Pages: [1] 2 3   Go Down

Author Topic: Luminence value for web output  (Read 4452 times)

michaelbiondo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 126
    • http://michael@michaelbiondo.com
Luminence value for web output
« on: February 20, 2015, 02:25:32 PM »

Ok, I know that this is mostly an unanswerable question but I gota at least try.
Think of it as the photographic equivalent of a controlled crash landing.
Does anyone have any thoughts about what to set  your luminance target for your display in preparing files for websites?
I am calibrating my monitors with  i1 software & hardware.

This is what I set up for targets...

White point = D65
Luminance = 120

Thanks!

MB

howardm

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1834
Re: Luminence value for web output
« Reply #1 on: February 20, 2015, 02:45:30 PM »

Since most people have their monitors turned most of the way up or are using factory defaults, I would probably start
higher than that, 120-140+ and see how that goes.

michaelbiondo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 126
    • http://michael@michaelbiondo.com
Re: Luminence value for web output
« Reply #2 on: February 20, 2015, 02:59:37 PM »

Thanks, I think that is a good starting point.
Most creatives I work with do have their monitors jammed up all of the way.
MB

D Fosse

  • Guest
Re: Luminence value for web output
« Reply #3 on: February 20, 2015, 06:06:19 PM »

Most creatives I work with do have their monitors jammed up all of the way.

Yes, but that's their problem, not yours. I go for 120. If pressed for a standard, that's what most people would agree on.
Logged

Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3944
    • waynefox.com
Re: Luminence value for web output
« Reply #4 on: February 20, 2015, 08:28:21 PM »

Yes, but that's their problem, not yours. I go for 120. If pressed for a standard, that's what most people would agree on.
“Most people” - if they are some how connected to the graphics/photographic industry and are familiar with the challenges of trying to use a computer display as a predictive device for printed output as well as the concept of color management, perhaps.

But that’s a pretty slim number of users out there. If you ask “most people” they would have no clue what you are talking about. And why should the great majority worry about using their displays this way, they have no need to use their systems to “predict” what a print would look like. And there is nothing wrong with using a display at a brighter setting or different white balance. Since that’s how the computer industry delivers the systems, that’s the “standard”, that’s the reality.

Additionally the environment of the computers is dramatically different for most users.  While most doing this type of work have their workstations in room with controlled lighting, the normal display is often in a brightly lit room.

The rationale for not doing anything is there is no standard so how can you do anything. While true, it ignores facts, one of them is 99.999% of the displays out there are much brighter, from 160-220 cd/m2, and often a little cooler.

So one can ignore that fact because there is no standard, or can decide that while there is not a standard, there is one fact, and that is the average conditions are vastly different than the settings we use when trying to use the computer as a device to predict printed output.

There have been discussions on this before.  I use to think there was nothing I should do.  Then I went to a computer store and pulled my gallery up on 10 or 15 different machines.  I started looking at my website when I was at friends homes.  And I realized all my images looked weak and flat.

So to the OP, I have two settings for my NEC, one is for printed output, one is to tweak things for web jpegs.  The web setting is 160 cd/m2, 6500k, sRGB. After experimenting with slight tweaks to my files for converting to web, I have settled on a very slight density increase, and a slight boost to vibrance and saturation. (these are settings within a photoshop action, not LR/ACR adjustments)  Subtle and still look pretty good on my 115 cd/m2 calibrated display.  Went down to the computer store, and things looked a little better. Still not perfect, but that will never happen. And I don’t try to “soft proof” every image, I pretty much just run with those settings.

Logged

D Fosse

  • Guest
Re: Luminence value for web output
« Reply #5 on: February 21, 2015, 06:31:34 AM »

The rationale for not doing anything is there is no standard so how can you do anything. While true, it ignores facts, one of them is 99.999% of the displays out there are much brighter, from 160-220 cd/m2, and often a little cooler.



Yes, but again, these people see everything this way. That's their environment, and if they're not bothered by that, why should your particular images be different?

