What's new

How Bright Should White Be? (aka Torch Mode Technique) (1 Viewer)

Guy Kuo

Supporting Actor
Joined
Mar 6, 1999
Messages
581
How many times have I read people complain that they simply don't know how high to set their contrast? AVIA shows you how to look for danger signs like blooming but the final instruction is often forgotten or ignored. Turn down contrast (white level) to the LOWEST point that still lets white appear white. Still, people think that their not recognizing blooming means they can keep their displays cranked up. Even worse some mistakenly think this means you should keep the display just under the point of blooming. NO. NO. NO! The goal is to set the display as DIM as possible but still bright enough to be enjoyable in controlled light conditions. That is a white level (contrast) such that white on screen has an apparent luminance of 13 to 20 foot lamberts.
I didn't state the target in AVIA in terms of Foot Lamberts because nobody other than calibrators has a video appropriate light meter in their kitchen drawer. Lacking such a device, the best and safest is to admonish people to turn contrast down too low and then back contrast up to the point that white just begins to look white again. That's usually just above 12 FL.
We get asked why the pattern doesn't have something that appears or disappears like we can for setting black level. You can't build in an end point like that for white level because there isn't a physical limit like the absence of light which defines black. White can be a huge range of luminance up to infinity. Black, well is black so you can have a hard end point built into the pattern. Once you hit black it's black. Lacking a light meter, the other way to solve this is to have people compare their white against a known brightness lamp. Your flashlight and my flashlight aren't going to be the same. Even if we get the same light bulb, they are going to vary with their age. So we've been stuck - until now.
A Way Out -- And into the Fire
A butane lighter which has the same flame height should produce roughly the same amount of light and be usable as a very rough measure of light level. Granted it is not the same color as the white on screen but you can visually compare relative brightness of lights despite color differences. Just don't set anything on fire while trying this at home. This is just for adults who know how to use a lighter and realize that paper burns. Keep a fire extinguisher handy and proceed at your own risk. You have been warned.
Turn off all lights in the room and display the Needle Pulses & Log Scale patttern in AVIA. Take a piece of white printer paper, in my case a piece of 20 lb Xerox "Primary Image" white paper and hold it up in front of you so you can compare the brightness of the paper versus the brightness of white in the pattern. Light your butane lighter (mine was a BIC). Move the lighter back and forth in front of the paper. I mean in front of the paper as in closer to you than the paper. At a distance of 2.5 inches from the paper, a one inch flame illuminates the paper about as brightly as 13 Foot Lamberts on the screen. A 3/4 inch flame does the same at about 2 inches. Adjust your white level (contrast) to make white that amount of light.
This should work as a rough estimate of how luminous white actually should be. It's certainly not a laboratory grade measurement but at least you get some idea of the light level we're aiming at for white.
Most people will be shocked at how low that actually is. This is a level for fully controlled (dim) lighting conditions. You can go up to a higher level for more ambient light, but be aware that you are wearing your CRT's more rapidly and making them prone to burn-in if you use white levels higher than 13 to 15 FL.
Once more, be careful with the lighter or you will experience an unwanted, new definition for what "torch mode" means in home theater.
------------------
Guy Kuo
Link Removed
Ovation Software, the Home of AVIA DVD
 

Michael TLV

THX Video Instructor/Calibrator
Senior HTF Member
Joined
Mar 16, 2000
Messages
2,909
Location
Calgary, Alberta
Real Name
Michael Chen
Greetings Guy
Do you mean that I should not be driving my car at 5999 RPM either?
laugh.gif

There probably needs to be an extended discussion of this in the next AVIA to drill it into the user's head.
Regards
------------------
fp.jpg

Michael @ The Laser Video Experience
 

Guy Kuo

Supporting Actor
Joined
Mar 6, 1999
Messages
581
I'm afraid we'd still have people red lining their sets if we spent an entire hour admonishing them to turn things down.
 

Jay Mitchosky

Senior HTF Member
Joined
Sep 6, 1998
Messages
3,729
Oops.
blush.gif

...some mistakenly think this means you should keep the display just under the point of blooming. NO. NO. NO! The goal is to set the display as DIM as possible but still bright enough to be enjoyable in controlled light conditions.
I am Jack's guilty self. I always knew that the goal was enough light to make white white, but I also simply attached this to the no-blooming doctrine: Find that point and back off a couple of notches. This is a revelation, but I wonder if there a better solution than simple flame-and-paper (read: safer, more consistent)?
One alternative could be to construct an optical comparitor with a twist - one that provides a reference for white level rather than gray. Use a Kodak white color card, or the Xerox paper you suggest, with a light source of appropriate intensity opposite. The question, of course, is what should that intensity be? Perhaps use a light with a dimmer control outside and adjust using the flame as an initial reference, and than maintain that setting for actual TV adjustment. Simple, consistent (ie. not wavering flame in and out relative to paper as you move) and repeatable. Plus easy to use with one hand while adjusting the TV with the other.
A second apparatus could be akin to antique ViewMasters. But instead of split-image photo cards at one end and a set of binoculars at the other we could use a white photo card/Xerox sheet at one with a light bulb at the other (separated by an appropriate distance using the torch method as a guide). Again the challenge is proper type and illumination of the bulb, but it could be calibrated in the same manner as above and the overall system would offer the same benefits.
As an alternative to the needle/log pattern could we also use the 100IRE block? A little less distracting as you're only looking at the white block. It's easy enough to go to the other pattern to identify blooming and bending thresholds and then use the lowest value identified.
------------------
--Jay
"No one can hear when you're screaming in digital."
My Home Theatre Pictures...
"You're no messiah. You're, you're a movie of the week. You're a ... t-shirt, at best."
 

