Gain and other confusing topics.

[wbassett]

I would like to think the gain post went over well, we will cover how gain is measured soon just to cover all aspects of it.

Viewing cone can be confusing too as we have seen, and thanks cynical for the questions and input, I think it cleared up some questions others may have had too.

Let's talk about texture now for a bit.

Texture isn't a bad thing, but excessive texture is. Texture can help with diffusion of light, but when is it too much?

I threw this together and will redo the images, and in color too when I get my Whibal card. I went with black and white because it provides better resolution and eliminates any color saturation that can detract from things.

Anyway, better pics are coming, for now all we really want to see is the texture.

Here is the finish on the Holovega. I used the flash for all pictures because my room setting was too dark to pick up the texture on these samples.

It shows some hot spotting potential with the flash, but again look at the overall surface. Also I apologize again and will get better pictures up in the next day or two.

Now is a shot of a commercial screen paint, and the panel was prepared by the company so this wasn't user made or anything like that, so no user error problems are introduced here. I can attest that you cannot see the texture when an screen image is projected on this surface (more on that in the formal review)

This is significantly more texture than what was seen on the HoloVega screen, but again, it is perfectly fine when I previewed it as a screen sample.

Okay next up, Da-Lite's Matte White screen. When I redo all these images I will include all of Da-Lite's screens as well and in color.


And here is a bit of a teaser... this is one I just painted this weekend.

Compared to the HoloVega the screen above has a very slight texture but you have to hold it within an inch of your eyes to see it, and (another teaser) this was done without spending $60 on a Wagner. Not that there is anything wrong with a Wagner, just that this was so simple and inexpensive it was almost laughable.

Texture... is it good or bad? It's not so much the texture (within reason, too much is bad) but surface sheen.

We've seen three commercial screens above and one had no discernable texture, one had a very slight texture, and the other was quite pronounced, but all perform very well (as we will soon see). As projector resolutions gets higher though texture will start to become a concern. With higher sheens they quickly become noticable.

Here is a sample of what is ideal though what is unacceptable:

[wbassett]

Screen Efficiency is a term I have been seeing mentioned a lot lately. Mostly it seems to be discussed with or as gain. I haven't really run across this term used by commercial screen companies, so there isn't a clear cut definition of this as a specification or measurable characteristic of a screen.

As far as a discussion, if it were a common term I would see it as the overall performance of a screen more than just one component such as gain. Color reproduction and image quality would be the biggest factors of 'efficiency' in my opinion. I actually used this term in the Neutral Gray thread to expressed color reproduction.

To recap the previous discussion about gain; the gain of a screen is used to obtain the recommended level of brightness for a particular setup. As I mentioned in one example of this, for a dedicated Home Theater room with total light control the recommended amount of fL at the screen is 12fL. So if a particular setup which includes the projector, room setting, and screen size yields say 9fL of light at the screen, then a screen with some gain would be required to bring it up to 12fL. If there are ambient light conditions it would need to come up even more.

That isn't an efficiency factor, it is a component of overall screen performance for a given setting. I say given setting because what one person requires may not be the same that another person needs.

So if I were to look at a screen and define efficiency, I would say it is how well a screen reproduces the image and color that the projector is producing. A more efficient screen will have better and more accurate color reproduction and a brighter image before gain is brought into the equation. (As I mentioned, gain is just one component used to meet specific requirements) Neutral is not a color per say. Hues are variants of color and have pigments that absorb certain frequencies of light so that the light that is reflected appears to our eyes as a color.

A neutral palate will absorb all light equally and reflect colors back equally without absorbing more of one color than another. That could be described as being more efficient since more of the projected light is being returned. It also would be brighter based on less light being absorbed. If you were to take two grays that were close to the same shade, but one is neutral and the other is a hue, the neutral would appear slightly brighter in most cases. Certain colors do appear brighter because of the light spectrum that is being reflected back, but keep in mind it is absorbing other colors as well. Also in this case, there would most definitely be a color shift in the projected image that is introduced. So it may have initially seemed brighter, but since colors are also skewed it wouldn't be a very 'efficient' screen as far as accurately reproducing colors.

Aside from that, I don't know of any official definition of efficiency. We'll have to call someone like Wendy and see if they have an official definition of it and if not, what their take on this topic is.

[wbassett]

Let me pose a question for people to think about.

If a projector is setup and factory calibrated to D65, what would be the most 'efficient' screen at reflecting back what the projector is producing? A D75 screen? A D50 or C screen? Or a D65 screen?

Hint: Gain was not one of the options.

I made an analogy of gain being like adding octane boost to your gasoline to improve performance. Some high performance cars require high octane fuel. If you can get gas at the pump with the required amount of octane, then you don't need to add any. Sure adding a slight amount above the required octane level usually does increase performance a little more, but if you add too much, or tried using JP5 jet fuel then you'll most likely cause damage rather than increase performance any higher.

Okay, some may be saying how can gain 'cause damage'? It won't cause physical damage to anything, but it can 'damage' the projected image. As we know the higher the gain is the smaller the viewing cone is. For most Home Theater setups a 180 degree viewing cone isn't a requirement, so a reduced viewing cone isn't always bad. However, there is a point reached when gain starts to introduce a color shift- especially for anyone not directly on axis.

So gain can be looked at as octane boost for a screen. You can add as much octane boost as you want to a Yugo though and it isn't going to win a race against a Corvette using regular gasoline with nothing added... If a screen isn't 'efficient' and doesn't perform well before adding any gain (octane boost ) adding gain isn't going to make it any more accurate and could actually amplify any issues.

		Adding a poly coating did improve Silver Screen, but it wasn't the slight increase in gain that made the difference, it was the slight shift in the color balance.

[Tiddler]

I observed that for some reason the Behr "Silver Screen" tinted UPW #1050 did indeed enjoy a more significant improvement in brightness than a more neutral (Wispy Gray) sample panel. Here is a photo of the two side-by-side to show the difference.



