Solux's arrived!!!
Did some initial tests with one bulb.
==At 16" a black thermocouple from a digital thermometer registerd and stayed at 105degrees for one hour. With just a very small amount of air -- temps dropped 10 degrees. A single pane of glass droped temps below 90.
My conclusion -- the heat projection is for real!!! This is the coolest halogen I've ever seen as far as forward heat projection.
Also, this is the whitest light I've ever seen!!! I am very very impressed. (And I have to admit -- I bought the 3500K, not the 4700's as they put out 50% more light, and less heat)
At 20', the focused beam is brighter and whiter than any ohp light I've ever seen. Infact, I'm so used to the color of typical halogens, these things look almost blue, and they are only 3500K!!
I focused one on a particularly rich dark picture that I have, and I now know why the smithsonian uses these bulbs.
I'll post more once I get my array built -- still waiting on fresnel.
Did some initial tests with one bulb.
==At 16" a black thermocouple from a digital thermometer registerd and stayed at 105degrees for one hour. With just a very small amount of air -- temps dropped 10 degrees. A single pane of glass droped temps below 90.
My conclusion -- the heat projection is for real!!! This is the coolest halogen I've ever seen as far as forward heat projection.
Also, this is the whitest light I've ever seen!!! I am very very impressed. (And I have to admit -- I bought the 3500K, not the 4700's as they put out 50% more light, and less heat)
At 20', the focused beam is brighter and whiter than any ohp light I've ever seen. Infact, I'm so used to the color of typical halogens, these things look almost blue, and they are only 3500K!!
I focused one on a particularly rich dark picture that I have, and I now know why the smithsonian uses these bulbs.
I'll post more once I get my array built -- still waiting on fresnel.
Sounds excellent. Did i read you correctly, you got a nice bright focus at 20 FEET???
Also remember that the LCD cuts out something like 50%+ of the light...so the real test will be to put the LCD in front of one.
Also remember that the LCD cuts out something like 50%+ of the light...so the real test will be to put the LCD in front of one.
yep, remember I have 10degree spots. My tests were bulb only, no optics. This thing will shoot 50' if I wanted it to.
This is the reason that I thought I could actually build a parallel light source array.
This is the reason that I thought I could actually build a parallel light source array.
WoW! I like it I like it.
So with a 3x4 array of these...we are talking approximately 36,000 lumens of nice mostly parallel rays of consistent intensity washing the entire screen evenly? I like it.
So with a 3x4 array of these...we are talking approximately 36,000 lumens of nice mostly parallel rays of consistent intensity washing the entire screen evenly? I like it.
slow down--- you're not going to get 36,000 lumens..... It is going to be bright though.
I had a post awhile back, where I figured about 350lumens on a 6' screen. The lumen rating obviously decreases with beamspread. At 24" each bulb puts out a 4" diameter circle of light at 3,000 lumens. So, I have a minimum of 3,000 lumens (more since I won't be 24" away) over a 108sq. inches (15" screen) with a 5x4 matrix of bulbs.
Simple math says that if I spread that light over a 10times bigger area, it will be 1/10th as bright. So... a 150" diagonal screen will be 300 lumens. After you figure in your lcd losses, 350lumens on a 72" screen is easily attainable.
Reference this link:
http://www.soluxtli.com/3500k.htm
So.... with these assumptions:
CP = intensity @ 1'
lcd transmissivity 50%
12" from bulbs to lcd lighting design
15"diagonal screen = .75sq.ft.
12,482CP * .50 *.75 = 4680 ANSI lumens
If you do my "guesstimated" math on a 350lumen / 72" screen, you get 5040lumens. Either way, looks like I should be in the 4000lumen range, unless my math is very bad.
PLEASE someone check my math and tell me if I'm wrong.
thanks
I had a post awhile back, where I figured about 350lumens on a 6' screen. The lumen rating obviously decreases with beamspread. At 24" each bulb puts out a 4" diameter circle of light at 3,000 lumens. So, I have a minimum of 3,000 lumens (more since I won't be 24" away) over a 108sq. inches (15" screen) with a 5x4 matrix of bulbs.
Simple math says that if I spread that light over a 10times bigger area, it will be 1/10th as bright. So... a 150" diagonal screen will be 300 lumens. After you figure in your lcd losses, 350lumens on a 72" screen is easily attainable.
Reference this link:
http://www.soluxtli.com/3500k.htm
So.... with these assumptions:
CP = intensity @ 1'
lcd transmissivity 50%
12" from bulbs to lcd lighting design
15"diagonal screen = .75sq.ft.
