Luminous Landscape Forum
Raw & Post Processing, Printing => Printing: Printers, Papers and Inks => Topic started by: Mike Guilbault on January 28, 2017, 03:50:59 pm
-
I have my new NEC PA272W-BK monitor and have calibrated it several times using an i1 Display Pro sensor and the SpectraViewII software. My target settings are D65, Gamma 2.20, Intensity at 80.0 cd/m2 and the default monitor Contrast Ratio which calibrated at about 835:1,
The images look great on the monitor, but still coming out too dark - not by much, but still dark - and the contrast is really high.
Am I missing something in the calibration?
-
I think you should be shooting for a contrast ratio of somewhere around 200-300:1 instead of the 800+
-
I have my new NEC PA272W-BK monitor and have calibrated it several times using an i1 Display Pro sensor and the SpectraViewII software. My target settings are D65, Gamma 2.20, Intensity at 80.0 cd/m2 and the default monitor Contrast Ratio which calibrated at about 835:1,
The images look great on the monitor, but still coming out too dark - not by much, but still dark - and the contrast is really high.
Am I missing something in the calibration?
You are soft-proofing, right?
-
I have my new NEC PA272W-BK monitor and have calibrated it several times using an i1 Display Pro sensor and the SpectraViewII software. My target settings are D65, Gamma 2.20, Intensity at 80.0 cd/m2 and the default monitor Contrast Ratio which calibrated at about 835:1,
The images look great on the monitor, but still coming out too dark - not by much, but still dark - and the contrast is really high.
Am I missing something in the calibration?
Your surround should be white, and should be generous. Is it? Don't make the picture so big it fills the monitor. I'm assuming you have soft proofing turned on.
Jim
-
I have my new NEC PA272W-BK monitor and have calibrated it several times using an i1 Display Pro sensor and the SpectraViewII software. My target settings are D65, Gamma 2.20, Intensity at 80.0 cd/m2 and the default monitor Contrast Ratio which calibrated at about 835:1,
The images look great on the monitor, but still coming out too dark - not by much, but still dark - and the contrast is really high.
Am I missing something in the calibration?
Monitor profile settings doesn't affect printing at all except if you make adjustments to the image based on what you see in the monitor. For a good overview of solutions on this often asked question see "Why are my prints too dark" at Andrew Rodney's site:
http://www.digitaldog.net/
-
I agree w/ Mr. Kasson. I used to have the surround the usual fairly dark grey but I've moved over to white or the lightest grey possible. This prevents me from seeing into the shadows as much as I would w/ a dark surround and forces me to bump up w/ Shadows/Blacks which I always seem to need (esp. w/ non-linear Epson behavior in the very dark tones).
-
I agree w/ Mr. Kasson. I used to have the surround the usual fairly dark grey but I've moved over to white or the lightest grey possible. This prevents me from seeing into the shadows as much as I would w/ a dark surround and forces me to bump up w/ Shadows/Blacks which I always seem to need (esp. w/ non-linear Epson behavior in the very dark tones).
That matches this post by Marcin: http://forum.luminous-landscape.com/index.php?topic=113411.msg933194#msg933194
-
Thanks guys. The surround may very well be a major part of the problem. I have it set as dark as it'll go. Andrew's "why are my prints too dark" has given me some ideas and I'm going to try to build a profile for my different applications. My problem, as I see it, is that I'm using one calibration/profile for everything. I do a lot of commercial work which is all delivered digitally, but also do portrait and fine art prints on various papers and canvas. Time for me to sharpen my colour skills and follow through with a proper calibration/profile for each area that I work in. It's going to take the two things I often find in short supply... time and money. ;)
I'm sure I'll have more questions, so please bear with me.
-
Remember also to put more light arround the display, it also has influence to what we see in darks and near blacks. I always strive to achieve neutral background that is visually close to 50% grey on the monitor (RGB 120,120,120 for TRC ɣ2.2 calibrated display)
(https://dl.dropboxusercontent.com/u/19059944/Spectrino2.jpg)
-
Remember also to put more light arround the display, it also has influence to what we see in darks and near blacks. I always strive to achieve neutral background that is visually close to 50% grey on the monitor (RGB 120,120,120 for TRC ɣ2.2 calibrated display)
(https://dl.dropboxusercontent.com/u/19059944/Spectrino2.jpg)
This is difficult to do if you are using a monitor white point of 6500K without affecting viewer adaptation. There are 6500K LED sources available, but I don't know of high-CRI ones, although they're probably out there. There are high CRI 5000K sources (both LED and fluorescent; I use the old Macbeth tubes), and that's one reason to calibrate your monitor to 5000K. It used to be that that gave a dim display, but today's monitors are so bright that you more often have trouble getting the brightness down than getting it up.
Then there's the reduction in contrast that ambient light causes, which is difficult to calibrate out.
All in all, I'm in favor of working in dim ambient light, and turning it off (I use a desktop remote) for critical judgements. I also work with a dark gray surround to make it easier to see into the shadows, and a white surround for those critical judgements. Having two different monitor brightnesses for the two cases is also OK in my book.
Jim
-
I am surprised with a cd/m2 at 80 you are having issues. I've never seen a flat panel display with a setting that low that wasn't too dark and end up with light prints. I've also calibrated many displays for many people, and to get a white point match I never end up with a white point at D65. How are you evaluating the print in comparison? Calibrating and profiling a display is intended to allow you to predict how a print will turn out, but the starting point is having a standard viewing area in which to evaluate the print.
I have 3 NEC displays, two 30" displays at my home side by side, and one 27" at my work. My target white point is always manual, which I tweak until my colors match the print on my viewing wall. It never ends up that cool. Currently my main display where I do most of my critical matching the white point is set to custom, x=0.356 y=0.373, 122 dc/m2 and contrast at 350:1. I have enough confidence in this display that my initial test prints are usually 24x30".
My 2nd display which sits side by side is set to emulate a "typical" consumer display ... basically what most displays look like if you buy one off the shelf . I know many who look at the images just bought a computer, turned it on, and have no clue as to setting up color or even brightness. So my second display is set up to emulate sRGB at 160 cdm/2, 6500k, native gamma and native contrast. I use this display to evaluate images I plan on posting to websites. i realize that there is no "standard" for all of those out there looking at these images, but I also realize that none of them are set up to emulate a printing environment so I know this is at least closer to the "average". Fortunately it really doesn't make much difference much of the time. Some of them look a little washed out so I'll add just a little density and perhaps saturation.
The challenge of calibrating and profiling a display to me is understanding that there are no "magic" numbers to set the display up with. The only thing you can do is keep tweaking things until you get a match. There is no specific intensity, contrast ration, or color settings. The only way you can get a "match" is to have a standard place in which to view the prints. Sometimes if you don't get a match it's because this evaluation area isn't setup correctly. Additionally many try to set the two side by side. I find it better to set them up so you can't actually see them at the same time. I have to turn my head to look at the print, then back to see the display.
Just some thoughts
-
This is difficult to do if you are using a monitor white point of 6500K without affecting viewer adaptation. There are 6500K LED sources available, but I don't know of high-CRI ones, although they're probably out there. There are high CRI 5000K sources (both LED and fluorescent; I use the old Macbeth tubes), and that's one reason to calibrate your monitor to 5000K. It used to be that that gave a dim display, but today's monitors are so bright that you more often have trouble getting the brightness down than getting it up.
I use 5000K High CRI LED (Ra >95%). It gives similar tone to D65 calibrated display. The 6500K light would be perceptually cooler than D65 calibrated display due to chromatic adaptation issues, caused by the different way our eyes react to narrow band spectra (like display) and wide band spectra (like daylight) stimuli.
-
I am surprised with a cd/m2 at 80 you are having issues. I've never seen a flat panel display with a setting that low that wasn't too dark and end up with light prints.
I am surprised that you're suprised, as we all know it's relative. There's no "luminance of truth" or the universal level of monitor brightness. Furthermore, the average level of daylight in living rooms is... between 60 to 120lx, which is equivalent to ~20-40cd/m² (lx/∏=cd/m²) of monitor brightness, so way less than 80cd/m²
-
I use 5000K High CRI LED (Ra >95%). It gives similar tone to D65 calibrated display. The 6500K light would be perceptually cooler than D65 calibrated display due to chromatic adaptation issues, caused by the different way our eyes react to narrow band spectra (like display) and wide band spectra (like daylight) stimuli.
Please explain that. I don't understand the point you're making. Feel free to be technical; I worked as a color scientist for IBM for six years.
It sounds like you're saying that the viewer of a 5000K LED-illuminated work is in the same state of adaptation as the viewer of a screen with a D65 white point. But that can't be right. So I'm missing something.
Jim
-
Please explain that. I don't understand the point you're making. Feel free to be technical; I worked as a color scientist for IBM for six years.
It sounds like you're saying that the viewer of a 5000K LED-illuminated work is in the same state of adaptation as the viewer of a screen with a D65 white point. But that can't be right. So I'm missing something.
