Pages: [1] 2   Go Down

Author Topic: 5000 or 6500 K for room light (and monitor)?  (Read 2712 times)

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
5000 or 6500 K for room light (and monitor)?
« on: June 17, 2017, 02:22:38 PM »

Hi!

I want to standardize my room light for photo editing, both wrt brightness and color. The print pipe line seems to be 5000 K. But my monitor calibrated to 5000 K looks yellow. So I have calibrated it to 6500 K. Which color temperature should I choose for the room light? Should I use 5000 K for both room light and monitor? I assume the monitor would not look yellow if the room light was 5000 K, too?

Thank you for your help!

aaron125

  • Newbie
  • *
  • Offline Offline
  • Posts: 16
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #1 on: June 17, 2017, 03:11:06 PM »

If it's Solux bulbs you'll be using, definitely go for their 4700K (maybe they have a 5000K bulb now, not sure, but their 4700K matches a 5800-6500K wp, depending on the exact monitor).

The reason that 5000K is so often talked about is because it's a holdover from the days when the only people who could afford to do any softproofing worked with news print and presses. Obviously in the current times, with the predominance of rc media and FWA, a wp of about 6500K is a much closer match to said media. That's also the reason why a luminance level of 80cdm is no longer set in stone, but anything from 100-160cdm is more common (not to mention that a certain specification calls for a luminance level of 160cdm).

As far as your room lighting, stay away from fluorescent bulbs as even with a CRI > 90, they still produce a very incomplete spectrum. Solux bulbs meanwhile, have an incredibly smooth and even spectrum, not favouring, nor neglecting any part of their spectrum.

For what it's worth, I used to calibrate my Eizo CG242W to 100cdm, L* and 5800-6100K, the variance depending on the specific media upon which I was printing. Now, with an NEC PA272, it's generally set to 120cdm, L* and 6000-6350ishK.


Sent from my iPhone using Tapatalk
Logged

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #2 on: June 17, 2017, 04:10:40 PM »

Hi Aaron,
thank you for your reply. - No it's not Solux for the room light, I intent to buy LEDs, their spectrum is not quite as continuos as that of the Solux, but I've seen some that looked quite OK.
That the Solux 4700 K matches a monitor white point of 5800-6500 K is something I don't understand.
Good light!

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2157
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #3 on: June 17, 2017, 04:55:23 PM »

Henning, this subject has been discussed over 10 times on LuLa.

Why are you asking this again? Clearly for some time you've been a contributor to these forums seeing you've posted 800 times.

Here's just one recent discussion... http://forum.luminous-landscape.com/index.php?topic=116755.0
Logged

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #4 on: June 17, 2017, 06:27:05 PM »

Hi Tim,
thanks for the link, I'll study it. Admitted: I forgot to search the forum first.

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #5 on: June 30, 2017, 04:54:06 PM »

I have now read that thread and a couple of others. So I have ordered 6 of these Yuji bulbs

https://store.yujiintl.com/collections/bc-series/products/bc-series-a60-high-cri-remote-phosphor-led-bulb-unit-2-pcs

By these, I hope to lit my room with 5000K to a brightness level of about 300 lux or so, which I then will use as my standard surround and viewing condition.

But I am very confused by a controversy between Andrew Rodney and Frans Waterlander.
http://forum.luminous-landscape.com/index.php?topic=113411.0
 
Andrews take is, that editing on a monitor in a dark surround will NOT make your prints look too dark.
GWGill (Argyll) replies to Waterlander, who claims the opposite:

"That's certainly consistent with expected color appearance phenomena - specifically the Stevens Effect and Bartleson-Breneman Equations. (See Mark Fairchild's  book, "Color Appearance Models", chapters 6.7 and 6.9).

And Andrew replies:

"I've got that book, the equations are out of my pay grade. But I do see this specifically on page 140: The Stevens effect indicates that as luminance levels *increase*, dark colors will appear darker and light colors will appear lighter. The Stevens effect can be demonstrated by viewing an image at high and low luminance levels.(Andrew Rodney:this image is or isn't an emissive display?) A B&W  image is particularly effective for this demonstration. (Andrew Rodney: Fran's used the wrong kind of image to test). At low luminance levels (Andrew Rodney:undefined), the image will appear to have a rather low contrast (Andrew Rodney: Contrast or perceived brightness?). White areas will not appear very bright and, perhaps surprisingly, dark areas will not appear very dark. If the image is moved to a significantly higher level of illumination, white areas appear substantially brighter and dark areas appear darker.

Nowhere do I see anything that states or suggests editing images on an emissive display in a dim environment *produces images or prints that are too dark*."

I don't have Fairchilds book, just this PDF:
http://www.cis.rit.edu/fairchild/PDFs/AppearanceLec.pdf,
and from this I can not see it, but dare I suspect that the quote in Andrews reply refers to a reflection print? Then it would make sense to me. But for the monitor image??