12,482CP * .50 *.75 = 4680 ANSI lumens
If you do my "guesstimated" math on a 350lumen / 72" screen, you get 5040lumens. Either way, looks like I should be in the 4000lumen range, unless my math is very bad.
PLEASE someone check my math and tell me if I'm wrong.
thanks
Right...i meant the LCD screen not the Projection screen!! I'll get back to you on your math...but looks like a good enough approximation....Hey jbell check out my new thread on combining rear and front projection and tell me what you think please. THanks.
I'll check it out -- finally found a projector design consultant and reprinted his article -- a great read!!! It also confirms some of my math.
"The following is quoted, and not original by jbell"
This is a mini-article I wrote a while back regarding how bright a projector needs to be in a given situation...
Video Projection Formulas – a Primer
I’m frequently asked to explain how much ‘horsepower’ (brightness) a projector needs to have in order to make the image on a screen look good. In a moment, I will give some formulas that are very necessary in determining projector brightness and screen size. However, two 'red flags' popped into my head.
First, it's a bit technical, and some will not understand the concept.
Second, while the formulas are not a big secret, they are a necessary part of how I make a living as a consultant. However, I can honestly say that MOST firms that I've run across do not perform all of the math required to make the estimates for screens and projectors as accurate as possible. How do MOST do it? They guess.
Trust me, I'm about to get to the formulas, but this is important.
Most A/V firms 'guess' in that they have done several (if not dozens) of
installations. Some of these individuals have developed an 'eye' for guessing approximate
values. Some are even very good at this, and, to be honest and try to sound
humble, I usually am very close by guessing myself.
But how much money is being invested here? Is that money worth a good
'guess-timate', or is it worth an accurate, calculated proposal? I prefer the latter.
So, here are the basic math formulas necessary:
Determine screen placement
Sightlines are paramount in determining screen location. Make sure everyone (especially those far away) have a clear, unobstructed view of the screen. In some cases, more than one screen is required.
I won't go into screen viewing angels, as this has a lot to do with screen gain (or lack thereof) and front vs. rear.
Determine screen orientation (front or rear).
Rear is always best, but requires space behind the screen that is
essentially useless for anything EXCEPT the projector (and possibly a
mirror).
Determine screen size.
Measure the distance from the screen to the furthest seat that is in
line-of-sight to that screen. If the majority of the content is video, divide that number by 8. That is the screen height. Remember that this is just one of many possible formats that will determine which number to achieve the screen height.
The screen height will be the same if the aspect ratio is 4:3 or 16:9 as
only the width is different (if done correctly, anyway). Make sure that the closest seat is twice the distance from the screen in comparison to the screen height. Example - 6' tall = 12' away for closest viewer.
Next, measure the distance from the screen to the closest audience seat that is in line-of-sight to that screen. The screen should be no taller than twice that distance. This isn’t always possible, but it is preferable. Remember, the furthest viewers take priority, as it is better for the screen to be “too big” for those close than “too small” for those in the back.
Determine square footage of the screen surface.
Take the screen height (in feet) and multiply it by the screen width (in
feet). Ex. 6' x 8' = 48 sq. ft.
Take the number of lumens that is estimated will be needed (this is the ONLY time we'll estimate lumens).
Example - 1000 (lumens) divided by 48 (sq. ft.) = 20.83
ANSI specifies that 18 (+ or - 2) is the acceptable number. This is assuming NO light is hitting the screen. Pitch black area. Dark. No light. Nada.
So we now know that 20.83 is our number for this example, and our next
measurement is the screen area itself - the amount of foot-candles hitting the screen surface from lights, windows, etc.
Let's assume 8 foot-candles is hitting the screen surface. We now take our number (8) and multiply it by 5 (our next formula). The answer? 40, of course.
THAT (40) is the number we must now reach to have adequate lumens being
projected onto the screen surface. In other words, we must project at least 40 lumens per square foot onto the screen.
So, going back to our first example, we have 48 sq. ft. of screen area. If we project 2000 lumens onto the screen surface, and divide that by 48, we get 41.66, which is enough to accomplish our goal.
Therefore, assuming no screen gain, we will need 2,000 lumens projected onto the screen surface from the projector in order to overcome ambient light.
Here’s another item to think about : contrast ratio. In a pitch-black room, lower contrast ratios are not nearly as important. But in a lit room, the better the contrast of the projector, the better the image will be. Contrast, in layman’s terms, is how black the blacks are in comparison to how white the whites are. We call this contrast ratio.
Contrast ratios are normally listed (if at all) in one of two ways on a projector spec sheet: ANSI contrast and ON/OFF contrast. The number to pay attention to is ANSI contrast.
