Film-Tech Forum ARCHIVE: How Many Lumens?

This is topic How Many Lumens? in forum Digital Cinema Forum at Film-Tech Forum ARCHIVE.

To visit this topic, use this URL:
https://ft-forum.com/ft/cgi-bin/ubb/ultimatebb.cgi?ubb=get_topic;f=16;t=000134

Posted by Robert E. Allen (Member # 1351) on 08-01-2003, 10:21 PM:

I am considering using a video projector in my upcoming project. I'm looking at a model that puts out 3500 lumens. It will be a 35 foot throw to a 9' x 22' screen. Will that be enough light. Also, since I've never used a video projector I would like to know how it adjusts for the different formats.

Thanks.

Posted by John Pytlak (Member # 331) on 08-01-2003, 10:28 PM:

Likely you will be projecting a 1.25:1 aspect ratio graphics image, so your actual image size (assuming the lens you have lets you) will be only about 11 x 9 feet, or about 100 square feet. So a 3500 lumen projector should theoretically give you about 3500/100 = 35 footlamberts on a matte white screen --- more than bright enough. It probably will be less, as the rated output is, shall we say, usually "optimistic".

Posted by Steve Guttag (Member # 268) on 08-01-2003, 11:13 PM:

Just curious...John...why do you say it will be a 1.25:1 image? Did I miss something in the original description.

In the video world...there are only really so many "native" formats.

For "data"...that is stuff that is computer in origin (like Power Point) then you have the resolution of the computer which can be all over the place. Most video projectors today fall in to the XGA 1024 x 768 category (4:3 aka 1.33) or the SXGA 1280 x 1024 (5:4 aka 1.25).

Thus if you are projecting data, then John is quite correct that you will more than likely be working with a 5:4 image..if your projector is a 4:3 native XGA machine than you will not be using all of your horizontal pixels.

For what I call motion video (DVD, DigiBeta, HDCAM, D-5...etc even VHS) you really only have two "native" ratios to worry about that have any real standard. 4:3 (essentially everything prior to HD) and 16:9 (essentially anything HD in origin, including some of the 16:9 enhanced DVDs).

Most large bright video projectors do not have 16:9 displays though they can be made to project that image. Even the current crop of DCinema projectors from Christie and Barco are really 5:4 SXGA projectors that have to be told what the native image is of the source selected and then an appropriate anamorphic lens is used to expand the image, just like scope for film.

Lets say you get one of the current crop of XGA native projectors out there. You have a 4:3 native image. If you are going to run either data (5:4) or 4:3 material, you are fine. If you plan to run 16:9 (or any ratio wider than 4:3) you have choices to make. Many people merely zoom out the picture so the letterboxed image fills the height. This brings about several negative aspects. First, you just enlarged the pixels so the picture will look worse (just like FLAT for film); second, you picture brightness just went down. Using similar numbers to John (remember my example was with a 4:3 machine) your 4:3 picture was at a theoretical 32.4fL (nice, by the way, for high-ambient light power point shows) down to something like 18fL...still not bad but nearly half of what you had.

Lets presume you want to run some DVDs that will have 1.85 or 2.39 scope titles...now you are going to really zoom out (most likely you wont be able to find a zoom lens that will allow this much range...normally there is one narrow sweet spot in a zoom lens' range where you can get 4:3 and zoomed 16:9 with one lens and one projector location. Your theoretical scope image is down to around 10fL. Most of your "light" is now lighting up black pixels. And to add insult to injury, if you need to read any of the menus to set up your projector prior to show...they will be on the ceiling and floor. You will also see just how unblack those pixels are.

An alternative is to use an anamorphic lens for your video projector. ISCO makes lenses for a large quantity of video formats including 4:3 and 5:4 native format projectors. Using them, you can keep your resolution and brightness. I'd put the money into the anamorphic lens and get a less bright projector that will be cheaper to maintain or run a 3500 lumen projector at half brightness (most projectors have an energy save mode). In this manner, you can have the REALLY bright picture for Power point in high-ambient light situations but maintain a great picture with reduced lamp cost in movie mode.

