Question About Gain's Effect - Home Theater Forum and Systems - HomeTheaterShack.com

Old 02-05-12, 02:06 AM Thread Starter
Shackster

Join Date: Jan 2012
Location: NE AR
Posts: 76
My System

I apologize if this is the inappropriate place to ask this since it is mostly off-topic, but I didn't want to start an unnecessary thread if this is easily answered. I looked for this for a bit, but didn't find an answer.

Gain.

I understand that we are limited by the laws of physics/ energy can't be created, destroyed, etc. But I am a little confused about the idea that higher gain (or lower gain) cannot increase or decrease the total contrast of the screen. My question is: Even though the contrast RATIO isn't changed, does the image not appear to have more contrast by virtue of the static values changing and having more total difference between them? Let me try to elaborate:

I'm going to assign some arbitrary values (hopefully not based on incorrect assumption):

Darkest Black = 5
Brightest White = 100
Contrast Ratio = 20:1 (I think)
Total Difference = 95

Let's double the gain on this thing...

Darkest Black = 10
Brightest White = 200
Contrast Ratio = 20:1
Total Difference = 190

Now, even though the ratio is the same, the total difference has doubled. Are we not perceptually affected by the absolute distance between the darkest dark and the whitest white on a number line?

It seems to me that the better the projector is at producing black, the more exaggerated this becomes...

Darkest Black = .1
Brightest White = 100
Contrast Ratio = 1000:1
Total Difference = 99.9

Now double the gain...

Darkest Black = .2 (maybe barely noticed because it was so dark to begin with)
Brightest White = 200
Contrast Ratio = 1000:1
Total Difference = 199.8!

Does this make sense what I'm asking? If I'm totally off in my approach, please don't hesitate to electronically school me. I would appreciate it.

Thanks guys! Did I mention I love this site? lol
Also, if this deserves a thread to itself, please, in the words of Patrick Stewart, make it so.

P.S. Perhaps wrongfully, I am ignoring viewing cone, hot spotting, and the like. I also don't understand how our perception is effected by the absolute intensities of light. I'm approaching this from the angle that maybe certain projectors are benefited by higher gain screens (e.g. A projector that produces near absolute black but also seems to be on the dim side may benefit by having its total output doubled by a high gain screen if the viewer accepts the compromise inherent in those screens).

Last edited by j0nnyfive; 02-05-12 at 04:15 AM.
j0nnyfive is offline

Old 02-05-12, 05:43 PM
Moderator Emeritus
Don

Join Date: Oct 2007
Location: Central PA
Posts: 3,772

Quote:
j0nnyfive wrote: View Post
I apologize if this is the inappropriate place to ask this since it is mostly off-topic, but I didn't want to start an unnecessary thread if this is easily answered. I looked for this for a bit, but didn't find an answer.
No worries! I did move it to it's own thread though.

BTW, did you read the thread Gain and other confusing topics. in the Sticky? This thread is a bit old, but the info is still good.

Quote:
Gain.

I understand that we are limited by the laws of physics/ energy can't be created, destroyed, etc. But I am a little confused about the idea that higher gain (or lower gain) cannot increase or decrease the total contrast of the screen.
Total PERCEIVED contrast CAN be changed, and changed drastically, depending on screen type and gray level. To take this to the ultimate lets say we are trying to use an almost black screen. Such a screen would absorb over 95% of the light hitting it. With a "normal" brightness PJ black levels would be intergalactic, but whites would suffer horribly and your gray levels (where image detail comes from) in the image would be almost non-existent. If the PJ were bright enough though even a black screen could produce a good image.

We need to get one thing clear in any discussion about image contrast or brightness. A screen is a PASSIVE device and cannot affect the lumen output or dynamic contrast OF THE PROJECTOR. Those are a product of the PJ alone. What a screen, and the total HT environment, CAN change is how we perceive the projected image when it is reflected off of the screen. This is the HUGE difference between an emissive visual display (TV set, PC monitor) and a front projector. The light from the emissive display goes straight to our eyes with no side trips. The light from a front projector MUST be reflected off of another object before we can perceive it as more than a concentrated light beam.

