or Connect
AVS › AVS Forum › Display Devices › Screens › DIY Screen Section › How close is close enough?
New Posts  All Forums:Forum Nav:

How close is close enough?

post #1 of 46
Thread Starter 

I rarely start threads but want to start this one to pinpoint just a single topic. And as totally unlike me I'm not even going to interject an opinion at this point. I'm just interested in what others in the hobby think, and as this has been coming up some in all different threads I thought why not have a place to discuss it.

The question is. How close is close enough? And I'm referring to screen neutrality.

Many here have experienced the benefits of a neutral gray screen and 95% of my posts have dealt with how, why and when I believe a neutral gray screen can work to your advantage.

But my question really is first and simply how close to neutral do we have to be before we can feel comfortable with calling it neutral and secondly more complex perhaps is how do we measure this color. Is 3 points of reference such as RGB enough or should we subject things to spectral graphing etc.

I feel the answer lay in our projectors ability to calibrate somewhat and also our own threshold to perceive.

What do others think?
post #2 of 46
Quote:
Originally Posted by bud16415 View Post

But my question really is first and simply how close to neutral do we have to be before we can feel comfortable with calling it neutral and secondly more complex perhaps is how do we measure this color. Is 3 points of reference such as RGB enough or should we subject things to spectral graphing etc.

As I mentioned in another thread, I had to stop trying to make my Silver Fire mix neutral since adding red to the mix was also increasing the blue content. In the end, I think close is close enough.

Spectral graphing would be the more accurate way to go, but RGB is the closest we can come (using a scanner and Photoshop) without sending samples out to be tested (which I assume is expensive).

Quote:


I feel the answer lay in our projectors ability to calibrate somewhat and also our own threshold to perceive.

I think you hit the nail on the head! Our perceptions will rule in the end, and this brings back a memory: I was visiting a friend years ago and the local news was playing on the TV (a CRT). I took one look at that screen and it was all I could do to NOT try to adjust it! It was so off-color it almost made my eyes water. I mentioned this to my friend and his reply was "Why? What's wrong with the picture?".
post #3 of 46
As long as the theater owner is personally satisfied with the results, the screen is close enough to neutral.

Virtually all of the solutions ever offered here are neutral within the range of calibration for the average projector. I'm not saying that true D65 neutrality is not a noble goal for a screen, but unless the projector is at least tweaked (by eye or THX optimizer) or calibrated (professional instrumentation), only half of the equation is being considered.

Garry
post #4 of 46
Quote:
Originally Posted by bud16415 View Post


I rarely start threads but want to start this one to pinpoint just a single topic. And as totally unlike me I’m not even going to interject an opinion at this point. I’m just interested in what others in the hobby think, and as this has been coming up some in all different threads I thought why not have a place to discuss it.

The question is. How close is close enough? And I’m referring to screen neutrality.

Many here have experienced the benefits of a “neutral” gray screen and 95% of my posts have dealt with how, why and when I believe a “neutral” gray screen can work to your advantage.

But my question really is first and simply how close to neutral do we have to be before we can feel comfortable with calling it neutral and secondly more complex perhaps is how do we measure this color. Is 3 points of reference such as RGB enough or should we subject things to spectral graphing etc.

I feel the answer lay in our projectors ability to calibrate somewhat and also our own threshold to perceive.

What do others think?

As long as you stay within the normal parameters of white, silver, or gray you should be fine. Getting the color balance acceptable is the easy part. Getting the deep blacks, 'pop', and image depth with digital is the hard part.
post #5 of 46
What are the smallest changes in RGB that we can detect visually?
post #6 of 46
Thread Starter 
Quote:
Originally Posted by tiddler View Post

What are the smallest changes in RGB that we can detect visually?

Garry was kind enough to make these gradients for me a while back and posted them into this thread where we had some discussion on this same topic.

http://www.avsforum.com/avs-vb/showt...=662348&page=2 starting at post 59



post #7 of 46
I stumbled on to this site called "Ask a color scientist"
http://www.cis.rit.edu/mcsl/outreach...p?catnum=0#737

It has some very interesting info on it, and it seems we may be barking up the wrong tree using RGB values since they are device-dependant.

Q: Suppose that we want to measure the difference between the color of a test object and a reference object. How would I choose between using RGB color space and CIE Lab for color difference?

A: RGB is a device color space and will vary from device to device. The values have no perceptual meaning and the meaning of differences between them will vary from device to device. CIELAB is a color space that describes color in terms of human perception independent of the device used to create the color. Color differences in CIELAB have perceptual meaning. So, if you want to know about perceived color, rather than an arbitrary color designation, you need to use CIELAB.
post #8 of 46
Just curious to see what people think of these images to test how much difference we can see.




Mind you the quality and contrast of your monitor will effect this too!

The idea would be for you to identify the lowest number you can see. For example I can just see "R4".

In the following slide I can see "r2"

post #9 of 46
I like the concept, but they are also a good example of what can happen to an image when it's processed. Here's same two images, with the background selected:





As you can see, jpg compression has removed a lot of the information! Do you have the uncompressed files?