Check out ISCO Anamorphic DLP/LCD Lenses

Steve

Posted by John Pytlak (Member # 331) on 08-01-2003, 11:37 PM:

quote:
why do you say it will be a 1.25:1 image

Just had 1280 x 1024 pixels on my mind.

Posted by Steve Guttag (Member # 268) on 08-02-2003, 12:15 AM:

And just what color are those pixels? [evil]

Posted by Michael Schaffer (Member # 1204) on 08-02-2003, 04:30 AM:

quote:
Just had 1280 x 1024 pixels on my mind.

That reveals John as a Macintosh user.

Posted by Daryl C. W. O'Shea (Member # 1303) on 08-02-2003, 05:09 AM:

John uses an IBM Thinkpad, and (based on his login ID) his system administrator at Kodak is twisted, although apparrently good at what they do. Well maybe not twisted, but certainly out of the ordinary.

I run 1280x1024... I never use Mac's anymore.

Posted by Jon Bartow (Member # 241) on 08-02-2003, 07:07 AM:

I also use 1280x1024 and never use a mac (with the rare exception of some occasional video editing for my church at a nearby studio.
Jonathan

Posted by Clayton Shedd (Member # 1668) on 08-02-2003, 10:11 AM:

i'm looking to show movies on the town green. i would start just a tad before dusk and am told i should use a 3000 lumen projector. indoors, with lighting half or more dimmed, 3500 lumen would be major overkill and very pricey.

Posted by Gordon McLeod (Member # 33) on 08-02-2003, 10:16 AM:

Not all manufacturers have always used the same method of determining lumen output

Posted by John Pytlak (Member # 331) on 08-02-2003, 01:07 PM:

Actually, I was thinking of the 1280 x 1024 DLP chip.

Posted by Michael Schaffer (Member # 1204) on 08-02-2003, 05:31 PM:

quote:
Actually, I was thinking of the 1280 x 1024 DLP chip.

That reveals John as a DLP user.

Posted by John Pytlak (Member # 331) on 08-02-2003, 06:11 PM:

quote:
That reveals John as a DLP user.

I didn't say that I was thinking good thoughts aboout the DLP chip. [Wink]

Posted by Frank Angel (Member # 248) on 08-03-2003, 05:13 AM:

Something smells rotten in videoland. First off, how come video hardware manufacturers have decided that the decades old SMPTE standard for evaluating overall brighness -- using foot lamberts -- is not good enough for video projectors? Perhaps it's not quite as easy to fudge foot lambert numbers? Here's my problem: seems to me there is something amiss in their Alice-in-Videoland world when you can say a projector with 3500 lumens will produce an image equal to 35 foot lamberts, more than double what would be considered a very bright image from a 35mm projector reaching the 16ftL benchmark. Then to confound even more, I just saw a demo of an Eiki LCD projector in our theatre which is rated at 11,000 lumens (that's eleven THOUSAND boys and girls) and low and behold, it was slightly less bright than our 35mm screen brightness, which is just under 16ftL. So is there something wrong with the math here or are the manufacturers simply making these lumen numbers up? According to John's calculations, a projector rated at 11,000 lumens shoud be able to put out 110ftLs? How is this possible? Besides, I can't understand how a projector with a 250 watt lamp can produce a brighter light output than a xenon arc lamphouse burning 2000 watts, notwithstanding the difference in aperture sizes.

And while I am at it, what's with this irrational abandoning of the decades old methodology of stating aspect ratios? For eons the simple declaration of width-to-height has been the norm. It is clear; it is directly understandable without any mental reconfiguring. Height is always 1; the width is always the variable. It is immediately accessable to the mind's eye. Did the vidiots think 16:9 is more understandable than 1.78:1? or 4:3 or 5:4, or are they just afraid of fractions? What are they going to do if they alter a format slightly like we had to do with scope -- 2.55 became 2.35 became 2.39, etc. What kind of bizzare non-fraction ratio are they going to have to come up with to express a small change like that? IMHO, it would have made a lot more sense if they had continued our perfectly workable and elegant way of expressing aspect ratios, even if it IS used by film people. We are not the enemy.