What has this to do with gain? By "gain" we almost always mean PEAK GAIN which is the brightest image the screen can produce. On regular specular (sometimes called angular) reflective screens this is always at the equal, but opposite, angle that the PJ is hitting the screen at; and it's always on-axis with the PJ. Think of a billiard ball rebounding off a cushion. For retroreflective screens the light from the PJ is always directed back at the PJ; i.e. the billiard ball would bounce straight back to where it started from.

For a screen to have a positive gain it must, following the rules above, produce an image that is brighter than a Lambertian surface (one that reflects light equally in all directions) the same color as the screen. A really flat finish paint comes close to this. BTW, a gray screen CAN have positive gain and still have a peak gain below 1.0 (take our Black Widow™ screen mix for example), more on this later. Since a screen is a passive device and cannot produce light on it's own, if it's peak gain is higher than it's normal Lambertian reflectance then it MUST be stealing that light from somewhere else on the screen, and that is exactly what is happening. When a screen does this it WILL have a VIEWING CONE, which means the brightness of the image will vary with the angle one views it from. The brightest image will be on-axis with the PJ and the further one moves away (off-axis) for that viewing position the darker the image will get. When the gain of a screen reaches a certain level (which varies with screen shade and how the added gain is achieved) the brightness falloff is so great that we can see it from a single position - this is what we call HOT SPOTTING since the image will appear to be brighter in the center than the edges. The thing is that even if a screen has enough gain to hot spot, that hot spot will still contain BOTH brighter whites AND brighter blacks! No screen can tell if the light it is reflecting is part of a light scene or a dark scene.

Quote:
My question is: Even though the contrast RATIO isn't changed, does the image not appear to have more contrast by virtue of the static values changing and having more total difference between them? Let me try to elaborate:

I'm going to assign some arbitrary values (hopefully not based on incorrect assumption):

Darkest Black = 5
Brightest White = 100
Contrast Ratio = 20:1 (I think)
Total Difference = 95

Let's double the gain on this thing...

Darkest Black = 10
Brightest White = 200
Contrast Ratio = 20:1
Total Difference = 190

Now, even though the ratio is the same, the total difference has doubled. Are we not perceptually affected by the absolute distance between the darkest dark and the whitest white on a number line?

It seems to me that the better the projector is at producing black, the more exaggerated this becomes...

Darkest Black = .1
Brightest White = 100
Contrast Ratio = 1000:1
Total Difference = 99.9

Now double the gain...

Darkest Black = .2 (maybe barely noticed because it was so dark to begin with)
Brightest White = 200
Contrast Ratio = 1000:1
Total Difference = 199.8!

Does this make sense what I'm asking? If I'm totally off in my approach, please don't hesitate to electronically school me. I would appreciate it.

Thanks guys! Did I mention I love this site? lol
All good questions, but I think I'll let Mech answer since he is much more familiar with video calibration than I am and this seems to be more down that ally.

Quote:
Also, if this deserves a thread to itself, please, in the words of Patrick Stewart, make it so.
Done.

Quote:
P.S. Perhaps wrongfully, I am ignoring viewing cone, hot spotting, and the like. I also don't understand how our perception is effected by the absolute intensities of light. I'm approaching this from the angle that maybe certain projectors are benefited by higher gain screens (e.g. A projector that produces near absolute black but also seems to be on the dim side may benefit by having its total output doubled by a high gain screen if the viewer accepts the compromise inherent in those screens).
Unfortunately, to get the complete picture of what is happening one can't dismiss the other parameters that effect our perception of the projected image.

I'll give you a link that will keep you busy for days, and probably give you a pretty good headache! It contains MASSIVE amounts of great information about color science and is a highly recommended read for those that care enough about this stuff to want to learn it for themselves.

http://www.cis.rit.edu/mcsl/outreach/faq.php

Here are a few more about black level and perceived black.

http://www.cs.ubc.ca/~mantiuk/pdfs/mantiuk10lpb.pdf
http://www.poynton.com/notes/brightness_and_contrast/