I have often pondered this difference. I agree with wbassett that the neutral gray would initially reflect more light than the SS and therefore the improvement seems less dramatic. I have also considered that possibly the specularity that is introduced by the polyurethane also reflects light before the underlying gray paint has any interaction with it. This would also explain why the combination of the SS base and polyurethane top coat together appear to be more neutral. I don't doubt that the slight yellow tint to the polyurethane also works in conjunction with the SS to create a more neutral gray but I think some of the light gets reflected before the base color even enters into the equation.

[mechman]

'Efficiency' - one of many definitions:

1. the ratio of the output to the input of any system

How would one measure it for a projector screen? You could come up with some wacky formula, but there's really no need. The measurements are in place for the manufacturers to accurately tell customers what their screen's performance will be - gain and viewing angle/cone. They also have the terms retro and angular reflective to describe the screens limitations with regards to set up. Cynical2 and I have had a long pm back and forth with regards to 'efficiency' and my line has always been - there's no way to measure it, so any numbers being thrown out (I've never seen any for any screen) are purely hypothetical. And to me it seems like it's a new 'buzzword' for gain. Buzzwords are, or maybe I should say were, big elsewhere. Things like "Ambient Light Rejecting" and "Light Fusion". Ambient light rejecting is however a characteristic of one manufactured screen - the High Power. And the High Power's qualities limit it to a narrow theater room crowd.

To me, three things can be used to describe a screen (of which I think the manufacturers only use two):

1. Color Reproduction - see neutrality
2. Gain
3. Viewing Angle/Cone

Now if someone were to come up with a wacky formula to add these three things up and spit out a value that can be compared to the user's environment and preference's... then we might be talking!

It's been described elsewhere as how well a light reflects the amount of light coming at it. IE if the projector is throwing 1000 lumens and you get 900 reflected back your efficiency would be 90%. To that I reply, we have gain measurements. Not to mention that a lot of painted and other DIY screens reflect the light back in a scattered pattern as opposed to uniform High Power.

mech

[Tiddler]

I agree that efficiency is not a term used in the projection screen industry. That does not mean it is not a factor. It just means it is not generally relevant to selling projection screens.

It often came up when discussing "light fusion". I do recall someone saying that a good mirror was only 95% efficient.

Here is the a question that may be able to settle this. Is it possible to have two different matte white screens that have 180 degree viewing cones and have one return more light than the other. If so, then that would indicate a difference in gain without changing the viewing cone. The only way I can see to do that is to increase the reflective efficiency. Increases in gain usually result in a narrower viewing cone because of the redirection of light back towards the viewing area. This is usually done through the use of surface sheen.

Now consider the 1xPearl and 2xPeal curves on this gain chart. Even if you say the reference is incorrect it does not matter. These then are simply relative gain measurements.



Again just look at the 1xPearl (yellow) and 2xPearl (cyan) curves.

Note that at around 20 degrees and greater they are right on top of each other. At narrower viewing angles the 2xPearl shows a higher gain. That gain was somehow achieved without robbing the wider angles. I can only conclude that there is a change in the reflective efficiency at angles less than 20 degrees as opposed to the available reflected light being redirected.

Now is this of any consequence to the average person picking out a screen? Maybe not. It may be of interest to DIY screen solution developers or investigators because it represents an increase in image brightness without reducing the viewing cone.

The issue of efficiency is related to a statement that is often made: Gain "Always" comes at the cost of viewing cone. These two gain curves show that that is not "always" true. The 2xPearl obviously has higher gain than the 1xPearl but indeed the viewing cones are identical.

By the way I did not expect this when I sent the samples off for testing. It came as a surprise to me and circumstances were such that there was no real discussion or investigation of this at that time. It will be interesting to see if future tests return similar results.

There are wall paints on the market that claim to be brighter than others. I believe some of them employ things like mica and ceramic micro spheres. Question is are they more efficient or are they just redirecting more light back to the source.

[wbassett]

tiddler wrote:

I agree that efficiency is not a term used in the projection screen industry. That does not mean it is not a factor. It just means it is not generally relevant to selling projection screens.

No disagreements there at all! It is a factor, but as mech pointed out there is no known way to measure this, and it's a good guess that even commercial companies would have a different 'definition' from one company to the next- so DIY developers also most likely have their own interpretation of 'efficiency'.

tiddler wrote:

Here is the a question that may be able to settle this. Is it possible to have two different matte white screens that have 180 degree viewing cones and have one return more light than the other. If so then that would indicate a difference in gain and no change in viewing cone. The only way I can see to do that is to increase the reflective efficiency. Gain is usually increased resulting in a narrower viewing cone by redirection the light more back towards the viewing area.

By definition, a unity gain screen has even light dispersion across the entire surface of the screen and a 180 degree viewing cone. With that in mind, if one screen is brighter, then yes it would indicate a difference in gain (but making it no longer a matte white unity gain screen). If the gain is minimal, say a 1.1 vs a 1.0 gain screen, there really wouldn't be a perceivable decrease in viewing cone to our eyes even though it would decrease slightly. This is just an example for discussion, but if a 1.1 gain screen has a viewing cone of 170 degrees vs 180 degrees for our matte white unity gain screen, I really doubt we would pick up on the decrease in cone, but it will still be there. Whether we see it by eye or not, the math still has to balance out. So does it make a difference? In this case no, the viewing cone for 1xPearl and 2xPear would by all intents and purposes look the same. If there was a larger difference in the gain between the two, then yes we would see it.

tiddler wrote:

Again just look at the 1xPearl (yellow) and 2xPearl (cyan) curves.

Note that at around 20 degrees and greater they are right on top of each other. At narrower viewing angles the 2xPearl shows a higher gain. That gain was somehow achieved without robbing the wider angles. I can only conclude that there is a change in the reflective efficiency at angles less than 20 degrees as opposed to the available reflected light being redirected.

The difference in gain is very slight, from the chart around .05 and that's not much at all. A .1 increase in gain, depending on the screen size can be as little as a 1fL of increase in brightness. That is based on a 106" screen size. For smaller screen sizes a .1 increase in gain can be 2fL or more of an increase, for larger size screens it can be less than 1fL increase. So there are a lot of factors to take into consideration.