I think the whole notion that you should try to compensate for the "average" setup is flawed to begin with. Aim for those who have things properly set up, and ignore the rest. I still think it's their problem.
Logged

Rhossydd

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2687
    • http://www.paulholman.com
Re: Luminence value for web output
« Reply #6 on: February 21, 2015, 06:48:13 AM »

I think the whole notion that you should try to compensate for the "average" setup is flawed to begin with. Aim for those who have things properly set up, and ignore the rest. I still think it's their problem.
It becomes your problem if you loose sales/reputation because your images are seen as too dark.
Logged

Simon Garrett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 735
Re: Luminence value for web output
« Reply #7 on: February 21, 2015, 06:51:31 AM »

There have been discussions on this before.  I use to think there was nothing I should do.  Then I went to a computer store and pulled my gallery up on 10 or 15 different machines.  I started looking at my website when I was at friends homes.  And I realized all my images looked weak and flat.

I don't quite understand this.  I assume your images look OK on your monitor (!)  I also assume that other images on the web look OK on your monitor.  So, if you go to another person's PC (uncalibrated, and high brightness), everything will look higher contrast and punchy - including your images, surely? 



Logged

Simon Garrett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 735
Re: Luminence value for web output
« Reply #8 on: February 21, 2015, 06:59:57 AM »

Yes, but again, these people see everything this way. That's their environment, and if they're not bothered by that, why should your particular images be different?

I'm with you on this.

It's sometimes said that there's no point using colour management for images destined for the web.  You get the skin tones spot on, but then most people viewing the web have monitors showing false and unpredictable colour, so why bother?

The answer is this, I think: someone with an unmanaged monitor sees skin tones incorrectly, but they are used to how "correct" skin tones look on their monitor.  Most professional content (news sites, galleries etc) show correct colours, so if yours are also correct, then they'll look "right" on an unmanaged monitor - that is, they'll look the way to which the user has grown accustomed on their monitor.  
« Last Edit: February 21, 2015, 07:21:21 AM by Simon Garrett »
Logged

LawrenceBraunstein

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 83
    • My SmugMug Galeries
Re: Luminence value for web output
« Reply #9 on: February 21, 2015, 07:13:00 AM »

As you can see, there are a lot of different opinions out there. None of them, however, really resolve the problem. I do a lot of printing and therefore prefer to have my main monitor calibrated to a luminance, white point, and contrast setting which best matches the paper/printer combination I use most often. However, I have a second monitor which is calibrated to a higher luminance and contrast setting. This is my ‘web’ monitor. Of course, the down side of all this is that I can’t really use both monitors to compare photographs. This doesn’t bother me so much since the second monitor is used during post-processing primarily to hold the library thumbnails in Lightroom while editing (or printing) photos in the develop (or print) module. Thanks for asking the question, though. I’m curious to read other ‘solutions’ to this unsolvable problem.

With best regards,

Larry
Logged

michaelbiondo

  • Full Member
  • ***
  • Offline Offline
  • Posts: 126
    • http://michael@michaelbiondo.com
Re: Luminence value for web output
« Reply #10 on: February 21, 2015, 01:42:03 PM »

Thanks everyone for their input.
I find myself leaning into the "can't beat em join em camp"
There are just too many clients of mine (and potential clients) using bright settings on their monitors.
I really like the idea of having one luminance setting for printing and one for web viewing.
This may work for prepping files for my website but what about file delivery to clients?
For that it may be a good idea to split the difference, say 180?
Hmm... need to give it more thought
MB



Wayne Fox

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 3944
    • waynefox.com
Re: Luminence value for web output
« Reply #11 on: February 21, 2015, 09:57:11 PM »

I don't quite understand this.  I assume your images look OK on your monitor (!)  I also assume that other images on the web look OK on your monitor.  So, if you go to another person's PC (uncalibrated, and high brightness), everything will look higher contrast and punchy - including your images, surely?  

Actually my images are a little flat on others monitors.  They look fine on mine ... but then that’s only because I’m trying to make them look like printed paper and I’ve gotten use to a display with a lower brightness setting.  The first thing most have to figure out when when trying to use the display to predict printed output is turn the brightness down, which requires a density change in the image file to look correct.  Turn the display back up and it can look a little washed out.

What f I were selling a high end display that was to be placed in a frame and display the images this way ... would I use the same settings and have the same workflow? Probably not.

  We’re the minority here, we are using the devices in a special way for a unique purpose, and as I said the only thing that is really known is pretty much everyone else has their displays brighter.  

And on my monitor there are many images on the web that don’t look OK, because it’s a high gamut NEC and things just get weird sometime when surfing the web.  i understand why, doesn’t bother me.  But if someone is looking at my image and it looks a little washed out compared to what it would look like when printed, they don’t understand that. They just think that’s the way it’s supposed to look.