To finish up about the magical 1.0 gain number. The standard reference target to which all other front projection screens are compared isn't really a screen at all, rather it is a target usually made out of magnesium carbonate or barium sulfate powder. These purified mineral compounds form a truly Lambertian surface when compressed into a disc or block. It is also extremely white and VERY close to a Munsell N10. The reflectance of such a target is given a reflectance value of 1.0 and screens are given the value of the ratio between the screen and the reference target (measured screen brightness / measured target brightness). A white screen that also has a gain of 1.0 would provide a similar image as if the screen was made from the reference material. A gray screen however, that has a gain of 1.0 would have some form of viewing cone since it is absorbing some of the light hitting the screen yet has a peak gain the same as a white screen. The Elite CineGrey screen material would be like this. In the case of the CineGrey the gain is gotten by a method that produces a very wide viewing cone for the added gain and most people in most HT situations wouldn't even notice that the image brightness is less when viewing off-axis. Another way of putting this is that the darker the screen is the narrower the viewing angle will be if a peak gain of 1.0 is maintained. The viewing angle will decrease more (and usually quite rapidly) if gains above 1.0 are to be had.

Another term you might run into is the screens "half angle", this is the viewing angle, right or left, where the image will only be 50% as bright as when viewing at 0° on-axis.
Harpmaker is offline
Old 02-06-12, 11:45 AM Thread Starter
Shackster

Join Date: Jan 2012
Location: NE AR
Posts: 76
My System

Wow!

Thank you so much for that information! I did read the post about Gain and other confusing topics, and I really learned a lot from it. I'm not sure if it answered my question or not, but that may be my being too slow at comprehending it all. This site's kinda like drinking from a fire hydrant. lol Love it!

Thanks for those links! I'm going to be busy for a few days (or more)! I'm really digging this stuff. To me, learning about this is about as much fun as experiencing it! I'm kinda the bookish type.

I think I understand most of what you talked about with the relationship between gray and specular gain and viewing cones. Even thought you may have answered my question in your reply, I may be too slow to get it. Just to make sure we're thinking the same things, I've been thinking about the question I asked, and I think I can ask it more clearly than I did before. I had a hard time communicating it the other night. Here's version 2 (for Mech as well):

Is there a point where a projector's best black level may be so good, that using a higher gain screen would not appear to raise the black level to any significant degree? As an extreme example, since black is the absence of light, lets say my projector can create a perfect black = 0. If I had a high gain screen of 1.5, then whatever percentage this increased my brightness by (from sweet spot of course), it wouldn't matter for my darkest black because 0 x anything is still 0, right? Can't increase the brightness of absent light. So, I'm just wondering if my logic makes sense.

So, to me, it seems that increasing the gain "spreads out" the gradients in brightness.

<---------------darker--------------------------------lighter-------------------------------->

1.0 gain: black..darker..dark..medium..light..lighter..bright..brightest.

1.5 gain: black........darker........dark........medium........light........lighter.. ......bright........brightest.

<---------------darker--------------------------------lighter-------------------------------->

In my little Ghetto Graph(tm) above, even though you increase the gain, black is still black. The distance between "black" and "brightest" increases with the 1.5 gain. Now, that would be the perfect projector if it could produce perfect Intergalactic Black(tm). lol But, in theory, it wouldn't have to produce perfect black. If it produced near enough black, I believe you would see nearly the same effect. I'm just speculating.
Here's a graph to illustrate my point:

1.0 gain: close2black..darker..dark..medium..light..lighter..bright..brightest.

1.5 gain: ~close2black........darker........dark........medium........light........li ghter........bright........brightest.

Even though we brightened the "close2black", it was so close to black, that, to us, it still appears to be very dark, only subtly lighter. At the same time, the brightest seems much brighter than it was before, and it is easy to see this. Another way to put it is that the "distance" between the darkest black and the brightest white is much greater now. In other words, the ratio is the same, but the distance is greater. I was wondering about you guys' thoughts on that.

Anyway, sorry for rambling. I just wanted to present my question in a different way just in case anybody reading this didn't get what I was asking. Again, thank you for the information! Now, to those articles...

P.S. If you answered my question above, I apologize for not catching it! You may have to hit me on the head a few times before I finally get something. lol I have to think a bit. Thanks Harp!