I’ll give another example to help illustrate contrast ratio.
Let’s say you have two projectors: Projector A is 2000 lumens with a 150:1 ANSI contrast ratio. Projector B is a 1400 lumen projector with a 500:1 ANSI contrast ratio. We’ll also say they are using the exact same lenses projected onto the exact same size screens, side by side in a well lit room.
For an easy simulation, let’s also assume the screens are the 4’ x 6’. Now, which image will appear brighter?
If you said Projector A, you’re wrong.
You see, the human eye perceives brightness in comparison to something dark. Therefore, because Projector B had a much greater contrast ratio, the whites would seem brighter next to those blacks. To our eye, Projector B would seem brighter.
Don’t believe me? Try duplicating this test with similarly equipped projectors and see for yourself that this is in fact true.
Beyond this, it gets quite complicated, and I probably have lost 85% of the readers
four paragraphs ago anyway. In fact, there's much more to it than what I've typed here. Therefore, I'm wrapping this up.
I do ask this one thing: if it's important enough to do, do it right – the first time - hire a consultant or design/build firm.
"The following is quoted, and not original by jbell"
This is a mini-article I wrote a while back regarding how bright a projector needs to be in a given situation...
Video Projection Formulas – a Primer
I’m frequently asked to explain how much ‘horsepower’ (brightness) a projector needs to have in order to make the image on a screen look good. In a moment, I will give some formulas that are very necessary in determining projector brightness and screen size. However, two 'red flags' popped into my head.
First, it's a bit technical, and some will not understand the concept.
Second, while the formulas are not a big secret, they are a necessary part of how I make a living as a consultant. However, I can honestly say that MOST firms that I've run across do not perform all of the math required to make the estimates for screens and projectors as accurate as possible. How do MOST do it? They guess.
Trust me, I'm about to get to the formulas, but this is important.
Most A/V firms 'guess' in that they have done several (if not dozens) of
installations. Some of these individuals have developed an 'eye' for guessing approximate
values. Some are even very good at this, and, to be honest and try to sound
humble, I usually am very close by guessing myself.
But how much money is being invested here? Is that money worth a good
'guess-timate', or is it worth an accurate, calculated proposal? I prefer the latter.
So, here are the basic math formulas necessary:
Determine screen placement
Sightlines are paramount in determining screen location. Make sure everyone (especially those far away) have a clear, unobstructed view of the screen. In some cases, more than one screen is required.
I won't go into screen viewing angels, as this has a lot to do with screen gain (or lack thereof) and front vs. rear.
Determine screen orientation (front or rear).
Rear is always best, but requires space behind the screen that is
essentially useless for anything EXCEPT the projector (and possibly a
mirror).
Determine screen size.
Measure the distance from the screen to the furthest seat that is in
line-of-sight to that screen. If the majority of the content is video, divide that number by 8. That is the screen height. Remember that this is just one of many possible formats that will determine which number to achieve the screen height.
The screen height will be the same if the aspect ratio is 4:3 or 16:9 as
only the width is different (if done correctly, anyway). Make sure that the closest seat is twice the distance from the screen in comparison to the screen height. Example - 6' tall = 12' away for closest viewer.
Next, measure the distance from the screen to the closest audience seat that is in line-of-sight to that screen. The screen should be no taller than twice that distance. This isn’t always possible, but it is preferable. Remember, the furthest viewers take priority, as it is better for the screen to be “too big” for those close than “too small” for those in the back.
Determine square footage of the screen surface.
Take the screen height (in feet) and multiply it by the screen width (in
feet). Ex. 6' x 8' = 48 sq. ft.
Take the number of lumens that is estimated will be needed (this is the ONLY time we'll estimate lumens).
Example - 1000 (lumens) divided by 48 (sq. ft.) = 20.83
ANSI specifies that 18 (+ or - 2) is the acceptable number. This is assuming NO light is hitting the screen. Pitch black area. Dark. No light. Nada.
So we now know that 20.83 is our number for this example, and our next
measurement is the screen area itself - the amount of foot-candles hitting the screen surface from lights, windows, etc.
Let's assume 8 foot-candles is hitting the screen surface. We now take our number (8) and multiply it by 5 (our next formula). The answer? 40, of course.
THAT (40) is the number we must now reach to have adequate lumens being
projected onto the screen surface. In other words, we must project at least 40 lumens per square foot onto the screen.
So, going back to our first example, we have 48 sq. ft. of screen area. If we project 2000 lumens onto the screen surface, and divide that by 48, we get 41.66, which is enough to accomplish our goal.
Therefore, assuming no screen gain, we will need 2,000 lumens projected onto the screen surface from the projector in order to overcome ambient light.