Frank

Posted by Bernard Tonks (Member # 809) on 08-03-2003, 06:47 AM:

quote:
Besides, I can't understand how a projector with a 250 watt lamp can produce a brighter light output than a xenon arc lamphouse burning 2000 watts, notwithstanding the difference in aperture sizes.

Exactly Frank, that has also been bugging me for a long time now. Our local Arts Centre with a spacious 130 seat auditorium uses an Eiki LCD projector with a 250 watt lamp and the light is remarkable for the size of the 4x3 ratio screen. Wide & scope pictures are of course letterboxed, but they get away with it! Films are run 1-2 nights a week which used to be on 16mm, but now only available from the distributors in DVD format.

Posted by Steve Guttag (Member # 268) on 08-03-2003, 10:47 AM:

Frank...a couple things here

As to light...lumens is lumens...foot-lamberts is foot-lamberts...there is nothing hokey going on here. foot-lamberts is reflected light so one must know what the screen material and square footage is in order to calculate...how could a projector manufacturer state an fL figure? They only can state what their projector is putting out, not what screen material and what size screen you are using.

If you look at John's (and mine) math you will see that both the screen reflectivity and square footage is given (matte white screen)

As to light output...there is wizardry going on there. How to get more light is always an issue. However one thing is constant...a point source form of light is more efficient than a larger source. So, if I can get the arc (or any source of light smaller) I can have a more efficient system. One thing xenon manufacturers and users have discovered is that more watts (an electrical expression) does not mean more light output. The current crop of 6KW xenons are generally brighter than last decade's 7KW xenons. The big reason? They 6KW has a smaller arc and can be more efficient (the electrodes are also smaller so the shadow they create is diminished).

One trend you will see with video projectors is the imagers are going to get smaller (as technology allows). This will increase efficientcy. As the imager (DLP, LCD or what have you) gets smaller, the light source must also get smaller (smaller arc) and we get closer to the theoretical limits of light efficientcy. Remember, with film we are constrained to established aperture sizes...not so with the vidiots. In fact, those people really don't have to think about it...all lenses are expressed in "throw-ratios" that already take into account the size of imager.

You will also see that future generation of "Dcinema" projectors will go away from film-style lamphouses since they will crave efficientcy. Right now, with the current DLPs you nead nearly 50% more "watts" for the DLP as compared to a film projector. That is, if your film projector has 16fL with a 4KW lamp...plan on 6KW for a DLP. Remember, for DCinema, they have to "deal" with 1.85 and 2.39 and not the easy-pezy sqaure stuff we abandoned in 1953.

Next thing going for video projectors in the light efficientcy wars...their apertures...at the moment...are almost all either 4:3 or 5:4. Compare that to a film aperture. They are not wasting a good portion of their light going through "light shields" and then smacking on an aperture plate. While not square, a 5:4 imager is only 1.25:1. Wanna see the difference in your film projector? Measure your brightness...take off the anamorphic attachment to your scope lens and measure again...see how bright your system is with just a small square picture.

As to ratios...the Video people are correct on that one. 4:3 is an exact expression. 1.33 is an approximation. Likewise for 16:9 versus 1.78. Anyone that knows how to use a slide rule understands carrying exact ratios is the way to go.

Lets face it film's facination with decimal equivalents has taken on the absurd...we use film expressed in millimeters but all other aspects are in inch form...even the expression of "footage" (regardless of country you are in) ensures some conversion. Our aperture sizes are such that .446 x .825 is for flat...what is with a number like .446?. Heck even the "Academy" aperture is .602" tall...couldn't leave well enough alone at .600?