I don't agree that the gain was achieved without robbing light from elsewhere, it's just that it's not a major increase in gain so as mentioned earlier we probably wouldn't see any change in the perceivable viewing cone, but again it is still there. We can't make more light than what originated from the projector, so if a screen is brighter on axis, that light did come from elsewhere even if our eyes don't pick up on that. Math and physics... the equations must balance out...

tiddler wrote:

Now is this of any consequence to the average person picking out a screen? Maybe not. It may be of interest to DIY screen solution developers or investigators because it represents an increase in image brightness without reducing the viewing cone.

I agree and think you nailed it, this is probably more of a concern and interest to us developers than the average person. Good call!

tiddler wrote:

The issue of efficiency is related to a statement that is often made: Gain "Always" comes at the cost of viewing cone. These two gain curves show that that is not "always" true. The 2xPearl obviously has higher gain than the 1xPearl but indeed the viewing cones are identical.

Again, and not trying to argue, but for the example in the chart shown we're not talking about a significant increase in gain so I really doubt there would be any perceivable change in viewing cone. In this example we're looking at a half gain with 1xPearl of .45 which is at roughly the 45 degree mark, and for 2xPearl a half gain of .48 which is round the 40 degree mark. I really doubt anyone would be able to see the difference by eye, but as I keep mentioning, it is still there.

[mechman]

And then we get into the other factors - Color Reproduction. The pearls have brighter whites on axis but your off axis shots show the screens with pearl and the ones' without looking pretty much identical. At which angle does that happen?

And you're trading the blacks for the whites. Whites always look white to me on my FG screen up until I either put a lighter gray up or a white screen up. Then they don't. Conversely the blacks always look great on the FG when compared to the white screens. And in my mind, with just a touch of ambient light the blacks take a hit ont the blacks. This is a trade off though - user preference.

Now the pearl question is, why do they make the shots seem more bluish. Is it a color shift or not.

tiddler wrote:

The issue of efficiency is related to a statement that is often made: Gain "Always" comes at the cost of viewing cone. These two gain curves show that that is not "always" true. The 2xPearl obviously has higher gain than the 1xPearl but indeed the viewing cones are identical.

This is true, yet when gain is usually being discussed it usually revolves around better gain figures than <1. I would think one could go from a .7 gain to a .95 gain by using a lighter shade of gray. Or by taking their quart of N9 and mixing it with a neutral white to make a N9.5 or even a N9.8.

mech

[wbassett]

I want to take a brief time out and say I think this is a very interesting discussion and debate.

I hope nobody takes anything personal or offense at some of the Point Counter-Point discussions.

There are a lot of topics that people find confusing, and as developers there are a lot of questions we need to investigate and find answers to. So let's not get upset or take things personally, not that anyone has yet, just reminding everyone this is a discussion is all.

So far I think everyone is bringing up some very good points and this is a very intelligent and civil group discussion.

[Tiddler]

wbassett wrote:

I don't agree that the gain was achieved without robbing light from elsewhere, it's just that it's not a major increase in gain so as mentioned earlier we probably wouldn't see any change in the perceivable viewing cone, but again it is still there. We can't make more light than what originated from the projector, so if a screen is brighter on axis, that light did come from elsewhere even if our eyes don't pick up on that. Math and physics... the equations must balance out...

You are forgetting the light that is absorbed even by a white screen. Efficiency is the ratio of reflected light to incident light. So a 90% efficient reflective surface absorbs the 10% of the incident light.

I do not recall any RGB numbers for any of the white surfaces being 256 256 256. As a matter of fact, UPW which is considered to be one of the whiter paints, all be it with a slight blue push, has RGB numbers around 235 to 245 does it not. If we assume it is 240 then it is reflecting 240/256 or 93.75% efficient.

If we are talking about a simple surface like paint then the less efficient it is the darker gray it looks right. I think the pearl clear coats are introducing something that is not quite so linear. Like car window tinting. This even more evident on a black car with a nice clear coat or metallic gel coat. The car is no doubt black but the sun reflecting off of it sure does not look black. In many cases it is so bright you can't actually look at it. Yet when you hit the same car with a more diffuse or lower intensity light it looks black.

No I don't think it is magic but these sorts of interactions are currently unexplained in any scientific way, by us to us that is. In fact I still don't accept the notion that the simple polyurethane top coat is just adding some surface sheen. I don't deny that it may well be introducing surface sheen but that is not the significant contribution. I say this because the effect seems to be dependent on the underlying surface roughness. The poly works most effectively over a very rough flat finish like UPW #1050. I still believe the significant action is taking place at the poly paint interface. Similar to getting a colored stone wet. The color darkens and the variation in colors become more apparent.

That reminds me, the fellow at the Behr Color Lab also indicated that there is more to it than that. I never did get his take on it. I should try to get him to comment on it.

Anyway while related to all this gain stuff surface wetting is even more of a stretch for people to grasp than increasing reflective efficiency.

It is an interesting discussion and I don't disagree with any of the basic concepts being put forward, I just think there may be more to it than we currently understand. Just like I know a one way mirror looks like a mirror on one side and like tinted glass on the other, but I don't really know why. Yes I know that the mirror side has to be brighter than the see through side and that if you swap which room is dark and which is light the one way direction will change but I don't know why.

[wbassett]

tiddler wrote:

You are forgetting the light that is absorbed even by a white screen. Efficiency is the ratio of reflected light to incident light. So a 90% efficient reflective surface absorbs the 10% of the incident light.

Actually I'm not forgetting about light being absorbed. If we looked at this as a math equation the absorbed light would be one of the variables of the equation. Again, a screen cannot create light and that isn't something I am making up just to be the devil's advocate.

As I demonstrated in the earlier post, there is a slight increase in gain, and a slight decrease in viewing cone.

I think what you are saying is the 2xPearl looks to be brighter than what a .05 gain would tend to indicate. There's no doubt that specular gain is increasing too, which could be a contributing factor as well. We do have to be careful though because if we bring the specular gain up too much then we stand the change of hot spotting.

We do know and can explain some things going on. Pearl is mica based and mica acts as a bunch of tiny mirrors, or a more accurate description would be prisms. They will have an affect on the light source, but they can't produce light and that is what I am saying.

