Gain vs energy

[mechman]

I'll echo what harp said!

[mechman]

robert_1967 wrote:

LOL .

But really, mech and of course all other out there, what do you think about the HTS suggestion?

GAIN X:Y:Z => X=Peak Gain at 0 degrees, Y=percieved acumulated reflected light from 0-30 degr, Z=percieved acumulated reflected light from 0-45 degr. (all figures related to the magnesium block)

RI => Reflective Index. Range from -1 to +1. -1 equals 100% retroreflective and +1 is a mirror. Zero is a ideal lambertian screen.

I think maybe Z should be 0-60 degrees. While a lot of folks have the long, narrow, straight room, some also have a long wide room. I have a fairly narrow room and I find myself sitting at 60 degrees quite often when guests are over.

Now, tell us non-engineers how we calculate Y and Z... in layman's terms!

[robert_1967]

Happy that you enjoy our artithmetics

Basically, what you could do is 10 deg incremental measurements 0-60 deg if that is enough in your opinion.

The easy way out is to do this for every 10 deg interval:
Gain 0 minus gain 10. This gives us the delta.
Divide delta with 2.
Add the half delta to Gain 10. This gives us the average Gain between 0 and 10.
Multiply this average with 10, as the interval between measurments is 10.

Repeat this between all measurements and incrementally add the results.
Then you get the area under the gain curve which could represent an acumulated persieved light reflection of the screen from 0-60 deg.

Custard, I thought excel had some tools for integrals but I cant find any. Can you suggest better, more accurate ways of estimate the integral of the curve based om measurements from mech?

[robert_1967]

Harpmaker wrote:

As it now stands, a screen's gain figure is almost meaningless since all it tells you is maximum gain and gives no hint of other screen attributes; and even a gain-curve won't tell you if a screen is retro-reflective or specular reflective.

Very true. But we do have presented an idea about the retro/specular stuff above.

Harpmaker wrote:

Perhaps something that should be taken into account in these proceedings is the "half gain angle", the angle at which the maximum gain figure is reduced to half it's value. If I remember correctly, this is the angle that commercial screen manufacturers use to come up with their viewing cone figure. I don't know if that angle should be taken into account in these formulae or simply given as a separate calculation.

I agree, but imo that is a very poor and blunt instrument to caracterize a screen. As you all have seen firsta hand the shape of the curve between zero and the point of 50% can vary dramatically. If you look at the loudspeaker industry the frequency range of a speaker is defined as between the two frequencys were it falls 3dB. But the 3dB rule does not say one word about how "flat" it is between. Lets not fall into the same trap.

Cust, any ideas on this? I do think it would be nice to have the 50% in there somehow...but how...

[mechman]

robert_1967 wrote:

Happy that you enjoy our artithmetics

Basically, what you could do is 10 deg incremental measurements 0-60 deg if that is enough in your opinion.

Can we stick to every 15 degrees? I could do 10 but then I'd have to redo a bunch of stuff. And in all honesty, this stuff is extremely time consuming...

[custard]

robert_1967 wrote:

Happy that you enjoy our artithmetics

Basically, what you could do is 10 deg incremental measurements 0-60 deg if that is enough in your opinion.

The easy way out is to do this for every 10 deg interval:
Gain 0 minus gain 10. This gives us the delta.
Divide delta with 2.
Add the half delta to Gain 10. This gives us the average Gain between 0 and 10.
Multiply this average with 10, as the interval between measurments is 10.

Repeat this between all measurements and incrementally add the results.
Then you get the area under the gain curve which could represent an acumulated persieved light reflection of the screen from 0-60 deg.

Custard, I thought excel had some tools for integrals but I cant find any. Can you suggest better, more accurate ways of estimate the integral of the curve based om measurements from mech?

there are some add ons for excel which caculate the area under the curve but i have not used them.
i have simply created the formula myself for one reading and then applied it to the rest.

with the data mech has provided and because we donot have equal intervals between measurements, the closest approximation that i can think of is that trapezoid rule again.

mech has provided readings at 30 and 60 degrees so calculating the values for total perceived light at these is fairly simple.


the 10 degree intervals are going to need more approximations and more calculations..
mech has provided readings for 15, 20, 30, 45 and 60.
so for the 10 degree intervals we are missing 10, 40 and 50.


if we assume a linear change between each interval:

then for 10 degrees we can approximate total gain as 2/3 of the total gain at 15.
the calculation would be:

[(gain at 0 - gain at 15) x 1/3] + gain value at 15 = gain value at 10 degrees

total gain at 10 = (gain value at 0 + gain value at 10)/2 x 10 (trapezoid rule)

for 40 degrees, total gain can be appproximated as total gain at 30 degrees plus 2/3 of the total gain difference between 30 and 45 degrees.
for 50 degrees, total gain can be approximated as total gain at 45 degrees plus 1/3 of the difference between 45 degrees and 60 degrees.


robert - what set of values are you using to calculate your total gain values for -75 to 75. i think the 250 watt readings are better to use than the 1000 watt readings in this instance.