Steve Schaffer

Senior HTF Member
Joined
Apr 15, 1999
Messages
3,756
Real Name
Steve Schaffer
Guy,
I've always just turned contrast down until "white" looks grayish, then back up until it just turns "white" and no further. Thank you for confirming that I'm doing the Right Thing.
I think one difficulty is gettin rid of our old concept of what Contrast and Brightness really mean.
Since the dawn of tv, we've been told that Brightness is how bright the picture is and Contrast is how much difference there is between dark and light objects on the screen. This does not jibe with the true concepts of white level and black level.
When compensating for varying room lighting, I've found I can leave brightess alone and just crank up Contrast about 8 notches to compensate for daytime viewing of regular tv, going from 21% to about 28-30%. At this higher setting, scanlines in bright areas do not dissappear on my ntsc set, so I'm pretty sure I'm not overdriving anything.
------------------
Steve S.
I prefer not to push the subwoofers until they're properly run in.
 

Guy Kuo

Supporting Actor
Joined
Mar 6, 1999
Messages
581
The new lamp based displays introduce a new variable. Not only do we have black and white level to adjust on the picture elements, but we sometimes get to adjust light output of the bulb. The association of "bright" is fast getting even more muddled as it now can also mean how bright the bulb is putting out light. Oh no! I've used the word "bright" in yet another context.
I've toyed around with trying to use one of those white LED's at Radio Shack to come up with some sort of reproducible light source. You run up against all sorts of problems. The LED's may vary in efficiency from one to another. Their light output will vary with supply voltage. The amount of soldering heat might change the light output. The white surface needs to be a standard distance from the LED. The optical lensing of the LED will change the intensity. We could try a regular bulb, but how standard are light bulbs and what about their dimming over time?
It will be tough controlling all those variables to the point that I could give a formula and everyone comes up with the same intensity light to look at. Ultimately we give up and try measuring the light to standardize our individual lamps. How do we measure them in a reproducible and low cost way? It's one thing if we had all the bulbs in one physical location, but we're talking about people all over the place building their own and trying to get the same light output. I don't have solution that is cheap enough to work yet.
The best I've come up with is if everyone went out and got the same Lumichrome 1XX D65 fluorescent tube and built enclosures with exactly identical dimensions and materials. The inside of the enclosure would need to be black felt lined to remove any effects of the enclosure color. The comparator screen could be a piece of Kodak gray card since that is a standard materail. We'd also have to vary the light output intensity by some sort of adjustable vignette (shade) for the tube. The design would have to use materials readily obtained anywhere, cheap, and easy to build. The design would have to be exactly followed by everyone. and a fresh (or standardly aged) tube used as they do dim with age. I'm not sure that's much more accurate, but at least it would the right color of white.
How many people would be willing to spend $200 making their own optical standard if it still wasn't quite a calibrated standard? The better you make the project the most expensive it gets. I've expended a few attempts towards this cause including an electronically dimmable version. Unfortunately, fluorescent tubes vary their color as one electronically dims them so its back to a shading system and enclosure.
This leads us to the "torch standard" - not very elegant, not very accurate, and certainly not terribly safe.
------------------
Guy Kuo
Link Removed
Ovation Software, the Home of AVIA DVD
 

Jay Mitchosky

Senior HTF Member
Joined
Sep 6, 1998
Messages
3,729
At a distance of 2.5 inches from the paper, a one inch flame illuminates the paper about as brightly as 13 Foot Lamberts on the screen. A 3/4 inch flame does the same at about 2 inches.
While I understand that the above is just an estimate, what would you suggest to achieve roughly 20fL. Armed with that you could identify (roughly) a range of contrast settings between the 13fL and 20fL you specify earlier.
------------------
--Jay
"No one can hear when you're screaming in digital."
My Home Theatre Pictures...
"You're no messiah. You're, you're a movie of the week. You're a ... t-shirt, at best."
 

Guy Kuo

Supporting Actor
Joined
Mar 6, 1999
Messages
581
Jay, it looks like a 3/4 inch butane flame from my BIC held 1.5 inches from the paper matches 20 FL. This is of course a pretty crude measure. YMV
 

Users who are viewing this thread

Sign up for our newsletter

and receive essential news, curated deals, and much more







You will only receive emails from us. We will never sell or distribute your email address to third party companies at any time.

Forum statistics

Threads
357,063
Messages
5,129,880
Members
144,281
Latest member
papill6n
Recent bookmarks
0
Top