You mentioned optical texture, and the mica flakes are doing that by diffracting and scattering the light more than what the paint itself is doing.

You're right, there are a lot of interesting things to delve into and find out what's going on and that's what we plan on doing.

[Tiddler]

I should also point out that even if we did make the screen more efficient, if there is such a thing, it would be a pretty small piece of the pie.

So what do you think the ceramic micro-spheres are for?

[wbassett]

tiddler wrote:

I should also point out that even if we did make the screen more efficient, if there is such a thing, it would be a pretty small piece of the pie.

So what do you think the ceramic micro-spheres are for?

Don't forget, you did ask!

What do I think they are for? Same as Alumunim Oxide and Titanium Dioxide... to make a brighter white/base and as a replacement for mica.

This should be in the teaser thread, but I have over five pounds of Titanium Oxide on its way to me, and I've had Aluminum Oxide for a little while now.

[Tiddler]

wbassett wrote:

. . . and I've had Aluminum Oxide for a little while now.

You might want to have a doctor take a look at that! :raped:

[wbassett]

My wife would tend to agree with you sometimes!

[mechman]

tiddler wrote:

You are forgetting the light that is absorbed even by a white screen. Efficiency is the ratio of reflected light to incident light. So a 90% efficient reflective surface absorbs the 10% of the incident light.

And this is measured and noted as gain. Light reflected is measured as gain. There's no need to coin a new phrase for something that already exists.

I do not recall any RGB numbers for any of the white surfaces being 256 256 256. As a matter of fact, UPW which is considered to be one of the whiter paints, all be it with a slight blue push, has RGB numbers around 235 to 245 does it not. If we assume it is 240 then it is reflecting 240/256 or 93.75% efficient.

Would it really though? How much of that goes back at the viewer? The High Power does not appear to be anywhere near a 255 white nor near UPW. So if it falls around 220 would we say it is 86% efficient then?

If we are talking about a simple surface like paint then the less efficient it is the darker gray it looks right.

This depends upon your definition of efficient. What's being tossed around here talks about light but nothing with regards to the color of the light. So I wouldn't agree with that at all.

I think the pearl clear coats are introducing something that is not quite so linear. Like car window tinting. This even more evident on a black car with a nice clear coat or metallic gel coat. The car is no doubt black but the sun reflecting off of it sure does not look black. In many cases it is so bright you can't actually look at it. Yet when you hit the same car with a more diffuse or lower intensity light it looks black.

No I don't think it is magic but these sorts of interactions are currently unexplained in any scientific way, by us to us that is. In fact I still don't accept the notion that the simple polyurethane top coat is just adding some surface sheen. I don't deny that it may well be introducing surface sheen but that is not the significant contribution. I say this because the effect seems to be dependent on the underlying surface roughness. The poly works most effectively over a very rough flat finish like UPW #1050. I still believe the significant action is taking place at the poly paint interface. Similar to getting a colored stone wet. The color darkens and the variation in colors become more apparent.

How would you explain Fashion Grey then?

That reminds me, the fellow at the Behr Color Lab also indicated that there is more to it than that. I never did get his take on it. I should try to get him to comment on it.

Anyway while related to all this gain stuff surface wetting is even more of a stretch for people to grasp than increasing reflective efficiency.

It is an interesting discussion and I don't disagree with any of the basic concepts being put forward, I just think there may be more to it than we currently understand. Just like I know a one way mirror looks like a mirror on one side and like tinted glass on the other, but I don't really know why. Yes I know that the mirror side has to be brighter than the see through side and that if you swap which room is dark and which is light the one way direction will change but I don't know why.

As stated above:

To me, three things can be used to describe a screen (of which I think the manufacturers only use two):

1. Color Reproduction - see neutrality
2. Gain
3. Viewing Angle/Cone

Now if someone were to come up with a wacky formula to add these three things up and spit out a value that can be compared to the user's environment and preference's... then we might be talking!

Come up with a formula for that and then we can talk efficiency, otherwise you're sugar coating what's already there with a 'buzzword'.

Maybe as we move along here we can come up with a formula that you can plunk all that data in, add some viewer expectations and get an 'efficiency' score. But in the past this term was tossed around a lot to avoid getting real measurements and because it sounded 'professional' to new folks. Now don't get me wrong, I have no problem with anyone using the word. I just think that it's being used incorrectly.

mech

[wbassett]

mechman wrote:

Maybe as we move along here we can come up with a formula that you can plunk all that data in, add some viewer expectations and get an 'efficiency' score. But in the past this term was tossed around a lot to avoid getting real measurements and because it sounded 'professional' to new folks. Now don't get me wrong, I have no problem with anyone using the word. I just think that it's being used incorrectly.

mech

I agree about terms getting thrown around at some other places and I tend to think gain is one of them that has been misrepresented for a long time. Broken record time, gain isn't a universal requirement, but more of a component used if needed to meet recommended specifications.

[wbassett]

Yes I know there is a thread completely devoted to grays, but this still seems to be a confusing topic.

A lot of what is being used is actually a color that we have been conditioned to think is gray. A neutral gray really isn't a color, which is what we mean when we say neutral.

There are many ways to make gray. The simplest and most basic is to add black and white. Obviously the more black that is used, the darker the gray will be. The problem is the pigments used in paints aren't pure black or white. Lamp Black tends to have purplish blue tone to it, like ink. So with this method we would get a color that looks gray because that's what we have always been told what gray looks is supposed to be.

Most paint manufacturers add other pigments to their grays to make them more appealing. Some companies lean more towards blue, others green, but the point is these are color hues and not a true neutral value.

Does all of this matter? Some say yes, others no. Some say close enough is okay, others say it's only acceptable. We do know that a projector can be calibrated and tweaked to look good on a variety of different screens, so yes acceptable is fine and will work. The question that comes up though is are we striving for acceptable or for the best that we can achieve? I would have to say most if not all of the DIY developers strive for the best.

