRT to the luminosity of the room for accurate print viewing. This was possible since a CRT has separate brightness (black-point) and contrast (white-point) controls allowing a proper calibration. A LCD does not do this. Both the whites and blacks are adjusted in one fail swoop via the brightness. in addition, LCDs are by nature brighter than a CRT will be. As I've mentioned before in other threads, the optimal compromise between whites and blacks requires the LCD to be around 120cd/m2 (verses ~95cd/m2 for a CRT). Due to that, your previous techniques for soft-proofing will not work with a LCD.

The best solution to getting around that for print viewing is to not set the monitor to match the print or even the ambient light to match the display but instead to get a viewing booth with an adjustable dimmer switch. This way you can turn up the brightens of the booth to match that of the LCD. This allows the room to maintain the proper luminosity for the display to render it's best yet have the print bright enough to match the display.

Quote
Is there a device and software which allows calibration based on a set brighness level so that I can achieve my goal? i.e. I tell the software how dark the screen must be after calibration and the calibration works according to that
ColorEyes does that but if the difference is too great, the profiles turn out very gnarly. Gnarly as in unusable. You can download a demo from their site but be warned. Having the profile compensate for luminosity will result in similar effects as altering the luminosity in the video card controls.
Logged
Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« Reply #3 on: April 09, 2006, 06:49:01 AM »
ReplyReply

Thanks Daniel, this was exactly what I was seeing, when the CRT was adjuted the whites stayed white, the blacks were as dark as I wanted them. As you've said this isn't possible with the LCD and I have been seeing the 'gnarly' effect in a  rather stepped approach to facial tones, far less 'analogue' in look and kinda nasty.

Viewing the print in brighter light isn't going to help. I took my test print and using the info palette of PS discovered that the shadow that needs to be totally blocked to match print brightness corresponds to an RGB value of 27,27,27 i.e. anything darker than this won't show up in print. I can darken the screen to reflect this using the video card but the shadows are still far too open on almost every other photo where the shadow has some color.

Even if I set the screen for the optimal 120 that you suggest, and I don't mind doing that, the screen is showing the shadows at far brighter than they can be on print under any lighting conditions and i have to print for 'normal' conditions, i.e. what the clients will view them under, I can't ask them always to view them outdoors at mid day.

I know for those who do all their editing in PS it would be possible to do all the work using the soft proofing for the printers profile. Fine.
What about all of us who are not editing in an enviroment that allows soft proofing, such as bridge/ACR or lightroom/aperture? How the heck can I adjust the brightness/shadows of 700 photos from a wedding in ACR to be batched to jpg's ready for print, when what I see is just not accurate?

Should I go back to the CRT? am I the only one trying to get a WYSIWYG solution outside of PS?

Is there any solution for CRT like control, i.e. contrast and brightness seperate so I can get the look I want, or should I give up on LCD for photo editing?

BTW is there any shareware that shows what the brighness of the screen is, so I can adjust for 120? The Spyder 2 doesn't show any custom options and I'm loathe to buy another calibrator until I'm sure that it will be better. Coloreyes unfortunately does not support the spyder.
Logged

Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7126


WWW
« Reply #4 on: April 09, 2006, 10:20:29 AM »
ReplyReply

Pom, I'm very interested in your experience with this because my CRT, a Trinitron, which I really like very much, is showing signs that it may simply konk-out one day, and then I shall be left with no choice but to buy an LCD, because there simply are no new CRTs on the market that deliver the quality needed for critical image processing. So I have been doing my homework in preparation.

From all that I've read and seen, it seems that at this moment the two contenders in the monitor market for up to USD 2000 or so are the Apple Cinema displays and the LaCie 321G (or a slightly smaller, similar device from NEC). The LaCie is supposed to have more bit depth (10-10-10), but I haven't done a close comparison under dim ambient light to see what difference it makes to the viewer. Both of them are too bright unless adjusted downward.

When Integrated Color Corporation developed their software (ColorEyes Display) which amongst other things features a software-determined setting for Luminance (called L*), they canvassed the market for the hardware device they considered optimal for the purposes of their software and they settled on the Monaco Optix XR/Xrite DPT-94 (same thing). I use this set-up for calibrating and profiling my CRT, and I find it on the whole fine, eventhough rendering deep shadow detail on matte paper will remain somewhat of a challenge for some images no matter what. I've heard that it works equally well on LCD monitors, but I have no direct experience with this option from which to report.