Or lens diameters...in metric they are 70.6mm in diameter...yeah, yeah but they started out in English measure...fine 2-25/32" (2.78125)...where the heck did that come from? 2.5, 2.75, 3" wouldn't one of them suficed? At least when the 50's rolled around they went to a solid 4" diameter barrel! The aperture was a nice rosey .715 x .910 too...yeah for CinemaScope! However that just wouldn't do so we went to a mag/opt and turned a nice easy aperture that was .715 x .838 or some such nonsense that has been played with ever since until it settled in a more civilized .690 x .825.

So don't give me any more of you white zone Bull sh#$%....er I mean decimal equivalent is better than an exact ratio. Lets face it, 1.85 is going to stick...it seems nice and finished...1.78 wouldn't stand a chance...it LOOKS like an equivalent....why do think there is all the fuss with scope? 2.39 doesn't have near the finsihed sound of 2.35.

Respectfully submitted....

Steve

Posted by John T. Hendrickson, Jr (Member # 849) on 08-03-2003, 11:57 AM:

Bravo, Steve! [thumbsup]

Posted by Randy Stankey (Member # 64) on 08-03-2003, 01:13 PM:

According to the story I heard, George Eastman is responsible for a fair bit of it.

The way it was told to me, when he was developing the first film that we call "35mm", he used panes of glass taken straight from the windows of his house. He progressively cut the pieces of glass in half until he came up with a strip that was "about right".

Logically speaking, all of the 35mm film in the world has the dimension(s) it does because of the size of the windows in the Easman house!

And, by extension... Since John P. works for Eastman's company... IT'S ALL HIS FAULT! [Big Grin]

J.K.

Posted by Frank Angel (Member # 248) on 08-03-2003, 03:06 PM:

Leave it to Steve to clear up the sweet mysteries of life. I say, round everything off to the nearest whole number and be done with it. [Big Grin]

So do 35mm projector manufactures ever use lumens in their specifications when they are describing their console/head systems? Seems like film people and video people could get on the same page with this. Plus, I do understand the difference between lumens and foot-lamberts but I still don't see why video manufacturers can't use the standard benchmarks that will tell you with some certainty how their projector will compare to film or even other video projectors when being used in real-life. I have a hard time seeing how useful lumen claims can be if a projector that claims to put out 11,000 lumens won't produce an image as bright as one that claims to put out 3500, all other things being equal. Shouldn't a projector manufacturer be able to say, my projector will be able to produce 16 foot-lamberts on an matte white with an image size of x? Then any idiot, be he film or video, can look at the numbers and say, this one is the one I need in my theatre.

As for that difference in light efficency between passing light through a film aperture and the DLP nano-mirrors struck me when I first heard about the SMPTE demo. They said they were projecting on a 40ft screen and using a 7000w lamp. I wondered how come no one pointed out that you don't need a 7000w lamp to adaquately light a 40ft screen when using a 35mm system. Seemed like just in terms of energy cost for an exhibitor, that was a BIG difference, but I didn't hear anyone making much of a point about it. I also wondered how aiming that much energy at a stationary object will effect it over a long period of time. What happens to those little mirrors after a year or two....but I digress.

As for scope, I hear guys refusing to call it 2.39 and they say the SMPTE can't make them. They are calling it 2.4:1, period.

And as for the curse of the "light shields," as many have said over and over, cropped wide screen was a stupid idea right out of the starting gate -- never shoulda happened. The only thing that ever made any sense is to use the appropriate anamorphic to get whatever size wide screen you settle on. You want 1.85 (yup, it's a nice number), then do it full-frame -- use all the film geography and the appropriate anamorphosis -- you waste no light and you let ISCO and Schnieder start up a new lens line -- time to develope a quality variable anamorphic attachment that will work for old scope and new wide screen. Then after everyone gets used to the new wide screen numbers, you go back to the SMPTE and say you need to reduce the aperture height to cover up negative flashes, and you get even newer numbers. [Big Grin]

Frank

Posted by Steve Guttag (Member # 268) on 08-03-2003, 03:45 PM:

Frank...