Here’s another item to think about : contrast ratio. In a pitch-black room, lower contrast ratios are not nearly as important. But in a lit room, the better the contrast of the projector, the better the image will be. Contrast, in layman’s terms, is how black the blacks are in comparison to how white the whites are. We call this contrast ratio.
Contrast ratios are normally listed (if at all) in one of two ways on a projector spec sheet: ANSI contrast and ON/OFF contrast. The number to pay attention to is ANSI contrast.
I’ll give another example to help illustrate contrast ratio.
Let’s say you have two projectors: Projector A is 2000 lumens with a 150:1 ANSI contrast ratio. Projector B is a 1400 lumen projector with a 500:1 ANSI contrast ratio. We’ll also say they are using the exact same lenses projected onto the exact same size screens, side by side in a well lit room.
For an easy simulation, let’s also assume the screens are the 4’ x 6’. Now, which image will appear brighter?
If you said Projector A, you’re wrong.
You see, the human eye perceives brightness in comparison to something dark. Therefore, because Projector B had a much greater contrast ratio, the whites would seem brighter next to those blacks. To our eye, Projector B would seem brighter.
Don’t believe me? Try duplicating this test with similarly equipped projectors and see for yourself that this is in fact true.
Beyond this, it gets quite complicated, and I probably have lost 85% of the readers
four paragraphs ago anyway. In fact, there's much more to it than what I've typed here. Therefore, I'm wrapping this up.
I do ask this one thing: if it's important enough to do, do it right – the first time - hire a consultant or design/build firm.
Hmm your math doesn't seem quite right..please explain:
"CP = intensity @ 1'
lcd transmissivity 50%
12" from bulbs to lcd lighting design
15"diagonal screen = .75sq.ft.
12,482CP * .50 *.75 = 4680 ANSI lumens"
but you are calculating only for 1 bulb which is covering not the whole screen, just 1/20th (5X4 array) of the screen..so calculating 3"x3" per bulb gives .06 sqft...so
12,482CP * .50 *.06 = 375 ANSI Lumens...no?
"If you do my "guesstimated" math on a 350lumen / 72" screen, you get 5040lumens"
I'm not sure what calculation you did there...please explain further...thanks.
"CP = intensity @ 1'
lcd transmissivity 50%
12" from bulbs to lcd lighting design
15"diagonal screen = .75sq.ft.
12,482CP * .50 *.75 = 4680 ANSI lumens"
but you are calculating only for 1 bulb which is covering not the whole screen, just 1/20th (5X4 array) of the screen..so calculating 3"x3" per bulb gives .06 sqft...so
12,482CP * .50 *.06 = 375 ANSI Lumens...no?
"If you do my "guesstimated" math on a 350lumen / 72" screen, you get 5040lumens"
I'm not sure what calculation you did there...please explain further...thanks.
Remember I have an array of 20 bulbs, so yes I do have a consistant brightness across the entire screen.
At 12,482CP @ 1 foot, that is per each bulb @ 3" diameter.
I will use a light grid that gives me 2.25 square light "grids"
You did the correct calculation for each bulb -- the key here is that ansi lumens is #of foot candles (lumens per square foot)
Because I have such a large screen (.75sq.ft) If I pass 12,482cp through it, I have the calc that I did.
12,482 (consistant across the entire screen due to the array of bulbs) times .50 (transmissivity of lcd) * .75 sq.ft == 4860ansi lumens.
My guesstimate of 350 (footcandles) lumens * 14.4 sq.ft (72"diagonal screen) = 5040 ansi lumens.
If you want to take this to the extreme, get a 19"lcd and an array of bulbs to light it, If you can find a converging fresnel, then you just increased your total ansi lumens because you have 1.2sq.ft. of lit up lcd, and not .75.
In that case you would have
12,482 *.5 *1.2 = 7489 ansi lumens
At 12,482CP @ 1 foot, that is per each bulb @ 3" diameter.
I will use a light grid that gives me 2.25 square light "grids"
You did the correct calculation for each bulb -- the key here is that ansi lumens is #of foot candles (lumens per square foot)
Because I have such a large screen (.75sq.ft) If I pass 12,482cp through it, I have the calc that I did.
12,482 (consistant across the entire screen due to the array of bulbs) times .50 (transmissivity of lcd) * .75 sq.ft == 4860ansi lumens.
My guesstimate of 350 (footcandles) lumens * 14.4 sq.ft (72"diagonal screen) = 5040 ansi lumens.
If you want to take this to the extreme, get a 19"lcd and an array of bulbs to light it, If you can find a converging fresnel, then you just increased your total ansi lumens because you have 1.2sq.ft. of lit up lcd, and not .75.