Garry
post #10 of 46
Quote:
Originally Posted by prof55 View Post
As you can see, jpg compression has removed a lot of the information! Do you have the uncompressed files?

Garry
I created them in PowerPoint and then saved the slides as jpg.

I tried saving them as tif files also:

 

ColorPerceptionTests.zip 262.00390625k . file
post #11 of 46
Here are bitmap versions.



Are they any better?
post #12 of 46
Yes, those are better:



post #13 of 46
Well so far I have learned that blue variations are the hardest to detect.

Also green pushes are the easiest to detect.

Interesting!
post #14 of 46
Thread Starter 
Quote:
Originally Posted by tiddler View Post

Well so far I have learned that blue variations are the hardest to detect.

Also green pushes are the easiest to detect.

Interesting!

That's pretty much what we concluded before when we talked about this and I don't remember all the details but longer wave lengths are hard for our eyes to get details from. Rods and cones and all that stuff

There is defiantly an advantage when missing neutral to error in a certain direction. Some of it I think has to do with what colors we perceive better and another part is what colors in a image trigger our brain to say that's not the right color. Example flesh tones that are too green or blue vs red.
post #15 of 46
The next set of six tests will be two colors varying differentially.

Just a reminder that this is very dependent on people's monitors as well as their ability to discern slight variations in color.
post #16 of 46
Quote:
Originally Posted by bud16415 View Post

That's pretty much what we concluded before when we talked about this and I don't remember all the details but longer wave lengths are hard for our eyes to get details from. Rods and cones and all that stuff

There is defiantly an advantage when missing neutral to error in a certain direction. Some of it I think has to do with what colors we perceive better and another part is what colors in a image trigger our brain to say that's not the right color. Example flesh tones that are too green or blue vs red.

Well I'm not sure we want to be defiant!

That's a good point Bud! So an image like prof55's lady with different color pushes may be more meaningful in trying to decide how close we should be.

I guess what this experiment shows is that if we were to try and compare a gray paint to a reference neutral gray then we would be less likely to detect a blue shift. Most able to detect if it is too green and red falls somewhere in the middle.
post #17 of 46
Quote:
Originally Posted by tiddler View Post

Well I'm not sure we want to be defiant!

That's a good point Bud! So an image like prof55's lady with different color pushes may be more meaningful in trying to decide how close we should be.

I guess what this experiment shows is that if we were to try and compare a gray paint to a reference neutral gray then we would be less likely to detect a blue shift. Most able to detect if it is too green and red falls somewhere in the middle.

Which poses an interesting question: Why do we never hear about anything but a blue push?

My theory is that most people have been trained to think blue-gray is actually gray, and they tend to select it for a "gray" screen.

Garry
ps keep my lady out of this...
post #18 of 46
Quote:
Originally Posted by prof55 View Post

Which poses an interesting question: Why do we never hear about anything but a blue push?

Garry

Well let's get right to the point then. I never observed a blue push so to speak with Behr "Silver Screen". I did observe that it had a redish look to it though. The complaints that I heard were always with regard to reds not looking right. I believe the problems some people observed were due to the variations in the spectral reflectance curve in the red region.

It also has occurred to me that if UPW is blue deficient to start with at 248 248 241 then adding Lamp Black alone should be helping to make it more RGB neutral id indeed Lamp Black will cause it to shift more blue. It will be interesting to see what the color match of the Munsell reference grays reveals for UPW.
post #19 of 46
Might be the terms folks use to express it, too.

A screen with a blue push doesn't look blue under projection. The most obvious tendency of a screen with a "blue push" is to shift reds toward purple. All colors are affected, but whites aren't perceived as blue because our eyes automatically perform a "white balance", and assign the lightest areas of the screens white.

Thus a "blue push" doesn't necessarily look blue...

Garry
post #20 of 46
If you did not respond to the polls in the Color Perceptions threads please do. The more people respond the more meaningful the results.

I will try to get the differential variance tests up over the next few days.
post #21 of 46
Here is a simple monitor calibration site. At least do the Brightness and Contrast calibration.


Click image to go to Display Calibration page.

I should have thought of this before starting the first six tests. Does anyone think we should start over and suggest the calibration first?
post #22 of 46
Quote:
Originally Posted by tiddler View Post

I should have thought of this before starting the first six tests. Does anyone think we should start over and suggest the calibration first?

It could be quite informative to see the differences in response after those with uncalibrated monitors (like me) did the calibrations. I think a new test should be done and only those that have calibrated monitors should do the poll.

In my case, my contrast was OK, but my monitor (a CRT) brightness was just a little dark and I couldn't tell the difference between the black and almost black squares in the gray scale.
post #23 of 46
Quote:
Originally Posted by Harpmaker View Post

It could be quite informative to see the differences in response after those with uncalibrated monitors (like me) did the calibrations. I think a new test should be done and only those that have calibrated monitors should do the poll.