While I can see that a dark environment will be the best for judging colors on the monitor, I fail to see the same for tonal edits. In fact, what made me start this topic, was a felt need for a standard brightness in my room for editing, based on experience like this:
 
For an image I edited in my room while it was dark, I was happy with this tone response curve: 1-TRC in dark. Next day, when I looked at it when the room was lit by dim day light, the image looked dull to me, and I changed the tone curve to: 2-TRC in light. I don't print myself, rely on a print service, but image 1 WILL print darker/less contrasty than image 2, will it not?

I am aware of that this topic has been the object of an extended quarrel already, but I just don't understand what I am missing.



Rand47

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1061
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #6 on: June 30, 2017, 07:47:34 PM »

I have now read that thread and a couple of others. So I have ordered 6 of these Yuji bulbs

https://store.yujiintl.com/collections/bc-series/products/bc-series-a60-high-cri-remote-phosphor-led-bulb-unit-2-pcs

By these, I hope to lit my room with 5000K to a brightness level of about 300 lux or so, which I then will use as my standard surround and viewing condition.

But I am very confused by a controversy between Andrew Rodney and Frans Waterlander.
http://forum.luminous-landscape.com/index.php?topic=113411.0
 
Andrews take is, that editing on a monitor in a dark surround will NOT make your prints look too dark.
GWGill (Argyll) replies to Waterlander, who claims the opposite:

"That's certainly consistent with expected color appearance phenomena - specifically the Stevens Effect and Bartleson-Breneman Equations. (See Mark Fairchild's  book, "Color Appearance Models", chapters 6.7 and 6.9).

And Andrew replies:

"I've got that book, the equations are out of my pay grade. But I do see this specifically on page 140: The Stevens effect indicates that as luminance levels *increase*, dark colors will appear darker and light colors will appear lighter. The Stevens effect can be demonstrated by viewing an image at high and low luminance levels.(Andrew Rodney:this image is or isn't an emissive display?) A B&W  image is particularly effective for this demonstration. (Andrew Rodney: Fran's used the wrong kind of image to test). At low luminance levels (Andrew Rodney:undefined), the image will appear to have a rather low contrast (Andrew Rodney: Contrast or perceived brightness?). White areas will not appear very bright and, perhaps surprisingly, dark areas will not appear very dark. If the image is moved to a significantly higher level of illumination, white areas appear substantially brighter and dark areas appear darker.

Nowhere do I see anything that states or suggests editing images on an emissive display in a dim environment *produces images or prints that are too dark*."

I don't have Fairchilds book, just this PDF:
http://www.cis.rit.edu/fairchild/PDFs/AppearanceLec.pdf,
and from this I can not see it, but dare I suspect that the quote in Andrews reply refers to a reflection print? Then it would make sense to me. But for the monitor image??

While I can see that a dark environment will be the best for judging colors on the monitor, I fail to see the same for tonal edits. In fact, what made me start this topic, was a felt need for a standard brightness in my room for editing, based on experience like this:
 
For an image I edited in my room while it was dark, I was happy with this tone response curve: 1-TRC in dark. Next day, when I looked at it when the room was lit by dim day light, the image looked dull to me, and I changed the tone curve to: 2-TRC in light. I don't print myself, rely on a print service, but image 1 WILL print darker/less contrasty than image 2, will it not?

I am aware of that this topic has been the object of an extended quarrel already, but I just don't understand what I am missing.

Hi, I'm not a math guy... but for me the bottom line is "good to excellent" screen-to-print match.  And one of the keys to this, IMO, is eliminating as many variables in the equation as possible.  One of the significant variables is the ambient light in the editing environment. 

Here's one way to think about the initial set up process (forgive any obvious bits):
  • Evaluation station w/ consistent K temp and luminance level (I use a GTI booth with 5000K at the unit's standard luminance value)
  • Consistent ambient light level in editing room (I prefer relatively low ambient light.)
  • Monitor calibrated to native (max) color gamut (In my case NEC PA302)
  • From this setup, I print a standard print evaluation file, making NO adjustments to the file, and printed on a nice wide gamut paper such as Ilford Gold Fibre Silk

  • Evaluate the print for a decent screen to print match.  If the print is "too dark" compared to the screen, I re-run the monitor calibration and decrease the luminance value (or visa versa).  Most monitors come "way too bright" out of the box, IMO  (e.g. I'm calibrated to 100 cdm2)
  • After a few calibration runs you should be able to achieve a really excellent screen to print match
  • You're done, except for soft proofing your own images for the paper/printer you're using

This approach works very well in my environment.  In your case, since you're using a lab, make sure that each "print run" you make between calibration adjustments has the identical parameters at the lab so that any variation in the print isn't due to the lab doing one thing one time, and another thing on subsequent runs.