In that case you would have
12,482 *.5 *1.2 = 7489 ansi lumens
Just reading back through the thread and thought i'd post this response to (rapsac, i think)
I think there is a gray area on whether to think of these solux's as point source or nice collimated beam, because of th nice 10degree spread.
Below, pic 1 was posted by (rapsac?) during the preliminary discussions on the effectiveness of these solux's. (sorry i have to find a site to put my pics on, so this will be a multi-post post)
I think there is a gray area on whether to think of these solux's as point source or nice collimated beam, because of th nice 10degree spread.
Below, pic 1 was posted by (rapsac?) during the preliminary discussions on the effectiveness of these solux's. (sorry i have to find a site to put my pics on, so this will be a multi-post post)
Attachments
correct -- that's the big unknown, and I will let you all know as soon as I try it.
That's the reason that I'm using a 5x4 10degree array and not a 3x4 17degree array. Also the reason that I'm using a 2.25" grid, so that the only thing exiting the grid is parallel light.
This may not work -- but success or failure, I'll post results.
That's the reason that I'm using a 5x4 10degree array and not a 3x4 17degree array. Also the reason that I'm using a 2.25" grid, so that the only thing exiting the grid is parallel light.
This may not work -- but success or failure, I'll post results.
But if you think more in terms of the way these solux's seem to be designed...i think this image (Pic 1) is more appropriate than a point light source ( espescially since it seems these lights have an even beam spread with no hotspotting...as per the specs).
The next pic shows the light array with a "credit-card" sized fresnel (sold for $.50 at dollar stores) per fresnel, in case (rapsac?) is correct in his analysis, i.e. they act more like point source.
And the last pic, JBELL, is meant to show that maybe placement of the larger fresnel (before/after the lcd screen) really has no bearing on screen door elimination.
The next pic shows the light array with a "credit-card" sized fresnel (sold for $.50 at dollar stores) per fresnel, in case (rapsac?) is correct in his analysis, i.e. they act more like point source.
And the last pic, JBELL, is meant to show that maybe placement of the larger fresnel (before/after the lcd screen) really has no bearing on screen door elimination.
Attachments
I've put some more thoughts into this multi-point lighting. I'm now 100% sure that it wont work this way 🙁
Anyone with a working PJ can verify this easily by using the monitor test program ctscreen (ftp://ftp.heise.de/pub/ct/ctsi/ctscjava.zip) and run the geometry test with the white lines on a black background.
Even with a single 'point' lightsource like a MH bulb you can clearly see a gray area around the white lines. This gray area is caused by the non-pointiness 🙂 of the bulbs.
And there is no way you can get better par. light than with a single bulb.
My advice: use one bulb and keep the rest as spares ;-)
Oh, and your assumption of tranmissivness of LCD panels is also a bit off, 5-10% is the accepted value.
Anyone with a working PJ can verify this easily by using the monitor test program ctscreen (ftp://ftp.heise.de/pub/ct/ctsi/ctscjava.zip) and run the geometry test with the white lines on a black background.
Even with a single 'point' lightsource like a MH bulb you can clearly see a gray area around the white lines. This gray area is caused by the non-pointiness 🙂 of the bulbs.
And there is no way you can get better par. light than with a single bulb.
My advice: use one bulb and keep the rest as spares ;-)
Oh, and your assumption of tranmissivness of LCD panels is also a bit off, 5-10% is the accepted value.
Rapsac
But you are still working on the assumption of a single big Fresnel between the lightsource and the LCD aren't you?
What about my idea of a small fresnel per solux bulb to collimate all the beams?
And 10% transmissivity still gives 1000 ANSI lumens.
But you are still working on the assumption of a single big Fresnel between the lightsource and the LCD aren't you?
What about my idea of a small fresnel per solux bulb to collimate all the beams?
And 10% transmissivity still gives 1000 ANSI lumens.
Using a single fresnel wont work 4 sure, unless you find a way to bundle the light from the multiple bulbs back to a single ligthsource.
Multiple fresnels would theoretically work when you would be able to bend the lcd + fresnels to point them towards your (also segmented) PJ lens.
But that would be pretty hard I guess 🙂
I've tried to take a pic from the test screen, but my camera doesnt catch the gray area, its just too dark for it.
Multiple fresnels would theoretically work when you would be able to bend the lcd + fresnels to point them towards your (also segmented) PJ lens.
But that would be pretty hard I guess 🙂
I've tried to take a pic from the test screen, but my camera doesnt catch the gray area, its just too dark for it.
- Status
- Not open for further replies.