So what would be the best? I posed this question before to get people to maybe think about this some:
Since projector manufacturers calibrate and QA their units to D65, and all the people that do professional reviews on projectors always check the pre and post calibration in reference to D65, here is the question-
Which screen would be the most accurate at reflecting an image the closest to what the projector manufacturer designed the projector to be optimal at?
A Screen centered on

• Illuminant C
• Illuminant A
• Illuminant F2
• Illuminant D65
• Illuminant D50

The answer should be pretty obvious since D65 is mentioned so many times throughout this thread and forum. But why?

Now that is a good question. We agreed a projector can be calibrated over a fairly large range. So yes we could calibrate to something like say C, and get a decent picture. Here's where the problem starts coming into play. Everything from the movie studio to the projector manufacturers are referencing D65, so even though you can get an image to look 'okay' it is not going to be optimal to what the director wanted the movie to look like, or how the projector was intended to present the movie or content. This difference may seem so slight to some that it's no concern. To others though it does make a difference seeing things as they were intended and with the best and most optimal image possible. The only real way to accomplish this is to keep everything at the same reference point, in this case D65.

Also by moving your projector settings away from its optimal setting, it will be slightly dimmer and less vibrant. In some cases colors can be way off depending on how much of a color push the screen has.

The next obvious question is what constitutes a D65 neutral gray? As measured by a spectrophotometer, the 'a' and 'b' color channels would both be zero for an absolute neutral, but absolutes are not realistic so a tolerance is established of +/- 1 from zero. So for instance a L*ab value of N8 would be around 80 0 0 where 80 is the Luminance value and both 'a' and 'b' are completely devoid of any color. An acceptable tolerance then would be 80 0 0 to 80 +/- 1 +/- 1. (Note: A luminance value of 81 is still in the Munsell N8 range).

Did I lose anyone? For most people this is way beyond what they care about because they just want a screen and not a color science lesson, but it is important to know that there are some tolerances that we can adhere to. These tolerances are something that weren't just made up, it's an accepted tolerance within the color industry.

Can we make a gray like this with just Red, Blue, and Green paints? Certainly. It's not 'hard' just tricky in the sense that there are some tight parameters that need to be maintained. First since we are talking about D65, the colors should be tested under D65. From here on out that will be an assumed given.

In the table below cells in the same row have the same color balance, only the intensity changes. All the colors in the first row are red, and red only with no trace of blue or green. Same as the other rows, the next row is green and only green, and so on.


The bottom row is what happens if all three colors are combined. We get grays! Now, this is absolutely critical and the ONLY way this will work. The colors must not only be pure colors, which we don't get with house paints, and the other major key, they MUST have the same luminance value. If any of the colors are off and have a different luminance intensity, you won't get neutral gray. If the color isn't a pure color, you won't get neutral gray. See how tricky this is? Yes it can be done, but the colors need to be the same intensity and purity. In addition to that, they all have to be equal amounts too, the guesstimate method isn't going to work. Again it can be done, but now you know how tricky and tight the tolerances are. There is a problem with using just Red, Blue, and Green though that will be addressed in an upcoming post.

As they say on Monty Python- Now for something completely different...
http://www.paintspot.ca/cgi-bin/paints.pl?t=2&g=91
All the Munsell's anyone would ever want! The problem is these don't exactly roll well.

Take a look at the paint list though. Can you tell what the luminance value is for any of those reds blues or greens? That's where the problem is. Now if these paints were tested by running them through a spectrophotometer, we could find out and match the colors that are pure and have the same intensity(luminance) and would know for sure we would have a neutral gray when they are combined.

Here is another concept- It is from the book Blue and Yellow Don't Make Green by Michael Wilcox.

Theoretically, when three PURE primary colors are mixed in equal intensities, only a tiny proportion of the light energy reaching the surface is reflected, while the rest is absorbed. The result is a very dark grey, approaching black.

In the same way, when two PURE primary colors are mixed in equal intensities, they result in a very dark grey, almost black. Why? Take the example of a PURE yellow and PURE blue. First, yellow pigment absorbs all light except the yellow. In a same way, the blue pigment absorbs all but the blue portion of the light. Mixed together, the yellow pigment absorbs all the blue light and vice-versa, therefore resulting in dark grey, approaching black.

This is why the book states "Blue and Yellow don't make Green."


So there are several ways to produce a neutral gray, which one is the best though? Well again that's a topic many people have debated as far as DIY screens go, but the answer really is simple. Neutral is neutral. It doesn't matter how it's achieved as long as when it's tested with a spectrophotometer it shows it is neutral.

Before moving on I want to address a statement made above from the book except "...only a tiny proportion of the light energy reaching the surface is reflected, while the rest is absorbed." This tends to make neutral sound bad, as if it will suck all the projected light in and not return anything. Keep in mind that a projector is a powerful device and emits a tremendous amount of light. Even with lower Lumen rated projectors the light is strong enough that if you were to look directly into the lens at close distance you could cause permanent eye injury- So never look into your projector lens while the unit is on!

Since we're hitting the screen with this much projected light, the screen color cannot absorb all of the light energy and a large percentage is reflected back. The darker the gray, the more light energy that is absorbed.

Here is the key though, since a neutral gray will absorb colors equally, it also reflects the excess light energy back equally, meaning more accurate color reproduction. A color hue (non-neutral) will absorb additional light energy based on the strength of the color shift. This not only has an effect on the reflected image's color balance, but it does rob some more light from the image brightness. Again, is this enough to matter? Depends. If the color hue is close to our D65 reference point probably not enough to be noticeable. The further away it gets though the more impact it will have.

For those with projectors that are proverbial 'light cannons' this isn't as much of an issue as for those that have lower Lumens that are trying to eek out every drop of brightness they can muster.

Again lots of ground has been covered in a short span, but to recap:

• Neutral is more efficient at reflecting back the full color spectrum that the projector is creating
• Neutrals can be made several different ways, but we can't tell by eye and they must be tested.
• We already have several available now.

This is sounding pretty difficult isn't it? Not really. We have many Off the Shelf paints readily available that are D65 neutral, and more and more are being added all the time- so they really are something anyone can obtain without resorting to anything complex.