Last edited by j0nnyfive; 02-06-12 at 11:52 AM.
j0nnyfive is offline
Old 02-06-12, 08:07 PM
Moderator Emeritus
Don

Join Date: Oct 2007
Location: Central PA
Posts: 3,772

Peak Gain is just a way of saying how much brighter a projected image will be compared to a white screen with a gain of 1.0 (actually I should say to a Standard Reference Target, but a white 1.0 screen comes very close). Gain affects ALL light hitting the screen. If you had a PJ that literally produced NO light output in black areas of the image then a 1.5 gain screen MAY produce a brighter white on-axis than a 1.0 gain screen, but ONLY up to the point of the TRUE contrast ratio of the projector after which point the black levels would rise as well as the white. A screen CANNOT increase true image contrast beyond what the PJ is shooting at the screen. All this also assumes an HT that has ALL room surfaces except the screen covered in black velvet!

This question, or ones similar to it, come up now and then; and we have been meaning to do actually contrast testing to prove it for ourselves. The problem is that Mech is the only one here that has a 1° spotmeter and that is needed to do the testing; and Mech is one busy fella.

So, at least in the real world, this is a moot question (but still a very interesting one!) since ALL digital PJ's that I know of produce visible light in black areas of the image which is increased in brightness right along with the white areas on a screen with gain.

The way projectors get the very high contrast ratios you read about is to take the white reading at the brightest mode the PJ is capable of with all irises fully open and all image brightening features enabled. They then set up the PJ just the opposite to take a black reading. This value means almost nothing since the PJ can't change settings like that while watching a movie. The best, so far, contrast ratio value is the ANSI Contrast value where the PJ shoots an image at the screen consisting of 16 black and white rectangles. The brightness levels of all 16 rectangles are averaged and a more realistic contrast ratio is produced. Forget about all those 1000000:1 contrast ratios you read about, they don't exist in the real world. I stand to be corrected (please provide documented proof), but I believe the true CR of digital PJ's is below 1000:1; usually WELL below. I read somewhere that the usual CR for a properly exposed and developed B&W photo is only 100:1.

I'll have to tap Mech on the shoulder to get his attention on this.
Harpmaker is offline
Old 02-07-12, 06:49 AM
Elite Shackster

Steve Mechelke -mech

Join Date: Feb 2007
Location: Empire Township, MN
Posts: 14,914
My System

Quote:
Harpmaker wrote: View Post
I'll have to tap Mech on the shoulder to get his attention on this.
I'm checking this out now Harp. No tapping necessary!
mechman is offline
Old 02-07-12, 07:08 AM
Elite Shackster

Steve Mechelke -mech

Join Date: Feb 2007
Location: Empire Township, MN
Posts: 14,914
My System

With regards to me measuring contrast ratios, since Rob brought it up in this thread I did some checking on this. I never did get any measures as my meters do not read low enough to measure contrast ratios properly.

Here are my spotmeter specs:

Quote:
Measuring Range Illuminance: 0.63 to 190000lux - 0.10 to 180000FC
Considering that a lot of the displays in the HDTV shootout were much lower than this - the highest MLL was .056. I'm still researching if it is at all possible with anything other than a high end spectroradiometer (\$15,000 and up). I think my OTC 1000 is capable of doing this but it would be a real pain to measure ANSI contrast using it.

I'll have more on this later.
mechman is offline
Old 02-07-12, 11:28 AM Thread Starter
Shackster

Join Date: Jan 2012
Location: NE AR
Posts: 76
My System

Hey guys!

I just wrote a long-winded message, then deleted it after thinking about it. lol I'm going to try to summarize my thoughts with bullet points. I'm just asking the same question in a different and more refined way. You guys are either way ahead of me and I'm just not connecting what you are saying with my question, or my question is so convoluted and nonsensical that, referenced with what you already know, it makes no sense whatsoever. I apologize if that's the case. lol Thanks for hanging in there with me!

Bullets:

- I think many times in conversations about enhancing picture quality with different screens, the distinction between contrast as a ratio and contrast as a simple mathematical difference gets muddied. I think when many people (newbies) talk about wanting to enhance their contrast ratio, they may be being too literal in their description of contrast, and also taken too literally. As I understand it, changes in gain cannot change the ratio, but it can either magnify or shrink the simple numerical difference between the brightest and darkest. I wonder if this distinction may be a sort of "missing link" in some conversations. Some people seem convinced that their screen changed their "contrast." And, correctly, we say that you cannot change the contrast ratio. But, they may be confusing absolute difference with ratio, and they probably don't care. I know this may seem pedantic, but I don't recall ever seeing this distinction discussed.

