Originally Posted by noob00224
From what I've been reading a negative gain grey screen will darken the image. It is possible to brighten up the image to what an 1.0 gain screen would be. This is done by calibration.
I would dearly love to see the source you've been reading that leads to this last conclusion.
The basic analogy I think of is like sculpting a specific shape out of a shapeless chunk of marble: an uncalibrated projector can make a lot of light but its not very accurate. (Consider how pleasant it is to watch the 2050's "Bright" mode.) The process of calibration involves "chipping away" the machine's inaccurate behaviors. Inevitably this involves trading off raw light output for accuracy. Always. Green light seems really bright but if it makes everything green, you do something to make it less green. Now everything looks better. And darker. That's calibration in a nutshell.
(True, in the middle of the applicable adjustment ranges, you can "reduce" the effect of an overly green condition by simply increasing red and blue to restore balance, but at the high end of the range you can't do this, it just drives the machine beyond its capabilities and causes inaccuracies that are even worse.)
There is no part of calibration, that I know of, that involves making a projector brighter. The exceptions to this rule are really rare and I can't think of any off the top of my head.
To keep the analogy going, you try to chisel away only what you need to reveal the perfect sculpture you had in mind *in its biggest possible size*. You don't continue chiseling and make a dwarf version of the same shape. If you do that, you should have just saved your money and started with a smaller block of marble.
People generally watch projectors as bright as they will reasonably go (or at least as bright as their tolerance for inaccuracy will allow). Never dimmer. Not on purpose.
So I've never heard of the ability to take a projector that has been optimized to put out a good picture and then just suddenly increase its brightness after it's been calibrated. At least not without undoing some aspect of the calibration. You can always make a picture brighter and less acurate, never birgter while maintaining accuracy.
People instinctively love brightness, make no mistake, and many times you make a change and people's taste for brightness will lead them to pronounce the result "better", but there was probably a tradeoff somewhere that makes things less faithful to the original image that was encoded in the signal.
When people calibrate their projectors, the results will generally be the same no matter what surface they are shining the projector on, unless there's a pronounced color shift or something. The surface is a "given" - your only task is to optimize the things you *can* change.
Shining a previously calibrated projector on a darker screen only has one effect: the image gets darker. There's isn't some button you can press that allows the projector to get you back up to your old experience. If it was calibrated right, there's nowhere left to go. It's already at the optimal balance of brightness and accuracy. You can't add rock to the original chunk of marble to allow a bigger result of the same shape.
Anyway, back to the specific topic at hand, ie how bright is a reasonably good looking picture on a HT2050 when it is new, your calculations seem to satisfy you that my real world measurements are in the same ballpark as one of the reviewers. I agree.
I'm not sure it's easy to directly measure lumens. The required device looks pretty formidable (see image here https://i.stack.imgur.com/MJqAP.jpg
) The best you can do is measure some other related property and go from there.
In this case, if the thing being measured is reflected light off of the screen surface, then the screen gain seems to have to enter the calcluation.
We can avoid that extra wrinkle by taking a measurement with the meter *facing* the lens. This results in a quantity with different units, lux, not cd/m2, but can still be used to get at the original thing your after, lumens. I think I have a few readings taken this way somewhere, or I can make some new ones if you think that would be helpful?