Soon we will have Munsell matches from just about every major paint manufacturer. There are some that have already been identified for a few manufacturers, and those that don't have any that meet the specifications, we'll have a color match done to the Munsell Neutral Gray fan deck shown to the left. Just because we color match though, we're not going to stop there, they will also be tested to confirm their balance.

I did not mean to side track this thread and turn it into another 'gray' thread. This is just another topic I see many people ask a lot of questions about and I wanted to throw together some quick information on the subject. For those interested in more information and discussions about grays you can check out the Neutral Gray thread.

Once the Fan deck has been color matched, these should work not only as excellent stand alone screen applications, but also as the base for advanced and more complex applications. Part of what I see with the advanced mixes are components intended to try and make a balanced or 'neutral' gray, so why not just start with the known neutral to begin with?

[wbassett]

This is a repost from the Neutral Gray thread, but since I just did a quick run down on grays I thought this was a nice follow up.

For those building a Home Theater room from the ground up, this is easier than for those that are working with an existing living room, but it is still very easy and well worth the little extra time even with an existing living room setup.

It is highly recommended that even for a living room that is going to be more of a multi-function room than a dedicated Home Theater to still paint the walls a darker color than the typical Off White that most walls are painted. This does a couple of things, first the darker colored walls help tone down any light reflected from the screen itself. The last thing anyone wants is to create their own ambient light from reflections off of white walls. Next it also helps to make the screen image look more vibrant and for us to see more depth and a vivid image. It is similar in principle to the black border adding to our perceived black levels and image contrast. Lastly... a nice Burgundy or other darker shade can add a flair of style to the room and decor.

If you decide to repaint the entire room as suggested, before painting the wall where the screen is going to be, and this applies to whether a person decides to go with a substrate, or to paint the screen directly on the wall, put a nice coat of Kilz2 primer on the wall. Why not? It certainly isn't going to hurt the final wall color the room is being painted, and it gives you a nice white reference screen. All the projector companies I talked to do their testing and setup calibration on a plain Jane white unity gain screen. Kilz2 will give you a nice white unity gain (1.0) screen, or in this case a wall with a nice white surface that you can test your projector on. Kilz2 is pretty inexpensive and this step is well worth it.

Turn the projector on and project an image on the wall, preferable a solid blue image, but a welcome screen such as this example will work.

Select the desired Aspect Ratio dimensions on the projector if applicable, then use the zoom function to set the screen size on the wall to the exact size you prefer.

















Once the screen placement and ratio have been setup, calibrate the projector. Most people neglect to calibrate and check the initial projector performance and limitations on a reference screen. This will show you exactly how well your projector performs and give you a baseline. AVIA or Digital Video Essentials are the two calibration discs most people use, but the THX Optimizer that is included with every THX certified DVD works nicely too if you don't have either of the calibration DVDs mentioned.

Here is a sample of one of the calibration screens.

Now sit back and watch some content. Make sure to watch both DVDs and any sports or television channels you like. Basically, spend a few nights watching the type of content you plan on using the projector for. This part is very important in determining what it is you want to do with your projector, and if there are any weak areas that could use some help with the right screen.

If viewing is going to be done only at night, or total light control is possible, this will show how well the projector black levels are as well as the color. If you are satisfied that the blacks are black and the image looks good then a white screen will work fine for your setup and environment, you're done- paint the wall the color of the rest of the room and select a white screen. You now know what your projector baseline is and if you are happy with the blacks, there is no reason to go any further with grays.

A lot of projectors, especially older ones (even ones that are only a year old too) have trouble with black though. If you feel even with the lights off and in total darkness the blacks look more gray, or just aren't what you expected and want, then even with total light control you may want to look at a gray screen. A light gray can have a dramatic effect on the black levels. The most annoying area on the projected image that this shows up is in the letter box area. If this area doesn't look black, it can be very distracting to some people. With my white screen, even with light control I wasn't happy with the letter box area and was going to build a complex masking system. Once I switched to a gray screen, there is no longer a need for a mask.

So if a gray seems to be in your future, which gray and shade? An N9 shade of gray will punch up the black level a surprising amount over a plain white screen. This may be all the darker in shade you want or need to go. Projector Lumen rating comes into play with grays too. What you are looking for is the point where the image has dark blacks, the whites remain white, and there is no loss in the color vibrancy or shadow detail.

If there will be times when some lights are on in the room, an N8 shade will help with the ambient lighting. Two very inexpensive ways to quickly test what shade is optimal are Winter Mist for an N9 shade, or Winter Mountain for an N8 shade. (Both of these are True Value colors and are $7-$9 a quart) I am not sure if Tiddler broke down the EasyFlex colors to what their Munsell rating is, but those tints can certainly be used as well. Simply use a pencil to mark off the corners of the screen image area, and then use painters tape to mask the entire area off. Roll on two coats of whichever shade you want to try (I would start with an N9 shade first) and then watch some movies again when it is dry. Don't forget to calibrate the projector for the darker shade or you won't get an accurate idea of the performance. (Note: with the right gray, you shouldn't have to make changes to the color balance, only he brightness and contrast levels. If any color balance changes are needed, they should be very slight or else the gray is shifting the colors) If this still isn't producing satisfying black levels, repeat the process with the N8 shade. For total light control though, I honestly can't see going beyond an N8.5 shade.

Projectors under 1000 Lumens should stick with an N9 shade of gray or white screen. 1000 to 1500 raw lumens can handle an N8 shade, and 1700 lumens and over can go to an N7 shade, but N8 seems to be optimal for both daylight and lights out performance for 1700 and higher lumens. Also keep in mind that there is a difference between raw lumen ratings and video optimized lumen output. My projector is rated at 1700 lumens, but video optimized it is more like 500 lumens. I am working on a lumen chart that should make all of this easier for people.

