Heya Guy's, this thread is mainly aimed at those who know abit about the ansi term, im curious to see if we can measure the ansi of our projectors with our eyes, now say take a 1000lm light bulb, place that 3 meters from the wall with a reflector behind it and reflect that on the wall, if we use that say as a base line, ( i know inacurate but its a base line) or if somone could measure that over a 100 inch image space into 4 parts within a meter square with a light meter we could use that as a baseline to determine how bright our projectors are, to me realy ansi lumens is all crap, u can have a 1000lm ansi pro projector and depending what source it is u wont always get 1000lm ansi, also with colours that will differ the ansi, source strenth ect, this is a tough subject that id like myself and the more exsperienced guys in here to learn, i can make a simple light meter up for us or u can make it yourself, it would cost about $5 but we do need a base point to start with and to learn more on this subject, getting back to pro projectors they are all a gimmic to me, there is no way u can get 1000lm ansi out of a 1000lm ansi projector all of the time, do they measure this with a dvd source? on what player with what signal strength? also is it measured on vga? or is it measured on the red channel of colouring? this is somthing id like to get clear, and weather u have a 300lm ansi projector or a 4000lm ansi projector all it is going to mean is that it is more watchable in a lightend room the higher u go with the ansi number, if your picture looks fine at 300lm ansi then all that means is u have to watch it in the dark. Any thoughts on this matter or ideas are much apreciated.
Trev
Trev
Ace3000,
If you look back in one of the two original long threads I believe there is someone that posted a method for calculating the screen lumens. I saw a more advanced method for calculating the LCD losses somewhere but that was a technical paper from IBM I think.
As a rule of thumb you can assume that each lens and mirror will reduce the light output about 4-5% and that the LCD will pass at most 30% of the light that is transmitted through it.
The only way that yo can assume that the total light output from the light is being used is if you have a near ideal light engine which is pretty hard to get.
I don't know if the lumen measurements for lights are given as a spherical reading or a section of a cone. That would be interesting to know.
Hezz
If you look back in one of the two original long threads I believe there is someone that posted a method for calculating the screen lumens. I saw a more advanced method for calculating the LCD losses somewhere but that was a technical paper from IBM I think.
As a rule of thumb you can assume that each lens and mirror will reduce the light output about 4-5% and that the LCD will pass at most 30% of the light that is transmitted through it.
The only way that yo can assume that the total light output from the light is being used is if you have a near ideal light engine which is pretty hard to get.
I don't know if the lumen measurements for lights are given as a spherical reading or a section of a cone. That would be interesting to know.
Hezz
yeah thats about right, thanks for the tip on the threads too i never did see them, most lcds put out at a maximum of 10% of the original light , some even less, i beleive the new poly si lcds wich now is incorporated in the sharp range of lcds is about 30% i know the one i have has that technology too but working on a10% from light of 14000lm and so thats 1400lm then 10% of that lost in the optics then thats 1260lm then u have the throw distance wich will wash it out god knows how much that will knock it back at various distances, but the reason i ask hezz and bring this matter up is because i want to make a rough system for the guys to go on and to make a couple of light meters, but we need to come to an agreement so i know how im going to calibrate this, i could do it side by side with a light meter but i dont want to go out and pay a heap for a light metter just to calibrate one ive just made on the cheap lol, im planing to make 2 one is a simple 3 led system while the other is a digital that i could make and sell for about $50 for the digital and about $5 for the basic, also if i compare my whites in my projected image to a 4200k 100w light then my projector puts out more then 1000lm projected at a size of 140inch, but thats just in the whites, ofcourse in other colours it differs, and u are right about how they measure normal lights, that is an interesting point to bring up, but lets find out.
Trev
Ace3000,
I guess that I don't have the patience to do any kind of a accurate calculation but my feeling is that the poorest DIY systems are getting about 8% of total light output and the best ones are mayble close to 20%.
This is assuming that there are no gross lens or light path misalignments.
Let me see if I can find that paper on LCD tranmissivity and if I do I will e-mail it to you as an attachment.
Hezz
I guess that I don't have the patience to do any kind of a accurate calculation but my feeling is that the poorest DIY systems are getting about 8% of total light output and the best ones are mayble close to 20%.
This is assuming that there are no gross lens or light path misalignments.
Let me see if I can find that paper on LCD tranmissivity and if I do I will e-mail it to you as an attachment.