[custard]

robert_1967 wrote:

Cust, any ideas on this? I do think it would be nice to have the 50% in there somehow...but how...

if i understand correctly you want the angle when total perceived gain is 50%

we need to calculate the total gain from 0 to 90 so an assumption is needed for the 90 degree value (or should i say 89.99999999).

we can either assume it stays the same as 75 degrees or it follows a linear relationship from 60 degrees to 75 degrees to 90 degrees.

next we need to formulate a table which shows the cumulative totals from 0 degrees to 90 degrees. this would be a simple extension of the area under the curve table we have already.

at 90 degrees we will have total cumulative gain 'T'

----------------------

T= total cumulative gain
H= 50% cumulative gain

therefore H=T/2

looking at the table we can also see which two cumulative figures H falls between, and we will also be able to see what angles these two figures are related to.

A= lower cumulative value
B= higher cumulative value
C= angle at lower cumulative value
D= angle at higher cumulative value
E= angle at H, the value we are seeking

then:

E = [(H-A)/(B-A) x (D-C)] + C

[wbassett]

Lots to digest that's for sure!

I can definitely answer your one question, #3
"Is it possible, in theory, to get a screen that reflects MORE energy in total compared to the ref?"

Sadly the answer is no because we cannot create light. We can only have at most exactly what the projector is putting out and even that starts to diminish as soon as it leaves the lens. We can however make the image brighter on axis but at the expense of the off axis brightness. That's gain and as much as some want to claim 1.8 gain screens with 180 degree viewin cones, it is simply impossible with a passive device such as a screen.

Gain has always been a tricky topic. That's why I started a thread about gain and other confusing topics. One that also seems to fit this discussion (from what I have read skimming through so far) is specular gain. Don't forget about that aspect of things either.

A real mind blower of a question though is "can a screen with an overall gain of less than one still be said to have 'gain'?"

The answer is yes, but then what about the laws of light and physics and our ever present viewing cone? Look at Black Widow as an example. Normally a plain flat/matte gray in the N7.5 range would not be expected to have an overall gain any higher than .75 at best. Because of the properties of this particular screen (BW) it has a gain of .9, yet the shade is an N7.5. That would indicate we have gain, and yes we do, but the viewing cone is still essentially a 180 degree area.

Under a gain of 1.0 we still have less than our reference material. The difference is the added reflective materials throughout the screen. Since we're not dealing with a specular type gain achieved primarily through sheen, we maintain our diffusive surface.

Once we go over a gain of 1.0 there is no other way to increase the on axis brightness than to 'take' light energy that is normally reflected off axis and refocus that back to the center viewing area. Since we're taking more energy and refocusing it, and now on axis we have greater than our reference material, the viewing cone is more apparent.

This is a very interesting topic, but don't forget gain is just a tool used to achieve the required brightness for a given setting and screen size. Far too many people talk about gain as if it is a knob to adjust the image and make it 'pop' as some like to say. Gain quickly can get out of hand and then we not only have hot spotting, but after a point we can also experience color shifting off axis.

I don't know if I strayed from the topic, if so my apologies! I'll read each post thoroughly after I finish some evening work I have to do. Good topic though from what I read so far!

[robert_1967]

I have guests over soon so I have not time to absorb all above. Will return.

Still a comment to Bill. I questioned ""Is it possible, in theory, to get a screen that reflects MORE energy in total compared to the ref?"

With the risk of beeing expelled from this forum I claim that the answer can be yes. Just to make my point, consider the idea of having a N6 gray as ref. Now if you measure the GAIN with the N6 as ref and test a white matt paint you will get the result that you indeed get more light back from the white compared to the N6.

I am not disputing the fact that we can only reflect the ammount of light actually reaching the screen and that is allway less than the PJ output. But if you think about it, a perfect mirror should reflect more light than a Magnesium block imo. There is a fraction of light that converts to heat in the magnesium.

But hey, look at me, now who is beeing academic here. I am sorry custy :-P

Lets keep the focus on our discussions on user-value-issues.

[Harpmaker]

robert_1967 wrote:

With the risk of beeing expelled from this forum ...

Not likely!

I am not disputing the fact that we can only reflect the ammount of light actually reaching the screen and that is allway less than the PJ output. But if you think about it, a perfect mirror should reflect more light than a Magnesium block imo. There is a fraction of light that converts to heat in the magnesium.

From a pure physics point of view this is correct, but when we stay within the parameters of real screen values (a mirror makes a horrible screen ) and keep the industry-standard magnesium carbonate reference target (which is a very bright white as well as a perfect diffuser); the answer is no, we can't have a screen show more total reflection than the reference; but we can have a screen that shows more on-axis reflection.

[wbassett]

robert_1967 wrote:

I have guests over soon so I have not time to absorb all above. Will return.

Still a comment to Bill. I questioned ""Is it possible, in theory, to get a screen that reflects MORE energy in total compared to the ref?"

With the risk of beeing expelled from this forum I claim that the answer can be yes. Just to make my point, consider the idea of having a N6 gray as ref. Now if you measure the GAIN with the N6 as ref and test a white matt paint you will get the result that you indeed get more light back from the white compared to the N6.

I am not disputing the fact that we can only reflect the ammount of light actually reaching the screen and that is allway less than the PJ output. But if you think about it, a perfect mirror should reflect more light than a Magnesium block imo. There is a fraction of light that converts to heat in the magnesium.

But hey, look at me, now who is beeing academic here. I am sorry custy :-P

Lets keep the focus on our discussions on user-value-issues.

One thing... never worry about being expelled for participating in a discussion.

I apologize about taking the thread off topic.

[mechman]

Here's a MgCO spectrum chart for reference.

[robert_1967]

Great, so there I have it! One gigantic block of Magnesium please to be mounted on the wall

...just kidding..