Considering the volume of work you are doing, it may be worth your while investing in this profiling/calibration combo to see whether it helps solve the problem. At the same time, however, I wonder whether it was just by chance that you were getting such well-matched results with your CRT by comparing images in Adobe Bridge or ACR with the printed output not using Soft Proof. I find I must use Soft Proof image by image for generally reliable, predictable output.
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
marc.s
Jr. Member
**
Offline Offline

Posts: 91


« Reply #5 on: April 09, 2006, 10:47:04 AM »
ReplyReply

If he can do it consistently with his old monitor he should be able to do it consistently with his new monitor.

I don't use softproof to check either, except with very saturated reds where I sometimes need to check and adjust. I've set up some default adjustment layers that I add to the files just before sending them in for printing and the results are very consistent (it took some meddling about to get those adjustment layers worked out though).
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7126


WWW
« Reply #6 on: April 09, 2006, 11:17:17 AM »
ReplyReply

Quote
If he can do it consistently with his old monitor he should be able to do it consistently with his new monitor.
[a href=\"index.php?act=findpost&pid=62209\"][{POST_SNAPBACK}][/a]


I think the problem he brought to our attention is that he is getting CONSISTENTLY misleading information from his monitor for the reasons stated. This appears to be not so much an issue of consistency from image to image, but one of predictability between what shows on the monitor and what comes out of the printer, particularly in the shadow areas.

Your use of "default adjustment layers" indicates that your system is not properly colour-managed, so you are approximating the results of a colour-managed workflow through colour/luminosity adjustment rather than by colour management. Fine if it works for you most of the time and you are satisfied with it, but it is bound to be less generally reliable accross a broad range of images and papers than would be a properly colour-managed system used with 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
marc.s
Jr. Member
**
Offline Offline

Posts: 91


« Reply #7 on: April 09, 2006, 11:47:08 AM »
ReplyReply

I know that my setup is far from the way it's 'supposed' to be from a properly colour managed system. But it works for me. My point was simply that I'm able to send my images to the lab without softproofing in photoshop (save for a few rare cases that I can spot in advance). So if he can tell what his images will look like without softproofing on his CRT setup (as can I on my CRT setup) then he should be able to tell what his images will look like without softproofing on an LCD setup.

I understand he has some different issues (that sound like they're fixable to me) setting up his LCD now. I was simply responding to the softproofing thing..
Logged
Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« Reply #8 on: April 09, 2006, 11:53:47 AM »
ReplyReply

I print using one system only and that is on a Fuji Frontier with Fuji Crystal Archive paper using three seperate labs. I've found that a properly calibrated screen will give print results that are 97% color accurate (I assume the differences are due to chemistry, etc) but are darker by quite a significant amount. This accuracy is no doubt due to built in calibration on the frontier for accurate results, I know the labs run color tests twice a day on each paper surface/size used which is then digitally analysed for calibration.
 I don't and never have had to soft proof given that I have gotten a true WYSIWYG solution provided I darken the screen to match the print brightness. This was fine with my CRT, my desktop image is a target picture that I use, when a certain shadow is blocked up completely then my screen is 100% accurate both in color and brightness. I don't really doubt at all that this would also be true of any decent calibrated CRT.

With the LCD calibrated with the spyder 2 I've found that the screen is too bright even at the lowest brightness setting, the shadows are far too open, the highlights are blown on screen even when according to the histogram, the same image on the CRT screen and the print, there is plenty detail. It's the equivelent of saying there just isn't the same DR on the LCD screen or even the difference in highlights between digital and neg film!
The colors are, though technically accurate, rather stepped, the tonality is too sudden, not graduated sufficiently and contrast adjustments are shown far too exaggerated. That is before I've adjusted the video card software!
Even with a custom curve using the software as opposed to their brightness slider, the color and black/white graduations are too sudden, not anywhere as smooth as they looked on the CRT which incidently matched the print perfectly for tonality and graduation.

It has got to the point where a 2nd hand 21" Dell CRT (with 3 month warranty) which can be calibrated with my present equipment for perfect results is going to cost me another 25 pounds while the cost of this ACD plus a Monaco Pro calibration system will rack me up to approx 800 pounds so far and I still won't have what I had with my CRT, or at least I doubt it. I chose the ACD as I had assumed that it would be better than my CRT for photo editing and the best within the price range. Now I find that a CRT is cheaper and better without needing a lot of expense and heartache to calibrate, so to hell with saving desk space. It's the results that count and at present I do not feel that I can trust my screen to give me an accurate representation of the photos I am editing. That is a horrible feeling for a busy digital photographer especially one who has just spent more on his new screen than on upgrading his entire system. Yes I am proficient enough to work based on the histogram alone for highlights/shadows but I'll be damned if I'm going to have spent that much on a screen just to have to work by the numbers.