Hezz
thanks hezz and also i think the ohp guys loose more light because of the diylabs triplet, and trying to light up a bigger image plane so that all has to be acounted for too, anyway buddy thanks for your help and tips
Trev
Trev
Ace3000,
I tried to e-mail you that PDF file but for some reson it was rejected by hotmail. Maybe it was too big of a file. I was pretty good information on LCD transmissivity even though it was about designing better LCD light engine. Here is the link. It is long so I hope it dosn't truncate it.
http://domino.watson.ibm.com/tchjr/journalindex.nsf/0/956ef858ef1eb21e85256bfa0067fa84?OpenDocument
Hezz
I tried to e-mail you that PDF file but for some reson it was rejected by hotmail. Maybe it was too big of a file. I was pretty good information on LCD transmissivity even though it was about designing better LCD light engine. Here is the link. It is long so I hope it dosn't truncate it.
http://domino.watson.ibm.com/tchjr/journalindex.nsf/0/956ef858ef1eb21e85256bfa0067fa84?OpenDocument
Hezz
DIY Projector Lumens
I don't know the formal method for calculating lumens, but I do know that a DIY projector with a 250w metal halide kit using a reflector without a condenser lens produces an image that is identical in brightness to a 500 lumens projector that I used to have. I think that with the condenser and after using a split fresnel lens that you could squeeze another 250 lumens out of the projector, which makes for a nice setup in medium to low lighting situations.
As near as I can tell, even the best DIY projectors only pass as much as 10% of the original light. The reason is that the LCD panel is of course inefficient, but also our basic light engine is inefficient because so much of the light is wasted into the sides of the enclosure. I've read about people worrying over the exact shape of their reflectors, but the truth is that even with a crappy reflector we are still wasting most of our light into the sides of the enclosure before the first fresnel lens. That's why I had condenser lenses made for us that would hopefully aim more of the light onto our fresnel lenses. So far that seems to work pretty well, although there is still dispute over what reflector to use with it.
In short, the best you could probably get from a projector with no reflector is about 5%, and with a reflector and condenser lens you could probably get 10-15%. That really puts into perspective just how inefficient LCD panels are in general. But notice that a reflector and condenser lens should double or almost triplet your output compared to a system with no reflector or condenser!
Happy holidays everyone!
Alan, DIY Labs
http://www.diylabs.org
I don't know the formal method for calculating lumens, but I do know that a DIY projector with a 250w metal halide kit using a reflector without a condenser lens produces an image that is identical in brightness to a 500 lumens projector that I used to have. I think that with the condenser and after using a split fresnel lens that you could squeeze another 250 lumens out of the projector, which makes for a nice setup in medium to low lighting situations.
As near as I can tell, even the best DIY projectors only pass as much as 10% of the original light. The reason is that the LCD panel is of course inefficient, but also our basic light engine is inefficient because so much of the light is wasted into the sides of the enclosure. I've read about people worrying over the exact shape of their reflectors, but the truth is that even with a crappy reflector we are still wasting most of our light into the sides of the enclosure before the first fresnel lens. That's why I had condenser lenses made for us that would hopefully aim more of the light onto our fresnel lenses. So far that seems to work pretty well, although there is still dispute over what reflector to use with it.
In short, the best you could probably get from a projector with no reflector is about 5%, and with a reflector and condenser lens you could probably get 10-15%. That really puts into perspective just how inefficient LCD panels are in general. But notice that a reflector and condenser lens should double or almost triplet your output compared to a system with no reflector or condenser!
Happy holidays everyone!
Alan, DIY Labs
http://www.diylabs.org
heya allan, i recon im getting about 25% of my light output from my projector, its definatley more then 10%, crusers setup on one of his projectors would have to be over 1000lm ansi, the one im making now would have to be atleast 700 - 1000lm ansi and yet it will soon be brighter as soon as i get some gear i orderd out of the bloody post lol. But yeah alot has got to do with the light engine and the reflector u use, the reflector im using and a few others i sold the precision reflector to is about the best money can buy, jcb is also using one similar to mine and they arent cheap, but i can garentee u that they wouldnt be far off of using the majority of the light from the bulb, in the future ill be fooling with parobolic reflectors that have a enhanced alloy coating, its just a matter of finding a lamp that has one to modd and not getting one like the diypro shop has got that is difused.
Trev
Trev
Alan,
after reading a published article from IBM i'm inclined to think that your estimates for the LCD are somewhat low. It's more likely closer to 20-30 percent that the LCD passes with most LCD's between 20 - 25 %. However these are based on the assumption that you are seding it polarized light. So if there is no light recycling you are probably passing between 12-15 % of the light going into the LCD.