Once you're dialed in in this basic way, you can explore variations in K temp and luminance of evaluation station based on any unusual "final destination viewing conditions" for the finished product (e.g. a home where the only light is incandescent, say around 3200K) and then make notes about different calibration parameters used to achieve screen to print match in varying "final display" conditions.  99% of the time my standard setup works well enough for normal variations in the client's final viewing conditions.

And just for completeness, monitor is 6500k, Gama 2.2, native gamut.

Rand
« Last Edit: June 30, 2017, 10:47:11 PM by Rand47 »
Logged
Rand Scott Adams

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2529
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #7 on: June 30, 2017, 09:30:48 PM »

My print viewing/evaluation table originally had a track of 4 Solux halogens.  I got tired of burning my fingers/hands from time to time though.  I got some MR-16 4700K LEDs for the track I have over the table.  Moving to LEDs also allowed me to add two more WAC fixtures to the strip without overloading the power supply.  I'm sure the spectral analysis isn't as top notch as the halogens but I'm happy with how it performs. 

NB. I don't print professionally 

 
Logged
Regards,
Ron

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #8 on: July 01, 2017, 02:22:54 PM »

Thanks to the both of you for your comments. -
WRT to room lightness:
Jim Kasson has very clearly demonstrated that an image looks more contrasty with a dark surround:
http://blog.kasson.com/the-last-word/in-image-editing-the-surround-matters/
This 'surround' refers to the on-screen surround of the actual image. I would think, and so is  my experience, the same is true for room illumination. Strangely, Jim himself does not take this into account when he, too, recommends to darken the room:
http://blog.kasson.com/the-last-word/in-image-editing-the-room-illumination-matters/
But in this last post, he talks about color, not contrast.

So my conclusion is:
Dark room for color editing, anticipated/standard viewing light in room for tonal edits. I think that the quoted passus from Fairchild does in fact refer to reflection prints, and that Andrew simply is not aware of this.
Good light!

Ethan Hansen

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 71
    • Dry Creek Photo
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #9 on: July 01, 2017, 02:40:27 PM »

For those using V4 profiles, the Perceptual Reference Medium uses the ISO 3664 P2 viewing condition with a neutral surround reflectance of 20%.

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2157
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #10 on: July 01, 2017, 07:58:50 PM »

Thanks to the both of you for your comments. -
WRT to room lightness:
Jim Kasson has very clearly demonstrated that an image looks more contrasty with a dark surround:
http://blog.kasson.com/the-last-word/in-image-editing-the-surround-matters/
This 'surround' refers to the on-screen surround of the actual image. I would think, and so is  my experience, the same is true for room illumination. Strangely, Jim himself does not take this into account when he, too, recommends to darken the room:
http://blog.kasson.com/the-last-word/in-image-editing-the-room-illumination-matters/
But in this last post, he talks about color, not contrast.

So my conclusion is:
Dark room for color editing, anticipated/standard viewing light in room for tonal edits. I think that the quoted passus from Fairchild does in fact refer to reflection prints, and that Andrew simply is not aware of this.
Good light!

The Jim Kasson first demonstration is not very helpful in real world application from my experience especially since we don't compare images in that type of setup because adaption is not factored in. And the differences between the three are so subtle that it won't have much influence on editing because adaptation will kick in a lot quicker viewing one image instead of comparatively.

Also I've noticed editing software improvements such as going from CS3 ACR to CS5 suspiciously appear to change black point appearance or maybe calibration software upgrades are defining display profile's differently for software to interpret black point when moving sliders and point curve adjusts.

Note the two different edits below. Why would I have made the CS3 look so crappy and plug up the shadow detail in the balloons. I do remember editing the CS3 version on the left years ago on a calibrated (old Xrite Display) 22in. Dell and the CS5 version on the right on my current calibrated (ColorMunki Display) 27in. LG both in the same darkened room. This isn't the only one I had to re-edit with similar differences. Something isn't adding up and it has nothing to do with editing in a darkened room.

Logged

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #11 on: July 02, 2017, 07:20:39 AM »

Hi Tim!
My choice of these 2 edits would be the CS3 one - by far. Apart from that, I am out of Adobe, exactly because of the things they do under the table. (And the cloud, of course.) This something which isn't adding up - I hope and think that PhotoLine saves me from that. Otherwise, I would be eager to learn what it is.
Good light - Hening.

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2157
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #12 on: July 02, 2017, 11:38:23 AM »

Hi Tim!
My choice of these 2 edits would be the CS3 one - by far.

That surprises and troubles me because the balloons on the CS3 version don't show separation between the red horizontal balloon on the bottom left and the angled to the left purple balloon just above it. Where they meet is one solid formless black mush.

To rule out subjectivity since you say you prefer (I'm assuming like) the CS3 version are you seeing the same formless black mush in the CS3?