This does seem like a lot of work, but in the end it is really worth it to know your projector and have a baseline on it no matter what screen method you go with. One huge advantage of DIY is the cost factor. It would be extremely cost prohibitive for a person to get a white screen, calibrate, and decide they don't like the performance... then get a grey screen like a GrayHawk, only to find out that is too dark for their tastes... the process can go on and on. Unless the company you are buying your screens from has a very liberal return policy, it could start running into quite a bit of money. Even for those wanting a commercial screen, this method is a very easy and inexpensive way for them to narrow down what shade they like the best, so they would only have to make a one shot screen purchase... however most people end up amazed at how well DIY screens perform and opt to just stick with it rather than spend hundreds, even thousands on something that looks the same, or maybe just marginally better.

Here are some more comments and a little more information on Grays and their use even in total light controlled situations.

[wbassett]

The Problem with Iridescence

Main Entry: ir•i•des•cence
Pronunciation: "ir-&-'de-s&n(t)s
Function: noun
1 : a lustrous rainbow-like play of color caused by differential refraction of light waves (as from an oil slick, soap bubble, or fish scales) that tends to change as the angle of view changes

The word is derived in part from the Greek word iris (pl. irides), meaning "rainbow", which in turn derives from the goddess Iris of Greek mythology, who is the personification of the rainbow and acted as a messenger of the gods.

		Snell's Law Refraction is the bending of the path of a light wave as it passes across the boundary separating two media. Refraction is caused by the change in speed experienced by a wave when it changes medium.

Pearlescent pigments manipulate light through interference, creating iridescent shifts of color similar to those seen with oil slicks, soap bubbles and peacock feathers. The most widely used pearlescent pigments are mica platelets coated with precise layers of iron oxide and/or titanium dioxide.

The occurrence and distribution of iridescent colors and the various theories regarding their manifestation in the natural world were for a long time discussed by scientists. Even Sir Isaac Newton in his book Opticks, published in 1704, put forward a reason for the iridescent nature of the color from the feathers of peacock tails. The word iridescence is itself defined as the change in hue of the color of an object as the observer changes viewing position.

An example of this iridescent color is the appearance of light reflected from an oil film on the surface of water. If the oil film is viewed from different angles the colors appear to shift and change. Iridescent colors are particularly striking for the observer. They can be among the purest and most brilliant and cannot be matched by even the brightest pigmented colors in their depth and intensity.

Warning, technical (boring) stuff!

Through our increased knowledge of light in general we have also gained additional insights into interference since this phenomenon was discovered three hundred years ago. The credit for formulating the principle of iridescence, or the "colors of thin films" as he called them, goes to Robert Boyle, contemporary and compatriot of Sir Isaac Newton. More than a century later, interference was explained in detail by the English physicist Thomas Young. According to Young's definition, very thin plates or films, such as a coat of oil on water, or the "skin" of a soap bubble, reflect some of the incoming light from their mirror like top surface. The rest of the light travels through the film and is then reflected by the lower surface. The light that enters the film is bent and deflected from its path by the film's greater density, or refractive index.

What actually happens when light enters a medium with a greater refractive index? The movement of the light is slowed down, the waves become smaller, and the traveling distance within the film is increased. In other words, the light cannot go straight through. After the light reaches the lower surface and is reflected, it has to make the return trip through the film, still at the same slowed-down pace. When this light finally rejoins the light that is reflected at the upper surface, it is out of phase.

How much out of phase it is depends on the distance it has had to travel within the film, which in turn depends on two factors: the thickness of the film, and the angle at which the light strikes the surface. If the phase difference between the beams reflected at the upper and lower surfaces happens to equal one full wavelength, or a multiple thereof, that particular wavelength will be reinforced when the light that went through the film rejoins the upper surface reflected light. Reinforcement will be strongest and the color purest if the waves happen to have the same amplitudes, that is, if their crests and troughs are of equal height. All the other wavelengths are either weakened considerably or, where crest meets trough, eliminated. This, of course, results in only one color becoming visible.

For instance, assume that a green ray, after having traveled through the film and back, has a phase difference of two full wavelengths when it rejoins the green light reflected at the upper surface. It is now in phase with this green light. The green color is therefore reinforced. Because the wavelengths of the other colors are different, they are by necessity out of phase and are therefore either eliminated or very much weakened.

However, if the phase difference between the two green light rays were one and a half wavelengths, they would neutralize each other and become invisible. In that case, some other color would appear.

Whew Right?! Even Jordie had a tough time with that one. It means as light travels it loses energy. When it passes through a medium, it not only loses more energy, but is bent (refracted) and unless the returning light is in phase, a color shift will occur. That's not supposition, it's the physics of light. (I recently finished a college course on this too, and can say many brilliant people have studied this)

[wbassett]

Pearlescents work by the process of interference. Nature was the first to make interference pigments. Mother of pearl is an excellent example. Some seashells also have the interference effect. Interference is the separation of white light into its component colors, like a prism. By making a series of thin layers that are clear and refract light, an interference pattern is set up. Sparkle and shimmer are two ways to describe pearlescence.

Manmade pearlescents are made from the mineral mica. The mica is processed into small particles and then coated with a very thin layer of titanium dioxide. The layer is so thin that it actually allows some light to pass through instead of acting like a normal pigment when it is used in paint.

What is Mica?

Mica is a natural valuable mineral product applied to a group of a complex aluminosilicate minerals having a sheet or plate like structure with different chemical compositions and physical properties. All micas form flat six sided monoclinical crystals with a remarkable cleavage in the direction of the large surfaces, which permits them to split easily into optically flat films, as thin as one micron in thickness. When split into thin films, they remain tough and elastic even at high temperature.

Physically Mica is transparent, optically flat, easily splittable into thin films along its cleavage, colorless in thin sheets, resilient and incompressible.

The problem: each layer reflects a small percentage at the front face of the particle, most of the light is transmitted (passed through) and refracted. When light hits the back surface of the particle, a small percentage is again reflected and most is transmitted. If the particles get stacked one upon another, the number of light reflections gets to be quite large. The resulting sparkle looks impressive. The down side is, color shift and light separation. This looks great on cars, but not projection screens.

Another problem is that by nature of iridescence and pearls as viewer’s changes angles, there will be changes in the color. This is caused by the alignment of the mica flakes and their properties. If the flakes could be aligned perfectly and all in the same orientation, there still would be color shifting from off angles but the main angle should appear bright and relatively accurate picture wise. If the flakes are not aligned perfectly or of consitant size, the color shifting could be noticeable even on axis.