I believe you will find that you will be able to predict video projector brightness with the Lumens spec better than film projector brightness with a wattage spec. If one projector states that it puts out 5000 ANSI lumens and another states it puts out 11000 ANSI lumens then I'd be willing to wager the 11,000 one will look about twice as bright on the same screen with the same size image as the 5000 one.

It is the film industry that mucked that up. Why, oh why, do we express our light output in WATTS? Watts is an electrical measure...Volts * Amps = Watts. While there is often a corellation between wattage and light output within the same lamp time (xenon short arc, for instance) you can equate watts between lamp types nor draw any real conclusions on actual light output of the projection system.

A lamphouse manufacturer can't really tell you what size you need to get 16fL give a particular size screen and material...they would would have to know the projector AND the lens you are using too (presuming no projection port glass since that will take away some as well). Different lamphouse manufacturer's use different reflector designs with understandably different efficientcies. You have projectors with conical shutters, flat shutters, barrel shutters, single blade, two blade, three blade...large, small ...etc...the efficientcy of the systems are all over the place.

The lamphouse manufacturer also has to think about you changing the lamp...some designes need contortionists to change those lamps there is so little room infront of the reflector, even with a pivoting lamp table!

So which projector does a lamphouse designer design to? Certainly the "house brand" of projector seems like a good idea, if one exists but that might limit your sales to those that buy the package.

Then there is the lenses...who makes the best lens? Define best? Can everyone afford that? Does the brightest lens offer the sharpest picture? I'm sure if we went to f1.0 lenses the picture will get brighter but the focus will zip away too. Remember too, film has pulsates when it is heated up so the film projection lens MUST have an adequate depth of focus that the video lens does not....

The video lens and hence the projection system, is designed with a specific projector in mind (or a new projector is designed for an existing lens) so once again a video system can more easily state their lumen output since everything is known.

When lamphouse manufacturers were stating how big a screen their products would light...they were using non realistic f/1.7 lenses on pearescent screens with 12fL...kinda like boasing a 36" TV only to find it is 36" on the diagonal.

The closest anyone could get to the Lumen output are those manufacturers that offer complete systems (Christie, Strong, Kinoton, Cinemeccanica....etc) but then they would have to qualify that with a particular lens (manufacturer, size and model) and that could be construed as an endorsement...and I'm sure the most efficient size/model lens for their particular projection system would be used that may not tell you what you want to know if they use a 90mm ISCO and you have a 50mm Schneider (again, those are THE only two companies to consider in film projection lenses).

Steve

Posted by Robert Throop (Member # 61) on 08-05-2003, 10:06 AM:

Years ago, National Carbon and some lamp manufacturers published lumen ratings for carbon arc lamps. They would specify the lamphouse characteristics (carbon combination, mirror size and current), the lens speed and focal length, and the aperture size. They would also state that no port glass, heat filter, lens attachment or shutter was used. Why can't the same thing be done with xenon?

Posted by John Pytlak (Member # 331) on 08-05-2003, 10:30 AM:

The lumen output for carbon arc lamps was published in the National Projection Carbon Bulletins, Issue #3, December 1957, pages 4-7. It's in the Film-Tech "Manuals" section:

http://www.film-tech.com/manuals/CARBONBULLETINS.pdf

For a 35mm scope image, the larger rotating carbon lamps could produce between 30,000 and 40,000 lumens depending on the trim and current, with the "record" being 42,000 lumens for an experimental 11mm "Ultrex" carbon at 195 amperes and 80 volts. That trim was consumed at 45 inches per hour!

The lumens were measured with a lens matching the f/number of the optical system, and no shutter.

Posted by Tim Reed (Member # 137) on 08-05-2003, 10:38 AM:

I remember some manufacturers also used to give the intensity in candellas.