I don’t worry so much about color ... viewers adapt to color quickly, and often we only think color is “off” because we are comparing it - if they’re not the same then one of them must be wrong, right?.  so I don’t try to adjust things a little ‘warmer” because most displays are a little cooler than mine.
« Last Edit: February 22, 2015, 07:56:28 PM by Wayne Fox »
Logged

Simon Garrett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 735
Re: Luminence value for web output
« Reply #12 on: February 22, 2015, 05:31:50 AM »

Actually my images are a little flat on others monitors.  They look fine on mine ... but then that’s only because I’m trying to make them look like printed paper and I’ve gotten use to a display with a lower brightness setting.  The first thing most have to figure out when when trying to use the display to predict printed output is turn the brightness down, which results in adding density to the image file. Turn the display back up and it can look a little washed out.

Well, several things could be happening here.  On a given monitor, if you calibrate to a higher brightness then the perceived contrast may well go up.  This is simply because peak white gets brighter but black stays the same.  However, it depends on the contrast setting on the monitor (effectively altering the gamma).  In my experience, most people turn the contrast up to get a more punchy image. 

At the same time, if you increase the brightness of an image, the perceived saturation will go down.  This could be one reason why images might look washed out. 

I did some tests: I created two calibrations for my monitor, one sRGB 100 cd/m2 and one sRGB 200 cd/m2.  On my monitor the greater contrast of the higher brightness outweighed the reduced saturation effect, and images did not look washed out at the higher brightness level, and I could not see any reduced saturation.  Images just looked more punchy - the opposite of what you observe. 

I can only guess that, when we start from images adjusted on colour-manged monitors with moderate brightness (around 100-120 cd/m2) then the appearance on general uncalibrated monitors is going to be unpredictable.  In the sample you tried you tend to see washed out colour, but in those I've tried the result has been the opposite. 

Logged

Peter_DL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 523
Re: Luminence value for web output
« Reply #13 on: February 22, 2015, 02:20:59 PM »

On a given monitor, if you calibrate to a higher brightness then the perceived contrast may well go up.

At the same time, if you increase the brightness of an image, the perceived saturation will go down.

I did some tests: I created two calibrations for my monitor, one sRGB 100 cd/m2 and one sRGB 200 cd/m2.  On my monitor the greater contrast of the higher brightness outweighed the reduced saturation effect, and images did not look washed out at the higher brightness level, and I could not see any reduced saturation.  Images just looked more punchy - ...

Agreed regarding the first statement,
from what I can tell:

the higher the monitor luminance, i.e. the closer it is to the scene luminance,
the less of tonal lifting is typically needed, e.g. referring to the common S-curve.

Quote: >>There are some new (and very expensive) display technologies on the market that have a real dynamic range of 10,000:1 and can produce extremely bright whites. With one of these, we could send a properly exposed scene-referred image directly to the display. The image would look just like we were there, …<< (Rendering the print , Karl Lang, see below link).

Images edited at high monitor luminance tend to look dark and flat at low viewing luminance (-> dark prints), whereas images processed at low monitor luminance (for good prints) tend to yield an over-processed, too contrasty look when changing to a high viewing luminance like e.g. with a slideshow on a TV screen.

Not sure though about the second statement regarding "perceived saturation going down with increasing brightness",
is it a known perceptual effect ?

--
http://www.lumita.com/labs/whitepapers/
http://www.lumita.com/site_media/work/whitepapers/files/pscs3_rendering_image.pdf

Logged

GWGill

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 422
  • Author of ArgyllCMS & ArgyllPRO ColorMeter
    • ArgyllCMS
Re: Luminence value for web output
« Reply #14 on: February 22, 2015, 05:31:10 PM »

the higher the monitor luminance, i.e. the closer it is to the scene luminance,
the less of tonal lifting is typically needed, e.g. referring to the common S-curve.
Hmm. I'm not sure that's right.

Here's my hand-waving type explanation:

For a display to be usable, its brightness needs to roughly match that of its surroundings. If a display is much darker than the surroundings, then you will struggle to see any mid tone detail in it. If it is markedly brighter, you get a lot of eye strain.