The problem, and one of the reasons why I decided to try the LCD is that CRT's are seriously on the way out. Finding a good monitor is getting harder and harder and they are technically obsolete. Will the new technology coming in the next few years prove to be more versatile for photographers than the present crop of super bright and over sharp/contrasty LCD's while being affordable and will the CRT's last that long?

Anyone want an Apple Cinema Display 20" in good condition for 390 pounds?

Anyone have a sony artisan for sale (I wish!)?  

Seriously though, any suggestions for which CRT's from the good era are considered worthwhile? might be worth trying to look for a lightly used one in good condition.
« Last Edit: April 09, 2006, 12:00:32 PM by pom » Logged

Dale_Cotton
Sr. Member
****
Offline Offline

Posts: 580


WWW
« Reply #9 on: April 09, 2006, 12:12:44 PM »
ReplyReply

Pom: what happens if you increase the brightness of the lighting of the room in which use the LCD? I have zero experience with LCDs, but I find that even with my profiled CRT, the shadows don't make any sense compared to the print, unless my ambient matches the brightness that i1 Display wants.
Logged
marc.s
Jr. Member
**
Offline Offline

Posts: 91


« Reply #10 on: April 09, 2006, 12:55:35 PM »
ReplyReply

According to Shootsmarter these CRTs are good:

http://www.shootsmarter.com/infocenter/wc033.html

But apparently there are mixed opinions on the expertise of that website. It's a starting point at least. I could only find one of those monitors here and it was rather expensive, so I'm probably going with the NEC 2090uxi. If I can't make it look good I'll return it.
Logged
61Dynamic
Sr. Member
****
Offline Offline

Posts: 1442


WWW
« Reply #11 on: April 09, 2006, 01:06:26 PM »
ReplyReply

Quote
With the LCD calibrated with the spyder 2 I've found that the screen is too bright even at the lowest brightness setting, the shadows are far too open, the highlights are blown on screen even when according to the histogram, the same image on the CRT screen and the print, there is plenty detail. It's the equivelent of saying there just isn't the same DR on the LCD screen or even the difference in highlights between digital and neg film!
That's a bit curious. It sound's like something is off-kilter either in the video card software or the Spyder calibration. I know based off DryCreek's calibrator reviews and my own experience with the first model of Spyder that it is not entirely accurate in calibrating highlights (it doesn't even measure them) but that should not be resulting in a loss of detail. What model of Spyder 2 do you have?

If I had to take a guess, it sound's like luminosity is being jacked up somewhere in software. Try opening up a grayscale test image in both PS and a non-color managed app (such as Windows Picture Viewer) and see if it appears equally blown out in both apps. I not, then I'd say it's an issue with the Spyder's profiling. If they appear the same then it may be an issue in the video card driver in which case, double check all the settings in the driver to make sure there is not adjustments occurring or uninstall and then reinstall the driver.

Oh, just thought of this. Are these images you have adjusted previously with the CRT or are they new images? I ask because it could be that the images are in fact too bright and your previous method of calibration could of hid that from you. Assuming you don't go for the economy printing, the lab could be adjusting the luminosity of the print down automatically and thus the appearance of a perfectly working system. Of course if you don't have any of your labs do corrections (print as is) then this theory goes down the tubes.
Logged
Mark D Segal
Contributor
Sr. Member
*
Offline Offline

Posts: 7126


WWW
« Reply #12 on: April 09, 2006, 01:11:48 PM »
ReplyReply

Pom if I remember correctly from what I've read, DR is more of an issue -generically - with LCDs than it is with CRTs.
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
Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« Reply #13 on: April 09, 2006, 01:37:31 PM »
ReplyReply

I'm using the images from 2 weddings and a barmitzva shot recently and adjusted using the CRT. I've had the prints recently and so they are fresh in my memory.

I'm still playing, having worked out how to calibrate each monitor seperately using the spyder software that ostensibly doesn't allow calibration of multiple monitors (unplug each in turn, calibrate and save profile then load them up seperately into the display properties for each video card - yawn!). Now I have my CRT calibrated using the same software and showing perfectly accurate results, next to the LCD also calibrated. This helps as I can have the same image on both screens and view the differences.

OK, basically I can reduce the contrast in the highlights on the LCD to get facial tones looking the same though the CRT is about 150K warmer and TBH more accurate to print. I've had to reduce the contrast in the blacks as well as darkening them but am having considerable difficulty getting the blacks back to neutral, they have an awful and unnatural green cast, unfortunately it's difficult to reduce this in the video cards curves menu without affecting the image globally to some extent, I wish I could show you the difference but other than taking photos of the two screens which I don't have time for, it would be difficult! The CRT has beutifully neutral blacks.