In my case, my contrast was OK, but my monitor (a CRT) brightness was just a little dark and I couldn't tell the difference between the black and almost black squares in the gray scale.

Did you just use the image or click on it to go to the Display Calibration website?
post #24 of 46
Quote:
Originally Posted by tiddler View Post

Did you just use the image or click on it to go to the Display Calibration website?

I went to the website.
post #25 of 46
Is closer than than Da-Lite High Contrast Matte White close enough? Surely closer than the EluneVision and Elite High Contrast Grays is close enough, Eh!?
post #26 of 46
I will try to compile the results from the Color Perception Test soon.

I was looking at some other color match images and noticed that the tolerance that I can detect seem to depend on the background color. With a lighter background I suspect the tolerance of around 3 point s may drop to one or two points. I will try one of the tests with a lighter background and see what happens.
post #27 of 46
Thread Starter 
Quote:
Originally Posted by tiddler View Post

I will try to compile the results from the Color Perception Test soon.

I was looking at some other color match images and noticed that the tolerance that I can detect seem to depend on the background color. With a lighter background I suspect the tolerance of around 3 point s may drop to one or two points. I will try one of the tests with a lighter background and see what happens.

Looking forward to the results.

I don't know if you remember the sample gradients that I had Garry make for me in my thread but I had him do them between some of the common off the shelf paints and then the average of those paints measured RGB numbers. That was also interesting to see.

Thanks for doing those polls.
post #28 of 46
+3 Red
-3 Red
+1 Green
-2 Green
+4 Blue
-3 Blue
+3 Red -3 Blue
-3 Red +3 Blue
+2 Red -2 Green
-2 Red +2 Green
+2 Green -2 Blue
-2 Green +2 Blue

It looks like a spread of 6 points maximum is ok as long as green is not one of the extremes. If green is one of the extremes then a spread of 4 points maximum would be advised.
post #29 of 46
Quote:


Well so far I have learned that blue variations are the hardest to detect.

Also green pushes are the easiest to detect.
Interesting!

Green, in fact, is the easiest to detect because it's the color that controls how much light/brightness that we see. Adjust ONLY the green on your device way up. If you could measure the lumens, they would increase subsantially. Do that with blue, not so much. With red, even less. Regardless of how bad your picture looks....They aren't pushes, they are changes in Luminance.

Quote:


I guess what this experiment shows is that if we were to try and compare a gray paint to a reference neutral gray then we would be less likely to detect a blue shift.......

Ah, see your problem is that you are testing a blue agaist a "so called" neutral background. I think if you test just the background,, you would see that we are more sensitive to the change in that color and not the intensity of the exisiting tests. That's why we always hear about a blue push. I.E.>>>>>

Quote:


Which poses an interesting question: Why do we never hear about anything but a blue push?


Quote:


Well let's get right to the point then. I never observed a blue push so to speak with Behr "Silver Screen". I did observe that it had a redish look to it though. The complaints that I heard were always with regard to reds not looking right. I believe the problems some people observed were due to the variations in the spectral reflectance curve in the red region.

Hmmm, could it be that most Projector Lamps have a red push to them already?...? AND.....people tried to correct by overdriving their saturation to compensate?

Quote:


It also has occurred to me that if UPW is blue deficient to start with at 248 248 241 then adding Lamp Black alone should be helping to make it more RGB neutral

Did you ever think that maybe UPW is balanced already and that's why MANY people use it? Remember adding a little red or green won't be as noticable to the eye. Change or add more blue and you'll see it. OHHHH, Blue Push again.

Why do the numbers have to equal on paper when our eyes don't see the primary colors equally.

Maybe people need to spend some time in the calibration thread and see what really happens when a color is thrown out of balance. with all of the money everyone has spent on experimenting, one could have easily bought any of the home user calibration set ups and not had to worry it their screen was neutral gray or pure white. A person can even get the software free.

I think you guys spend more time talking shop than getting in your cars and DRIVING THEM. For Pete's sake, go get a life...............
post #30 of 46
Quote:
Originally Posted by Highside View Post

Maybe people need to spend some time in the calibration thread and see what really happens when a color is thrown out of balance. with all of the money everyone has spent on experimenting, one could have easily bought any of the home user calibration set ups and not had to worry it their screen was neutral gray or pure white. A person can even get the software free.

I think you guys spend more time talking shop than getting in your cars and DRIVING THEM. For Pete's sake, go get a life...............

Only a few of us who have enjoyed playing around with paints and such have spent a few hundred dollars. To suggest the many who have benefited by only needing to spend $50 for a very good screen should go out and spend hundreds on a calibrator does not make sense. I think you have missed the point.

I find the green and brightness comments interesting. The tints have been decidedly green but did not show a green push, only a surprising brightness and whiteness in the images. Interesting.

I do have a life. It's in my basement mixing paint and discussing whether my gray paint is grayer than his gray paint. And who is this Pete anyway and what color are his walls, ceiling, flooring, and furnishing?
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: DIY Screen Section
AVS › AVS Forum › Display Devices › Screens › DIY Screen Section › How close is close enough?