Also there is no reason why the light engine has to be as ineffecient as you claim. The large light bulb setups were very ineffecient but I think we can do a lot better.
Hezz
after reading a published article from IBM i'm inclined to think that your estimates for the LCD are somewhat low. It's more likely closer to 20-30 percent that the LCD passes with most LCD's between 20 - 25 %. However these are based on the assumption that you are seding it polarized light. So if there is no light recycling you are probably passing between 12-15 % of the light going into the LCD.
Also there is no reason why the light engine has to be as ineffecient as you claim. The large light bulb setups were very ineffecient but I think we can do a lot better.
Hezz
heya HEZZ, i wana try and design somthing soon in a couple of months when there is no plans to do to make up a different kind of light engine, have u ever seen how much light reflects off of the frensel? wow that is alot, now with a pyramid we can reuse some of this but not all so its brought my atention to make somthing like this: basically a filter that is a one way filter , this means light can only go in one direction, i want that light that gets reflected from the frensel reflected back to the lcd but at the same angle, im looking into using somthing like a one way mirror or a dichoric mirror in its correct terms, but with a percentage of like 90% transmisive with a 10% reflective surface, this should work, also placing a polariser plane before the first frensel, this first polariser will take the heat and not the lcd's polariser thus making the lcd run much cooler although we can get away with this with just a peice of ir filter glass, anyway thats whats on the cards for the new first part of the year in the light engine department, as for the polarised light geting thrown at the lcd makes a lcd brighter then a non polarised light , well thats all crap, if u have 4 polarisers all lined up exactly before the lcd u wont gain light nor loose light, the lcd has a polariser on it that polarises the light already, the ibm site basically just tells u how they have made the light recyclable with the polarising films, i have some of that stuff here and its how they can get a lcd with a brightness of 400nits compared to a lcd with 280nits, and thats how they do it, no brighter back light but just the light is recycled, now in projection that has nothing to do with what we are doing as the light in a lcd's back light is coming in from the sides and then difused and filterd to come out strait and evenly across the screen, our light comes from directly the back of the lcd with a minimal angle, the new idea above with the reflecting recycling method should be similar the what ibm have done but ofcourse with different angles.
Trev
Trev
Hezz said:
Actually I agree with you. When I posted that, I was recalling the simple calculations I did a long time ago before we figured out fresnel splitting. The whole reason fresnel splitting is so good is because you are able to feed the LCD more or less parallel rays of incident light, thus greatly increasing the efficiency of your system in many cases. I also feel that this increases the effective contrast of LCD panels because each pixel is able to filter more light when it enters at a good angle.
The reason that my own projector is around 500 lumens is probably because I wasn't using a split fresnel or condenser lens the last time I tested it, and also because I had a crappy LCD panel back then (it was a really old proxima panel that could only show about 512 colors).
DIY light engine
Guys,
I believe this or a variation of this will be the most efficient light engine that we can design with a resonable cost. I feel that it will utilize about 80% or more of the total light of the bulb. The important things are that the light arc be at the bottom of the reflector and the bulb be as small as possible.
With this arraingment most of the light from the arc will be directed through the condenser to the fresnel. Any stray light is trapped in the light pipe and recycled for correct polarity. It is important the the main reflector be reflective on the inside and outside. And any brackets inside the pipe should be of reflective material. Also the longer the fresnel focal length the better this will work because it will reflect less off incident light.
Hezz
Guys,
I believe this or a variation of this will be the most efficient light engine that we can design with a resonable cost. I feel that it will utilize about 80% or more of the total light of the bulb. The important things are that the light arc be at the bottom of the reflector and the bulb be as small as possible.
With this arraingment most of the light from the arc will be directed through the condenser to the fresnel. Any stray light is trapped in the light pipe and recycled for correct polarity. It is important the the main reflector be reflective on the inside and outside. And any brackets inside the pipe should be of reflective material. Also the longer the fresnel focal length the better this will work because it will reflect less off incident light.
Hezz
Attachments
Hezz
nice design buddy its similar to what im going to design but my design would be with easier parts to get and a few filters in different places , but overal nice design bud.
Have a merry Christmas!
Trev
nice design buddy its similar to what im going to design but my design would be with easier parts to get and a few filters in different places , but overal nice design bud.
Have a merry Christmas!
Trev
Hezz, I have worked on something like your lightenginee now for two weeks, but I can say that I failed. It seems that it´s not possible to focus the light and then spreed it from a focus point, because you will get a projection of the reflector on the lcd and that projection will get anlarged though the fresnels and projected on the screen mixed with the lcd-picture. Maybe some kind of diffuser can solve the problem.