A kindly soul is lending me a Monaco so I will hopefully manage to run another calibration which will have better and more neutral blacks. The idea of having to fork out that kind of money for a calibration device just to find out it's as useless was seriously troubling me, this way I can at least trial it before handing over the money. Having the calibrated CRT next to me for reference is heaven when trying to tweak the LCD!

I have a glimmer of hope about being able to get this ACD to work for me, I'll keep you posted. BTW Marc.S, if I do get it right then it might be worth you hanging on to the idea, always better when someone has done the grunt work!
Logged

marc.s
Jr. Member
**
Offline Offline

Posts: 91


« Reply #14 on: April 09, 2006, 04:23:20 PM »
ReplyReply

Pom, if you let the ACD be at its bright level (where it sounds like it wants to be) instead of turning down brightness, are the colours still off? Just wondering if it's only when you turn down brightness that everything goes haywire.

I will probably use the Spectraview software that NEC makes for their monitors because it takes advantage of its special features. The NEC uses the Eye-One puck as well so it's just a software thing there.

Of course, I may still get scared off and find some second hand CRT somewhere
Logged
Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« Reply #15 on: April 09, 2006, 05:02:34 PM »
ReplyReply

well, that was fun!

Firstly the weird colour shift to green in the shadows was a Spyder issue, totally corrected with Monaco, no shift at all. I calibrated using color eyes but prefered the monaco software version. Color eyes measured the native white point of my ACD at 7454K which suprised me somewhat, I got far better results calibrating to D65 in any case with both software solutions for the monaco.

I profiled both the CRT and LCD with the monaco using the monaco software at D65 (6500k), the CRT is still about 250K warmer and I can't seem to do anything about it which is a shame.

With the LCD once it was calibrated I had to make a custom curve using the video cards software to match the CRT for both contrast and shadow/highlight brightness but once that was done, and it took a long while, the colors were as I said within 250k of each other (I have to print up another test pic to check which is more accurate, my test pic is B&W as I was using it for print brighness only) and the shadows/highlights were near enough not to make any difference. Most importantly there is no noticeable color shift at all in the shadows in comparison to the same image on the CRT other than that caused by the WB imbalance.

So lessons learned, keep away from spyder 2, utterly useless, Monaco works and its software is accurate, you won't get an exact match with hardware between a CRT and LCD though I have to admit that the CRT is both more accurate for tonality and more pleasing to the eye and finally, if you want to match for print brighness prepare to spend time tweaking video card software curves though having your calibrated CRT up there with the same image/s on both screens is essential for matching purposes.

I'm not counting on this being the end of the road but I'm far further down it than I was 4 hours ago (with supper in between!)

Is an LCD worth it looking with hindsight? If you work with PS proof colors 100% of the time then maybe, if you like the look or the look of your desk then fine. If you have a working CRT that you are happy with then to be honest the advantages are far far outweighed by the disadvantages when you consider the cost and time consumed to get results even close to what you are used to from your CRT.

However once you have a widescreen on your desk it's going to be damn hard to give up...  
Logged

Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« Reply #16 on: April 09, 2006, 05:14:50 PM »
ReplyReply

Just found a print from the last wedding, brightness is fine, contrast still a drop high in the shadows on the LCD, the warmer CRT is more accurate, in fact give or take perfectly accurate, the LCD is too cold, I have to work out how to warm it up using the curves which unfortunately I'm no expert at!
Logged

marc.s
Jr. Member
**
Offline Offline

Posts: 91


« Reply #17 on: April 09, 2006, 05:31:18 PM »
ReplyReply

Good to hear you found the main problem! What's all this about making custom curves with the video card? Is that part of the standard procedure and detailed in the calibration guide, or is it just something you figured you have to do?
Logged
Ben Rubinstein
Sr. Member
****
Offline Offline

Posts: 1733


« Reply #18 on: April 09, 2006, 05:52:33 PM »
ReplyReply

If you want to lower the brightness more than the buttons on the screen will allow as I needed to do, you have to use the video card software. The Nvidia software allows you to either use the brightness/contrast sliders or do it yourself with a curve. The curve is far more accurate as you can control what it is doing better, i.e. darkening which shadows without dulling the whites etc.

I'm still working on getting it perfect or as near to such that it doesn't bug me, getting the LCD warmed up is pretty difficult to this curves novice, the curves dialog allows for changes to either the luminosity or RGB and I can't work out how to warm it up by approx 200K using the RGB curve without it screwing up the luminosity curve.
Logged

marc.s
Jr. Member
**
Offline Offline

Posts: 91


« Reply #19 on: April 09, 2006, 07:08:56 PM »
ReplyReply

Try contacting Nvidia support or ask in a graphics card geek forum
Logged
Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad