Editing in a dark room: results

[digitaldog]

I'll make one more comment and I probably have to move on (I'm being asked politely to do so).


The one person who's suggested a sound methodology thus far is Doug. Setup the environment with proper control to test various levels of ambient light. Control the surrounding, and placement of the lights (one can easily sabotage the results by pointing a light striking the display, use a hood etc). Start with a couple well controlled reference images. Specify (or not?) the edits that each person should apply. Get a good cross section of users. You can have someone with the best color perception on the planet who hasn’t a clue how to edit an image, ruin the experiment!. Consider that it is critical to involve human perception! Soft proof Make prints. Properly.


I'll also point out that people who report 'my prints are too dark' very, very often find, they are not too dark. They appear too dark compared to the display. Sure, some people do produce dark prints. That shouldn't ever happen with a color reference image (for example, the Roman 16s). Are the prints too dark or not? IF we agree that most often, they are not, there is a visual disconnect between viewing the print next to the display (which involves all kinds of variables, adapting to reflective/emissive output, etc), I can't see how we can disagree this test must include what people perceive. Where is the data that users report their prints are too dark in dark conditions? The other end of the scale, there's a huge number of users all over the net who report this, even to this day.


Is the experiment going to involve Soft Proofing? It should. What RI? BPC? What's the quality of the profile? Where do you see Frans stating his test was conducted with soft proofing on or the results were a dark print?


There's a ton of variables here. And Doug started providing a methodology to prove or disprove what Frans already believes.
Based on the ISO, based on Dr. Lang, based on 20+years of editing in dark conditions and setting up digital darkrooms for clients, I see nothing to indicate that editing in a dark room (again, undefined) results in images and prints that are too dark. For me, for my customers and it seems, a group posting here.


I'll leave it at that. Before I get locked out.

[Frans Waterlander]

And why would this thread be locked? Because people have opinions that differ and they discuss the issues? Come on! If we all stay civil here then I can't think of any reason why this thread would be locked.

I agree with GWGill's point that viewing conditions effects are a continuum. In my opinion, the ambient brightness should roughly match the monitor brightness; that way the ambient serves as a anchor or reference point for our eyes and this will reduce the tendency to turn up the image brightness under bright ambient conditions and turn down the image brightness under dim ambient conditions. How do you achieve this brightness match between ambient and monitor? In my case it's simple: I have two SoLux lights hanging from the ceiling to the right of my hooded monitor in line with the plane of the monitor screen, so no light falls on the monitor screen. There is a good brightness match between the monitor and prints viewed in this light (as it should be). These lights illuminate the general area to the right of the monitor, not just the prints, so I have my reference ambient at about the same brightness when I leave my SoLux lights on while editing. This works like a charm for me and I suspect it would work well for many others. If you have a viewing booth that doesn't illuminate the ambient much, then I would propose to put some image or material with average toanlity in the booth and leave it on during editing.

Why don't we here complaints from many people on this issue? I guess they either have a reasonable match between monitor and ambient, they have learned to live with the handicap or they waste paper, ink and time with trial and error. I don't put much value on recommendations for ambient values that vary all over the place (<32 lux, <16 lux, 4 lux). Those values may or may not work, depending on the monitor settings and how well people have learned to live with the problem.

[Doug Gray]

And why would this thread be locked? Because people have opinions that differ and they discuss the issues? Come on! If we all stay civil here then I can't think of any reason why this thread would be locked.

I agree with GWGill's point that viewing conditions effects are a continuum. In my opinion, the ambient brightness should roughly match the monitor brightness; that way the ambient serves as a anchor or reference point for our eyes and this will reduce the tendency to turn up the image brightness under bright ambient conditions and turn down the image brightness under dim ambient conditions.

I don't think operating where many visible objects in a room are at a luminance equal or above the monitor is a good idea. One white adapts to the brightest objects around and you want that to be the monitor. I like to run with a Photoshop neutral backing set to L=70 but it is still brighter than most everything else in my field of view.

There is an interesting point that hasn't been brought up. Ambient room light is illuminance. This is the light hitting a surface and is measured in Lux.  Luminance is the light being reflected from a surface and is measured in cd/m^2.  This is what your eyes actually see, not Lux. A perfect, reflecting (white, not mirror) surface will reflect Lux/Pi cd/m^2. For example, in a proofing booth with 500 Lux of illumination the measured cd/m^2 from the surface of a blank, Baryta print paper is about .92*Lux/Pi which is about 146 cd/m^2. If your monitor is set to 160 cd/m^2 this is very close to what soft proofing pure white on Baryta using "show paper white" will produce.

Now most people, including me, do not run their monitors at 160 cd/m^2 nor do they run soft proofing at 500 Lux even if those are recommended in graphic arts. I run a bit over 300 Lux for proofing and 100 cd/m^2 for the monitor. My room light runs around 3000K and Lux values depend entirely on where, and with what orientation I measure it. I don't really consider it very important except that I want the Lux that illuminates my monitor to be low enough that any reflected light is below .25 cd/m^2. It turns out that almost all the reflected light is around a 20 degree axis to the screen and roughly 2% of that is reflected back. Outside of that there is virtually no light reflected back. I can illuminate the screen with 200 Lux at an angle of 45 degrees and yet measure a luminance increase of only about .1 cd/m^2. But light coming from my direction is critical and gets reflected back hence I wear a dark or black shirt to reduce light reflected from me. So what matters to me is not directly the ambient but the luminance from reflected ambient and that is dependent on both the intensity and the angle. If the angle is more than about 20 degrees from max. reflectance it doesn't affect what I see on the screen.

Other than that both the monitor's white point and the neutral surround I use (I have a 30" monitor and just use the display to provide a surround) has a luminance significantly greater than the luminance off most objects in my general viewing range. I find this is necessary so that I properly white adapt. If I have particularly deep shadows I want to look closely at I'll drop the surround way down and turn the lights off if necessary but I don't edit with the lights off.

How do you achieve this brightness match between ambient and monitor? In my case it's simple: I have two SoLux lights hanging from the ceiling to the right of my hooded monitor in line with the plane of the monitor screen, so no light falls on the monitor screen. There is a good brightness match between the monitor and prints viewed in this light (as it should be). These lights illuminate the general area to the right of the monitor, not just the prints, so I have my reference ambient at about the same brightness when I leave my SoLux lights on while editing. This works like a charm for me and I suspect it would work well for many others. If you have a viewing booth that doesn't illuminate the ambient much, then I would propose to put some image or material with average toanlity in the booth and leave it on during editing.

Why don't we here complaints from many people on this issue? I guess they either have a reasonable match between monitor and ambient, they have learned to live with the handicap or they waste paper, ink and time with trial and error. I don't put much value on recommendations for ambient values that vary all over the place (<32 lux, <16 lux, 4 lux). Those values may or may not work, depending on the monitor settings and how well people have learned to live with the problem.

[GWGill]

I don't think operating where many visible objects in a room are at a luminance equal or above the monitor is a good idea. One white adapts to the brightest objects around and you want that to be the monitor. I like to run with a Photoshop neutral backing set to L=70 but it is still brighter than most everything else in my field of view.

I think there is a difference between the recommendations for a normal office work environment, and a critical viewing environment. For the former (going by ASA and personal experince), the general recommendation is to have a similar level of light for both paper documents and electronic displays, to minimize visual fatigue. Arranging things so that the illumination used for documents minimizes display glare is the trick.  Electronic displays originated in CRT technology days, and inherited a lot of the assumptions that go along with Television - and television standards assume a viewing environment that is quite dim (2.2 gamma image encoding, CRT 2.4 gama native response). Things like ISO recommendations seem to follow the TV lead, with minimization of glare being one motivating factor.

[digitaldog]

I agree with GWGill's point that viewing conditions effects are a continuum.

As I do and have said so. As does the ISO, both Karl's (both color scientists) referenced in other posts. Stay on topic! You have a theory that editing in a room that is too dim (undefined by you), results in images that are too dark and prints that are too dark. You tried rather poorly to prove this in the first post here. You were soundly and unanimously shot down. Read (re-read?) posts #2,3,5,6,10,16 and 25, in this thread alone. Note your lack of an answer to back up your theory in post #25. So in terms of peer review, you've moved nowhere. Doug and I have provided a few posts suggesting how you might begin to prove your theory. I suggest you get to work.

You have a couple options.

• Continue writing OT; not concentrating on the dismissal of your ideas presented here that so many have with your first post attempting to back up your theory. Hopefully we'll all move on and ignore this tactic.
• Take the suggestions of your peers and come up with a sound, scientific, unbiased testing protocol and present your actual data to your peers. Post #1 didn't cut it.
• Stop stating that editing in a dim room results in dark image and prints. You will not get called out and shown the work of others like the two Karl's, the ISO, the SpectraView software that do not back that up, just the opposite.
• Continue to state editing in a dim room results in dark images and prints, continue to be called out to prove your theory.

What's a bit telling is Frans deleted his original text in the original thread http://forum.luminous-landscape.com/index.php?topic=113202.0 :
Quote from: Frans Waterlander on August 31, 2016, 02:53:58 PM
I take issue with Andrew's statement that you can't have too little ambient light.
He wasn't able to delete our replies with that text!

Let us examine your very own text that started us down this rabbit hole that wasn't deleted by you:
Care to explain how, in a room that's too dark, you won't edit your images to be too dark, Andrew?
Care to explain using sound testing methodology it will? That hasn't occurred.

[digitaldog]

Why don't we here complaints from many people on this issue? I guess they either have a reasonable match between monitor and ambient, they have learned to live with the handicap or they waste paper, ink and time with trial and error.

Just assumptions on your part.

I don't put much value on recommendations for ambient values that vary all over the place (<32 lux, <16 lux, 4 lux).

It's clear you don't. What is also clear is you can't prove why others should agree with your recommendations or lack thereof! Yet the ISO and two color scientists referenced DO provide recommended values. None state anything about the results being prints or images that are too dark. Prove them wrong.

Those values may or may not work, depending on the monitor settings and how well people have learned to live with the problem.

Without data, you're just someone with an opinion.

[Czornyj]

FWIW - my practical observations of a printer and a color geek are that:
- dark monitor background may open shadows slightly (Bartleson-Breneman effect),
- ...but working in bright environment is even worse due to excessive monitor contrast decrease

As a result I ended up working in dark environment with highlighted monitor background:

[Frans Waterlander]

So, Andrew, we agree that viewing condition effects are a continuum. As the ambient becomes lighter, the image will appear darker, resulting in a tendency to brighten the image during editing; as the ambient becomes darker, the image will appear lighter, resulting in the tendency to darken the image during editing. My test images appear to confirm this. So where is the sweet spot? I suggest that the sweet spot is where the ambient brightness is not higher or lower but equal to the monitor brightness.

Brightness balancing is an important part of fine-tuning the monitor and print viewing set-up so the image on the screen and the print match. So, I suggest that the print viewing set-up is the ideal ambient for image viewing and editing on the monitor. I further suggest that the print viewing setup be placed right next to the monitor to act as the reference point or anchor point for our eyes.

If the print viewing set-up is a viewing booth, then it should be placed next to the monitor set back enough to avoid booth light to fall on the monitor screen and be left on during all editing. The booth should illuminate a print with average tonality or an 18% gray card.
If prints are viewed with e.g. SoLux lights, then I suggest those lights shine down and are set back to avoid illuminating the monitor screen. They should be on during editing. If you have a stand for prints, then put an average tonality print or 18% gray card on it. If you don't have a stand for prints, but handhold them during viewing, then the area that is lit by the lights acts as your ambient.
The use of a monitor hood is of course essential. The rest of the room should be as dark as possible to minimize reflections in the monitor screen.

Another way to create a reference ambient is to use a large 18% gray area around you images on screen. The danger in this approach is that you will want to view your images at larger magnifications and compromise your reference ambient

[GWGill]

Another way to create a reference ambient is to use a large 18% gray area around you images on screen. The danger in this approach is that you will want to view your images at larger magnifications and compromise your reference ambient

Note that in Viewing Conditions speak, "surround" is not the same as "ambient":

[Doug Gray]

Just a thought but it seems to be that this effect, to the extent it's real, should be the same with prints. Prints made with images with what amounts to a gamma adjustment based on working in a dark v well lighted room should also look properly adjusted when viewed against a black wall and a lighted wall. So perhaps a test could be done with sets of prints and having people choose which looked better against black paper and against white.

[digitaldog]

So, Andrew, we agree that viewing condition effects are a continuum. As the ambient becomes lighter, the image will appear darker, resulting in a tendency to brighten the image during editing; as the ambient becomes darker, the image will appear lighter, resulting in the tendency to darken the image during editing. My test images appear to confirm this.

We don't agree, you've provided no proof, and I'm out of here.

[Frans Waterlander]

We don't agree, you've provided no proof, and I'm out of here.

For the record:
Reply #38, GWGill: "...and there is no reason to think that reducing ambient levels below 10% will have no effect - viewing conditions effects are a continuum after all."
Reply #41, Frans: "I agree with GWGill's point that viewing conditions effects are a continuum."
Reply #44, Andrew: "As I do and have said so."
Reply #47, Frans: "So, Andrew, we agree that viewing condition effects are a continuum."
Reply #50, Andrew: "We don't agree."

[GWGill]

We don't agree, you've provided no proof, and I'm out of here.

I've provided a couple of posts supporting the proposition with established color science, while you have made no attempt to disprove it. This leaves one with the impression that like any good conspiracy theorist, no amount of proof will convince you.