Jim
For what it's worth I did a test with a Solux running at 4200K v monitor. The right hand side of the monitor was covered with a highly flat reflectance paper and illuminated at 45 degrees by the Solux. The left hand side was adjusted so that the XYZ from a spectro matched that of the right hand side reflected light from the Solux. The perceived chromaticity and brightness between the left and right side were quite a close match. At least within less than a couple hundred Kelvin max. If the narrow spectra of the monitor v the broad spectra of the reflected light on the right side created different color temps (like D50 v D65) then I would not have expected them to match,
That said, I'm somewhat astonished at the perceptual similarity of D65 monitor image v a D50 illuminated image. That motivated doing the above test. I'm not sure exactly what causes the perceptual similarity but when an broadband illuminated white object is brought within an inch or less of a screen patch with the same measured XYZ values there is a sudden perceptual jump to a match. When they are distinctly separate physically, the two look way different.
-
I downloaded DigitalDog's reference print (with the white dog). I calibrated fine to match Epson Cold Press Natural and Enhanced Matte. The Epson prints were done using the Epson ICC profiles for the respective papers. I'm now trying to match Breathing Colour Lyve, using BC's profile.
The red is coming out orange. Does that mean the profile is off?
-
I am surprised that you're suprised, as we all know it's relative. There's no "luminance of truth" or the universal level of monitor brightness. Furthermore, the average level of daylight in living rooms is... between 60 to 120lx, which is equivalent to ~20-40cd/m² (lx/∏=cd/m²) of monitor brightness, so way less than 80cd/m²
yes it's relative, I think my post was pretty clear about that. but I'm surprised because the challenge is the prints are too dark even at that dim of a setting. I've calibrated dozens of displays for myself and others over the years and I believe the lowest number I've ever used is 95. If 80 is too bright, meaning the relatively correct number would be more in the line of 65-70 to get a match, I think at that point it would be difficult to get any decent saturation on the print, resulting in a pretty washed out print in many viewing environments.
I'm not sure how Mike is evaluating the prints themselves, but rather than dimming the display even further, I would see about getting a little more light on the prints.
Regarding your numbers, for the heck of it I set my display at 30. While it may match a dim living room, I can't imagine trying to work with it. There wasn't enough light to see any detail even in lower midtones.
-
I evaluate the prints under different light. Unfortunately, my studio is lit by fluorescent, but I check under an OttLight at my work station as well as natural light coming in through large windows and glass doors at the front of my studio.
-
Hi Mike,
I've just read this thread and I think I might have an answer of sorts to your problem, although you may have already followed this procedure. The NECs are excellent displays IMO, but there are some settings you have to take care of before you do you first calibration with Spectraview(or any software). It's been a long time since I've had to do all of this with a new NEC, although I have three of them currently. You will find all of this info in the NEC manual as well as the Spectraview Guide. Since I have neither here at the moment, as I'm in the process of moving, I cannot readily recall the exact settings to take care of, but you will find the procedure in the manuals as mentioned. The idea here is to make sure all of the user available display settings should be set to neutral before calibrating. Then you lock out the possibility of anyone changing any of those settings, including yourself of course. It may be tempting, under certain circumstances, to tinker with the display settings to lighten, change contrast etc., but of course that would negate the calibration settings. Please take a very careful look at the manual and guide Mike, since your numbers seem to be in the range of a good calibration, although your prints would seem to indicate otherwise. And as Wayne mentioned, make sure you have at least the minimum illumination for your prints. I've learned in this thread that there are apparently 5000K LEDs with a CRI perhaps as high as 95. I'll be setting up a new viewing station at home, so I'll certainly be looking into the possibility of the LEDs instead of fluorescent, since the new energy efficient 5000K bulbs cannot get any higher than CRI 80.
I hope this helps in some way Mike.
Gary
-
if someone has a URL for a 5000K (or adjustable) LED lamp w/ high CRI, pls post it.
-
if someone has a URL for a 5000K (or adjustable) LED lamp w/ high CRI, pls post it.
+1
Gary
-
Gary
I'm pretty sure I followed the guides when I first set up the monitor, but it may be worthwhile starting from scratch again. I've set up a few different calibrations for different papers I'm using, according to the article at DigitalDog - makes a lot of sense in how Andrew describes it.
Those look great, but it's the Lyve canvas that I'm having trouble with now. The Reds are coming out Orange and too much contrast. I don't know if it's the profile (I have two different profiles from BC for Lyve and they both look the same) or something else. At this point my head is kinda muddled with colour - I don't know if I can trust my eyes. lol
-
After printing on the exact same printer for 8+ years, I think there is another factor with this prints are too dark issue: We need to train ourselves. I still occasionally catch myself digitally developing an image that really is too dark. I can somewhat get away with in on screen. Heck, sometimes images I post on-line get rave reviews. Then after several months I go to print that image and it is too dark. But it's not really the print that is too dark, it is the image.