Attachments
Mathias,
I understand what you are saying. I think that the way to do it is to design it so that it is close but then put the condenser focal point back an inch or two form the fresnel focal point so that the light is at nearly the correct angle but so that the light is out of focus. Some posts by Ace3000 have given me the impression that this is necessary or otherwise the light patterns from the reflector are focused too well throught the LCD.
Also the reflective polarizer should help a lot in this regard because it will bounce back into the light pipe out of polar light and it will take a lot of varied light paths before it comes back out thereby diffusing the light into a more even field.
Hezz
I understand what you are saying. I think that the way to do it is to design it so that it is close but then put the condenser focal point back an inch or two form the fresnel focal point so that the light is at nearly the correct angle but so that the light is out of focus. Some posts by Ace3000 have given me the impression that this is necessary or otherwise the light patterns from the reflector are focused too well throught the LCD.
Also the reflective polarizer should help a lot in this regard because it will bounce back into the light pipe out of polar light and it will take a lot of varied light paths before it comes back out thereby diffusing the light into a more even field.
Hezz
Guys'
Here is a refinement to the above design. It uses a pre retangular trapazoidal collumator to collumate the light from the bulb into retangular shape the same aspect ratio as the LCD before it goes through the condenser lens. Then the condenser lens focal point is set back a small amount from the fresnel focal point to defocus the light somewhat and make a more diffuse light image. Then the stray and non polarized light is recycled in the large trapazoidal collumator.
Hezz
Here is a refinement to the above design. It uses a pre retangular trapazoidal collumator to collumate the light from the bulb into retangular shape the same aspect ratio as the LCD before it goes through the condenser lens. Then the condenser lens focal point is set back a small amount from the fresnel focal point to defocus the light somewhat and make a more diffuse light image. Then the stray and non polarized light is recycled in the large trapazoidal collumator.
Hezz
Attachments
nice design hezz, that is getting a little large though lol well for my small stuff anyway, it would be great in your setup your making with the frame, that polariser stuff u are talking about is like a mirror, u cant shine light into it directly it has to go through on a very wide angle so i think in this case it wouldnt be too good in the above setup but somthing similar with a diferent angle aproch would work, all we need to do is to get the light through it and for the light reflected back from the frensel to get reflected back to the lcd, i think a simple filter could do this but its just a matter of wich one, u can bet your *** u would see a big difference in picture brightness especially on the bigger lcd's and ohp setups as they waste a tonne of light, i think the more we are looking into this, u get too see realy how much light we are actually wasting, (me anyway).
Trev
Trev
@ Hezz..
..could you, or someone else, explain something about that polarizing film? Is it the one fixed on external side of lcd (I'm referring myself to my fp 567s Benq) or is it one of the films between lcd and its enclosure seen when I stripped my monitor?
Thanks a lot!
Regards
Michele
..could you, or someone else, explain something about that polarizing film? Is it the one fixed on external side of lcd (I'm referring myself to my fp 567s Benq) or is it one of the films between lcd and its enclosure seen when I stripped my monitor?
Thanks a lot!
Regards
Michele
Same Idea...
Hey, I was thinking that if people were up to it, I could work on a design for a simple circuit that would let use measure the relative brightness of our projectors in a way that we could all compare. If people tested things that they knew the ansi of, it would be pretty easy to create a scale. I need to do a little more research before I can give a price for one of these things, but I think It would probobly be around $15... Most of the money comming from whatever I use to display the output... the numeric LCD display and driver... but with those little calculators that are often given away for free, I doubt that the displays from one are too expensive. I just think that many people would like having a scale to measure the light output anywhere in the light machine/plus screen with a measurable value. I know I do. THe biggest problem with one of these devises would be the scale that they are on... I would have to come up with some kind of standard to test them on... hmmm...
any ideas let me know
Hey, I was thinking that if people were up to it, I could work on a design for a simple circuit that would let use measure the relative brightness of our projectors in a way that we could all compare. If people tested things that they knew the ansi of, it would be pretty easy to create a scale. I need to do a little more research before I can give a price for one of these things, but I think It would probobly be around $15... Most of the money comming from whatever I use to display the output... the numeric LCD display and driver... but with those little calculators that are often given away for free, I doubt that the displays from one are too expensive. I just think that many people would like having a scale to measure the light output anywhere in the light machine/plus screen with a measurable value. I know I do. THe biggest problem with one of these devises would be the scale that they are on... I would have to come up with some kind of standard to test them on... hmmm...
any ideas let me know
- Status
- Not open for further replies.