Angles of View wrote:

An excerpt from Da-Lite’s Angles of View
If we again go back to Volume 1 in Angles of View, we learn that these types of screens have a diffusive base with platelets of mica strewn across its surface in a regular fashion. These crystals are also coated with Titanium Dioxide (TiO2), which then makes them highly reflective and behave like thousands of tiny little mirrors. We also learned through this article that these materials reflect light incident to their screen surface in a fashion that is equal but opposite the angle of incidence. Keeping that in mind, if one of those particles land on the screen surface at a very severe angle, this is one potential cause of a bright spot or sparkle, depending on your viewing position. The second potential issue is if the particles are too large and do not allow at least a portion of the light to strike the diffusive surface behind the reflective particles. This too can be a source of a bright spot or scintillation.

A number of years ago, this particular issue was presented to the Chemical Engineers at Da-Lite. They conducted a number of tests to replicate the issue of the fabric sparkling. As a result, they made changes in the way the fabric was made to ensure that the particles were placed on the screen in a much more even pattern. In addition, they began using much smaller particles to ensure the fabric would work with future generations of high resolution video projectors. This is what has brought us to where we are today. As you can see the size of the materials used to create a good number of Da-Lite’s fabrics are extremely small and even the highest resolution projectors available today will not cause a resolution or scintillation problem.

But it works right? Yes mica and Pearlescent are used, but in an extremely controlled manner. Everything from their size, orientation, and even concentration are precisely measured. DIY isn’t that precise and is where some problems can start to be encountered. Too little and the screen could look anywhere from ‘glittery’ to no real improvement. Too much and there will most likely be color shifting, especially off axis.

That wasn't a slam at DIY, the commercial companies have resources and money that we just don't have.

One solution would be to use non-interference pigments. Instead of being based on mica, the core of the particle could be aluminum oxide, which is thick enough to not transmit light and therefore opaque. It is then coated with the same thin layer of titanium dioxide that the interference pigment was but the optical results are much different. The reflections are reduced and the degree of color separation is minimized.

Don't take my word, here's some knowledge from a Color 'Doctor'- Richard S. Burrows, Ph.D.

Dr. Burrows wrote:

Let’s play a numbers game for a minute. Suppose you have a 700 lumen projector. Twenty years ago, this was a state of the art projector. Say your projector screen is sized that 10 foot candle hit the screen. The pearlescent screen will reflect 4% of the light that hits the front surface and transmit the rest onward. Four percent is not much, just 0.4 ft-lambert reflected. This is hardly noticeable. But on the same size screen, if a 2000 lumen projector is used, the light on the screen is now almost 3 times, so that 4% reflection is now over a foot-lambert. That reflection is more easily detectable, especially since it is color shifted. So noticeable, that it will become a distraction. That’s for one mica particle. Well, there are a couple hundred thousand of those mica particles laying on that screen.

Remember that screen gain does not have to interfere with or compromise your home theater.

Note: Richard Burrows, Ph.D. has over 12 years of experience dedicated to research and development of projection screen materials. As a Doctor of Chemistry he has immersed himself in the field of optics and how new technologies can improve front projection screen performance. Dr. Burrows currently consults with several companies. That means he's one smart dude!

Even these require some care in the amount, quality of the material being used, and they still need to be oriented properly, but color shifting would at least be minimized. Does Aluminum work? You tell me...

That was not a shameless plug, just a demonstration of the first tests being done (that I know of) with aluminums. Expect much more on this soon.

This is definitely an area of DIY that needs and deserves some more study. We are finally starting to settle on grays and know we have a multitude of readily available neutrals anyone can get. We also know that there is a wide margin of acceptable tolerances, but I really think the goal most DIY developers strive for is more than just acceptable. Non-interference pigments are in my opinion the next step in making DIY even better.

DIY is becoming more mainstream every day. Before opening this up for discussion I'd like to just say that people put a lot of thought and spend a lot of money on their projectors, and an equal amount of thought should be put into the screen. Many have said this is and it couldn't be truer- It's the total package, the projector, screen, screen size, room layout, and light control... all of these elements contribute to a great Home Theater.

Okay, discussion time if any... and I have a feeling there is going to be plenty. Please though, if a debate is desired, please back up statements so it doesn't turn wild. I tried very hard to research this and back up things. This isn't 'attacking' anyone or anything thing, rather it was an explanation why iridescents can only take things so far and we are on the cusp of some really big things... but it may require some adjustment in the way we think and do things...

[mechman]

wbassett wrote:

But it works right? Yes mica and Pearlescent are used, but in an extremely controlled manner. Everything from their size, orientation, and even concentration are precisely measured.

I don't think we can stress that enough. Some folks think that because they have the same ingredients it must be the same. Not so. I recall reading or hearing somewhere, may have been from the 'Angles of View' or it may have been from one our contacts at Da-Lite, that they coat one side of their mica - as well as their glass beads - and use a magnetic field to align them properly. This would be way beyond the realm of DIY! At least until we get some physicists involved...

mech

[mechman]

wbassett wrote:

Does Aluminum work? You tell me...

Still gotta test it with bright whites too! Amongst other things...

mech

[wbassett]

Yes, but got myself in a bit of trouble with the wife, anniversary time and other things that she felt shouldn't have been spent doing testing. I will test with bright scenes as well as ambient scenes and I have a poly coated panel all ready as well as a Gray Screen and will spray a Winter Mountain panel soon to compare this to. (That's going to be specifically for the spraying thread and an evaluation/review of the Wagner, but I will use it for comparisons)

I do think though that those shots show that non-interference methods are something that really needs to be investigated. Even so, it's not going to be a dump and go recommendation. I sincerely think it's a balancing act, but again, I do see some dramatic improvements and that should be investigated.

[cynical2]

wbassett wrote:

Does Aluminum work? You tell me...

That was not a shamless plug, just a demonstration of the first tests being done (that I know of) with aluminums. Expect much more on this soon.

Freudian slip?