-Related to all that, I'm wondering that if a person already has a very good projector that is capable of respectable black levels, and they purchase a high gain screen... I'm wondering if they are under the impression that their ratio changed (even though it didn't) simply by the mere fact that their black levels were below some threshold where they could hardly perceive the black level rise. Because they couldn't perceive this black level rise (at least compared to the huge rise in the brighter colors), they mistakenly say "My ratio is better."

-I'm wondering what that threshold would be. lol Even if we don't have adequate instrumentation to measure in this range, is my logic sound?

Thanks for your patience guys! I will eventually get this! I just know it.

Last edited by j0nnyfive; 02-07-12 at 11:44 AM.
j0nnyfive is offline
Old 02-07-12, 11:02 PM
Moderator Emeritus
Don

Join Date: Oct 2007
Location: Central PA
Posts: 3,772

Quote:
j0nnyfive wrote: View Post
Hey guys!

I just wrote a long-winded message, then deleted it after thinking about it. lol I'm going to try to summarize my thoughts with bullet points. I'm just asking the same question in a different and more refined way. You guys are either way ahead of me and I'm just not connecting what you are saying with my question, or my question is so convoluted and nonsensical that, referenced with what you already know, it makes no sense whatsoever. I apologize if that's the case. lol Thanks for hanging in there with me!
No apologies necessary! When you start digging into the nuts and bolts of PJ and screen performance things can get complicated quickly.

Quote:
Bullets:

- I think many times in conversations about enhancing picture quality with different screens, the distinction between contrast as a ratio and contrast as a simple mathematical difference gets muddied. I think when many people (newbies) talk about wanting to enhance their contrast ratio, they may be being too literal in their description of contrast, and also taken too literally. As I understand it, changes in gain cannot change the ratio, but it can either magnify or shrink the simple numerical difference between the brightest and darkest. I wonder if this distinction may be a sort of "missing link" in some conversations. Some people seem convinced that their screen changed their "contrast." And, correctly, we say that you cannot change the contrast ratio. But, they may be confusing absolute difference with ratio, and they probably don't care. I know this may seem pedantic, but I don't recall ever seeing this distinction discussed.
After thinking about this for awhile, I think what you are terming "contrast mathematical differences" is commonly called "dynamic range", which, while connected, is different from the contrast ratio. I think I've been guilty of talking about dynamic range and contrast ratio in the same breath and not differentiating between the two. A quick&dirty definition of contrast ratio is the brightness difference between 100% black and 100% white projected onto the screen. Neither of these values can carry actual video detail since they are absolutes. Dynamic range on the other hand is where the image details exist because of the difference between them; these are the "gray levels" between black and white. The more levels of gray your image has the more detail the image will have. One of the reasons that photos taken of a screen in use usually look different than the same screen viewed with the human eye is that the eye has more dynamic range that most cameras. We can see detail in darker and/or lighter gray levels than the camera can.

If too dark a screen is used both the contrast ratio and the dynamic range of the projected image can be lessened, condensed or clipped because the discernable gray levels are reduced from what the projector is producing. Many times a PJ as it comes from the factory isn't taking advantage of all the details it is capable of producing, but calibrating it will fix that.

Quote:
-Related to all that, I'm wondering that if a person already has a very good projector that is capable of respectable black levels, and they purchase a high gain screen... I'm wondering if they are under the impression that their ratio changed (even though it didn't) simply by the mere fact that their black levels were below some threshold where they could hardly perceive the black level rise. Because they couldn't perceive this black level rise (at least compared to the huge rise in the brighter colors), they mistakenly say "My ratio is better."
Many things enter into the picture (pun intended ) when it comes to PERCEIVING an image. I have seen people talk themselves into believing their screen was the best screen in the world just because they PAID so much for it! After all, a \$6000 screen just HAS to be better than a \$1200 screen right?

Increasing the gain of a screen will only make the image brighter when viewed on-axis with the PJ. Both blacks, grays and whites will be brighter. If the viewer can't perceive the difference between blacks on a lower gain screen and on the higher gain screen they they would perceive an increase in contrast ratio. That same perception may or may not hold true for the gray levels.