I don't know if this is due to editing on my LG with CS5 or CS3 and the Dell. Or your calibration black point could be brightening the blacks in the CS3 version and extremely over brightening the CS5 to where it looks washed out, low contrast.

I've provided a screen grab of how these two appear in my Firefox browser and Apple's Color Meter RGB sampler to locate the blackest black in the balloon separation detail between the two.

Could you please confirm if you're seeing the same crushed blacks in the CS3?
Logged

rdonson

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2529
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #13 on: July 02, 2017, 12:19:15 PM »

Tim, it's clear to me.
Logged
Regards,
Ron

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #14 on: July 02, 2017, 12:22:58 PM »

Well I looked at the image as a whole. And laziness kept me from adding "maybe I would lift the shadows for the balloons separately a bit". One thing is seeing every detail everywhere as clearly as possible, another thing is an image with an appealing contrast.

Here is how these 2 balloon clusters look in Safari on my calibrated Eizo CG243W.

If I search for the darkest area between the 2 balloons you mention, my Digital Color Meter reads as low as 0-0-3 in the CS 3 version. But also in the CS5 version, I can find a spot as low as 0-1-4!  Only in a minor area though. (I can not reproduce these spots in a timed screen grab).

No reason for trouble :-)

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2157
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #15 on: July 02, 2017, 12:34:28 PM »

Thanks for confirming, Ron.

I re-edited this image in CS5 just now and it's clear to me that there were great improvements engineered into ACR's Fill, Contrast, Black sliders and added sensitivity to refined Point Curve adjusts. I was able to darken the bricks in shadow and boost definition and retain clarity and tonal separation in the balloons which I do remember running into a brick wall with CS3 which I thought I'ld bottomed out in the Raw file from under exposing to retain detail in the sunlit bricks.

All this editing in a darkened room surround didn't make much of a difference.
Logged

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2157
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #16 on: July 02, 2017, 12:44:24 PM »

Well I looked at the image as a whole. And laziness kept me from adding "maybe I would lift the shadows for the balloons separately a bit". One thing is seeing every detail everywhere as clearly as possible, another thing is an image with an appealing contrast.

Here is how these 2 balloon clusters look in Safari on my calibrated Eizo CG243W.

If I search for the darkest area between the 2 balloons you mention, my Digital Color Meter reads as low as 0-0-3 in the CS 3 version. But also in the CS5 version, I can find a spot as low as 0-1-4!  Only in a minor area though. (I can not reproduce these spots in a timed screen grab).

No reason for trouble :-)

Thanks, Hening.

Your Color Meter of the black points is as it should be because I converted the screengrab to sRGB which of course is going to change the numbers in the shadow areas especially close to black. And your screengrab jibes with what I'm seeing.

CS5's Black slider and Fill combo allowed far more movement in the Black slider before clipping than CS3 allowed.

This is the issue with software upgrades and legacy Raw image Schewe and other advocates of Raw processing warned about. Image to image will vary and so no blanket statement can be made about how much detail new improvements to software will yield.

But that can be quite cumbersome when you have over 1000 Raw images edited in old software. I'm still not done re-editing 1000 Raw images in CS5 and now we have PV2012 improvements. Forget it!
Logged

Hening Bettermann

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 837
    • landshape.net
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #17 on: July 02, 2017, 12:54:09 PM »

> I re-edited this image in CS5 just now

I agree that this edit combines the best of the two!

Tim Lookingbill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2157
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #18 on: July 02, 2017, 01:08:17 PM »

Thanks, and getting back on the subject of ambient room lighting surround influencing edits, I've posted a screengrab of what I started out with on that shot.

To be honest I didn't rely on my eyesight to make the editing judgments on setting contrast ratio between the highlighted bricks vs those in shadow. I kept in mind from memory the brightness level from a setting sun and whether the bricks in sunlight retained definition or was just glaring. Usually when viewing a scene lit like this our eyes adapt to the bright bricks and when we focus on say the balloons they first appear dark but then our eyes adapt and they soon appear lighter.

That's difficult to remember just how bright those balloons should be but I know they were well defined and had full tonality for that kind of lighting.
Logged

ChrisMax

  • Newbie
  • *
  • Offline Offline
  • Posts: 15
Re: 5000 or 6500 K for room light (and monitor)?
« Reply #19 on: July 15, 2017, 11:09:54 AM »

I just had a problem with reddish looking prints and went crazy trying to solve the problem.  Well there this lamp by my desk with a 6400K light and after a few weeks of frustration decided to hold a print under it.  So many prints and lots of ink later under that lamp light the prints looked like what was on the screen!  It was one of those eureka moments.  The bulbs in my basement studio are 5000k and were definitely the cause of the problem.  So I have purchased 6500K bulbs and am with those that say use 6500K lighting.
Logged
Pages: [1] 2   Go Up