A display in bright surroundings generates a lot of flare & glare in what we see, partly from reflection from the display itself, and partly because of the optical properties of our eyes. The bright surroundings also set our eye's brightness adaptation at higher level, making us less sensitive to darker tones (we are most sensitive to differences from what we are currently adapted to). So shadow detail tends to disappear, unless it is given a lift. Adding this lift corresponds to displaying the image with a lower power curve or gamma value.

A display in dark surroundings generates much less flare & glare in what we see, and our eye's brightness adaptation is set at a much lower level, so we are much more sensitive to the shadow detail. So if shadows detail is not to be over-emphasized, it needs to be reduced. Making this reduction corresponds to displaying the image with a higher power curve or gamma value.




Logged

Simon Garrett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 735
Re: Luminence value for web output
« Reply #15 on: February 22, 2015, 05:52:49 PM »

Not sure though about the second statement regarding "perceived saturation going down with increasing brightness",
is it a known perceptual effect ?

Surprisingly, yes it is.  Here's a series of strips from 5 photos merged together of the same plain red sheet, each with different exposure:



Note how the darker ones appear to have greater saturation?  If you look at them in Photoshop, as one goes from lower to higher exposure, the saturation goes down and brightness goes up (hue stays the same, obviously).
Logged

Peter_DL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 523
Re: Luminence value for web output
« Reply #16 on: February 23, 2015, 01:26:40 PM »

Here's a series of strips from 5 photos merged together of the same plain red sheet, each with different exposure:
If you look at them in Photoshop, as one goes from lower to higher exposure, the saturation goes down and brightness goes up

... however the red channel is clipped at 255 in the +EV patches.

Could be an issue from gamut conversion, the file as downloaded is tagged with sRGB.

--
« Last Edit: February 23, 2015, 02:31:23 PM by Peter_DL »
Logged

Simon Garrett

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 735
Re: Luminence value for web output
« Reply #17 on: February 23, 2015, 03:12:36 PM »

... however the red channel is clipped at 255 in the +EV patches.

Could be an issue from gamut conversion, the file as downloaded is tagged with sRGB.

--

Hmm... it's OK in Lightroom (no clipping of the red channel, and it's and sRGB jpeg, so should be no gamut conversion issues).  However, I've done it again:



I've loaded the jpeg into Photoshop, and the brightest patch has R=252, G=0, B=55.  The patches are 0.3EV apart. 

The effect perhaps isn't very clear on this, but originally I did this test for a course I was doing (written by Michael Freeman - author of "The Photographer's Mind" and many more).  What he said was:

Quote
What you should be able to see is that the actual colour changes with the exposure. Under-exposure produces a ‘stronger’ colour, and in professional photography this is quite a common technique.

Perhaps more realistic is to take a typical image.  I find that underexposed images often appear more saturated than correctly exposed ones, even if there is no clipping of colours. 

Logged

Peter_DL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 523
Re: Luminence value for web output
« Reply #18 on: February 23, 2015, 04:31:59 PM »

In LR/ACR this impression can result from the default S-curve which has a characteristic side effect on color saturation,
increasing it for the shadows and midtones while decreasing it for the highlights (insofar corresponding to the behavior of a RGB S-curve).
By varying the camera-exposure the data are initially placed under different sections of this curve.

--
« Last Edit: February 23, 2015, 04:38:47 PM by Peter_DL »
Logged

Peter_DL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 523
Re: Luminence value for web output
« Reply #19 on: February 23, 2015, 05:01:33 PM »

For a display to be usable, its brightness needs to roughly match that of its surroundings. If a display is much darker than the surroundings, then you will struggle to see any mid tone detail in it. If it is markedly brighter, you get a lot of eye strain.
...
A display in dark surroundings generates much less flare & glare in what we see, and our eye's brightness adaptation is set at a much lower level, so we are much more sensitive to the shadow detail. So if shadows detail is not to be over-emphasized, it needs to be reduced. Making this reduction corresponds to displaying the image with a higher power curve or gamma value.

Hmm. I'm not sure that's right.

With a scene of broad dynamic range, and a white luminance far above a monitor’s candelas,
all tones usually look dark and flat with a linear, scene-referred rendition on screen:
/> the lower the monitor luminance,
/> and the brighter the surrounding light - the worse it gets,
… and the higher the need for a brightening S-curve, or a more elaborated tone mapping technique which adds brightness and contrast to the mid tones while preserving shadow and highlight details,
in order to obtain a pleasing rendition.

My 2 cents.

--

Logged
Pages: [1] 2 3   Go Up