Dave
-
After printing on the exact same printer for 8+ years, I think there is another factor with this prints are too dark issue: We need to train ourselves. I still occasionally catch myself digitally developing an image that really is too dark. I can somewhat get away with in on screen. Heck, sometimes images I post on-line get rave reviews. Then after several months I go to print that image and it is too dark. But it's not really the print that is too dark, it is the image.
Dave
I separate out the issue of printing, and "too dark," issues from the monitor surround, and hard/soft proofing by the simple expedient of printing a known reference image.
The advantage of a known reference image is that comparing them is, for the most part, not dependent on the illuminant. A previously created reference image and a print of the same will appear to be identical outside of possible metameric failure. The latter only occurs if the spectral response of the print and reference image differs materially and it usually is a secondary effect at most and is always insignificant if the paper has no OBs and the two are viewed in daylight.
My reference image of choice is a ColorChecker and the compared print is a print of the same ColorChecker image using Absolute Colorimetric. When I do that I don't even look at the CC image on the monitor. It's monitor appearance has no affect on printing. If the print and Colorchecker match, and they will if color management including the profile is correct, then any and all issues of prints being too dark, etc., are due to editing and viewing on the monitor side. It's that simple.
-
For what it's worth I did a test with a Solux running at 4200K v monitor. The right hand side of the monitor was covered with a highly flat reflectance paper and illuminated at 45 degrees by the Solux. The left hand side was adjusted so that the XYZ from a spectro matched that of the right hand side reflected light from the Solux. The perceived chromaticity and brightness between the left and right side were quite a close match. At least within less than a couple hundred Kelvin max. If the narrow spectra of the monitor v the broad spectra of the reflected light on the right side created different color temps (like D50 v D65) then I would not have expected them to match,
Metameric matches under different illuminants can exist perfectly well with spiky and smooth spectra, and if those matches are obtained for a sample that is the white point for the image, the white point is defined by the same filter response as that which converts the rest of the image to triplets. Therefore, wide vs narrow spectral responses will not shift the white point.
I think you're saying that you don't observe such a shift.
In all of the soft proofing studies that I know of that presented the monitor and the proof print in the same field of view an a way that forced a single adaptation, the white point of the print and that of the monitor had to be the same for effective soft proofing.
That said, I'm somewhat astonished at the perceptual similarity of D65 monitor image v a D50 illuminated image. That motivated doing the above test. I'm not sure exactly what causes the perceptual similarity but when an broadband illuminated white object is brought within an inch or less of a screen patch with the same measured XYZ values there is a sudden perceptual jump to a match. When they are distinctly separate physically, the two look way different.
You're talking about adaptation plasticity. And yes, it's pretty amazing. Which is why you can get away with soft proofing for D50 illumination on a D65 monitor.
Jim
-
if someone has a URL for a 5000K (or adjustable) LED lamp w/ high CRI, pls post it.
http://indiecinemaacademy.com/complete-led-color-database-cri-tlci-cqs-tm30-15/
you can use a LED panel too in the office
-
Metameric matches under different illuminants can exist perfectly well with spiky and smooth spectra, and if those matches are obtained for a sample that is the white point for the image, the white point is defined by the same filter response as that which converts the rest of the image to triplets. Therefore, wide vs narrow spectral responses will not shift the white point.
I think you're saying that you don't observe such a shift.
Yes, that is exactly what I'm saying.
-
if someone has a URL for a 5000K (or adjustable) LED lamp w/ high CRI, pls post it.
I’ve tried solux bulbs at different temperatures. I’ve moved away from lamps that are too cool, I struggle with how cold the image looks.
I’ve been testing bulbs from Lumicrest (http://lumicrest.com/product-category/led-lights/led-par-lights/) and comparing them to the solux. The CRI rating is one of the highest I’ve seen in an LED (95) but also looking at the individual measuements and detail of the rating (http://lumicrest.com/pdf/Photometry/Apturi_Par20_4000K_95CRI_Report.pdf) they seem to be pretty solid in all 15 wavelengths that are tested. I’ve been pretty happy with them. I actually prefer the 3000k bulbs over the 4000k.
-
Well, for whatever reason I can't get a good print out of Lyve canvas. I tried some Epson Exhibition Canvas Matte that I had and it printed beautifully. Almost a perfect match for the monitor. On the Lyve canvas, it really dumped a lot of ink on it - so much that it felt very wet after printing. I've never felt it this wet before.
I think I've got it under control now. Thanks guys for all the help and suggestions. I learned a lot over the last few days. ;)
-
I replaced the original 4700K Solux w/ their 3500 and like it. It matches my 3000K living room well. I also have one of those cheap Amazon 4temp, 5intensity level LED desk lamps.