Don't forget that we get our perception of contrast from an images black level and not it's white level. A B&W image (this applies to photographs too) can appear "muddy" and lacking in contrast if the blacks aren't black enough, but get the blacks right and the whites can be quite gray and still appear to be white (it's a type of optical illusion).

Whenever we talk about image perception I like to show the image below.

Take a look and you will swear that blocks A and B are different shades of gray. THEY AREN'T! They are both an RGB 107 neutral gray. So much for trusting out eyes...

Here is another illusion that has little to do with the current contrast discussion, but it's so neat I'm going to show it anyway.

This is NOT a moving image, but a single static image; the rotation we are seeing isn't real!

Quote:
-I'm wondering what that threshold would be. lol Even if we don't have adequate instrumentation to measure in this range, is my logic sound?

Thanks for your patience guys! I will eventually get this! I just know it.
Good questions! Your logic seems sound to me, but just realize that a screen can't expand an images actual CR or DR, it can only contract it. How contrast is perceived is a different story.
Harpmaker is offline
Old 02-08-12, 11:48 AM Thread Starter
Shackster

Join Date: Jan 2012
Location: NE AR
Posts: 76
My System

Those images are trippy... whoa...

Embarrassingly enough, I had to copy that checkerboard image into Microsoft Paint, zoom in, cut and paste from the 2 squares and compare them myself. Whoa. I can't believe it. They look so different in the picture.
That moving picture. That's really moving. I know it is. lol Crazy man.

I think I may be making a fundamental error in my assumption of how gain works. Correct me if I'm wrong here. Assuming you are sitting perfectly on axis (and are ignoring hot spotting), does gain increase the brightness of the image uniformly across all shades of gray (including black and white)? So, does it make sense to say "X% increase in gain will increase the brightness of any projected shade of gray by X%?" For instance, could I increase the gain of the screen so that I would get a predictable 20% brighter image no matter what shade of gray I show on the screen? I could show white, and I would know it would be 20% brighter than before. I could show my darkest black, and I know (assuming it isn't perfectly black) that I would get 20% brighter black than before? Is that how it works? That was an assumption I was making in my question (that the change in gain effects all shades (brightnesses) by the same percentage). If this assumption is wrong, then my question is nonsensical. lol

I'm going to have to study up on dynamic range now! I'm soakin it up! That's interesting about the clipping. As a side question, where would be a good site to go to learn about calibration (for free if possible)? I would love to learn how to calibrate! That would be awesome. Thanks again!
j0nnyfive is offline
Old 02-08-12, 12:22 PM
Elite Shackster

Steve Mechelke -mech

Join Date: Feb 2007
Location: Empire Township, MN
Posts: 14,914
My System

Quote:
j0nnyfive wrote: View Post
Correct me if I'm wrong here. Assuming you are sitting perfectly on axis (and are ignoring hot spotting), does gain increase the brightness of the image uniformly across all shades of gray (including black and white)? So, does it make sense to say "X% increase in gain will increase the brightness of any projected shade of gray by X%?" For instance, could I increase the gain of the screen so that I would get a predictable 20% brighter image no matter what shade of gray I show on the screen? I could show white, and I would know it would be 20% brighter than before. I could show my darkest black, and I know (assuming it isn't perfectly black) that I would get 20% brighter black than before? Is that how it works? That was an assumption I was making in my question (that the change in gain effects all shades (brightnesses) by the same percentage). If this assumption is wrong, then my question is nonsensical. lol
For the most part, and in my experience, yes. Especially in the diy world. The amount of 'gain' you add to whites is added to the blacks. There are very few that can keep the darker blacks while making whites a tad bit whiter. You have to find the right balance of materials in order to do this. This is why mixes such as silver fire fail. They have way too much mica in them -among other things. On the commercial side of things there are materials that can do this as well - the Supernova for instance.

But these are not 'leaps and bounds' improvement, especially in the diy area. They are very slight.

IIRC somewhere buried within this forum are some images of silver fire alongside the same shade of gray sans the mica. Those shots should help you visualize it. They are probably in one of the silver fire threads.
mechman is offline

 Bookmarks

 Tags effect , gain , question

Message:
Options

## Register Now

Random Question
Random Question #2

User Name:
OR

## Log-in

Human Verification

In order to verify that you are a human and not a spam bot, please enter the answer into the following box below based on the instructions contained in the graphic.