Hey, can I ask where you downloaded those drivers from? The ones I got from Leadtek sure don't seem to work very well.
dscaler should look awsome if properly tweaked. Actually i heard on avs that the radeon 7500 is best because there are some driver issues with the 7500.
Dscaler looks better than tv because you actually can get 480p and not the 480i that our tv's have
Dscaler looks better than tv because you actually can get 480p and not the 480i that our tv's have
BARBUCHA,
Well done!!! 😀 😀 😀
Gunawan,
Your lenses would be needed if the reflectors were spherical. If they are parabolic, then the rays should already be parallel enough... (See the light array found previously by Xblocker!)
If your approach is the same as mine - to treat the LCD as a CRT projector - then the two prime reasons to get the rays going parallel are:
1) Maximise throughput through the LCD;
2) Stress the fresnel as little as possible by using it the way it was intended - to converge roughly parallel beams to a focal point.
This approach should maximise brightness and clarity. Your original suggestions for the placement of the fresnel & primary objective should work optimally.
I would use as long a focal length fresnel as possible, placed as close to the LCD as will avoid moire patterns, and then choose the best objective lens you have with a focal length that gives tha appropriate zoom for your room when combined with the fresnel.
The closer you get the fresnel to the panel, the less the fresnel optics will be stressed: the majority of the light from a given pixel will only be deflected by a few of local points on the fresnel, rather than its whole surface.
In support of this, the more parallel you can get the light beams entering the LCD, the better.
I think that with a more diffuse light source, you may be able to put the fresnel closer without moire being a pain, but I don't think this would give optimal throughput for the LCD, and the emerging beams might not be good for the fresnel quality.
The natural slight diffusion through the LCD panels will be enough for a good fresnel, and will reduce moire.
eebasist,
I think that he bulbs, if suitably arranged, will not produce hotspots. If the LCD is being treated as a giant CRT projector, then you could live with the natural slight diverging beam that any non-perfect parabolic reflector would produce, and place the lights such that the beam patterns overlap enough to give uniform illumination.
For multiple small bulbs, the best pattern is hexagonal - it is the closest way to pack circular objects, and would reduce the intensity differences between the center of the beam, and the gaps between them.
Bill.
Well done!!! 😀 😀 😀
Gunawan,
Your lenses would be needed if the reflectors were spherical. If they are parabolic, then the rays should already be parallel enough... (See the light array found previously by Xblocker!)
If your approach is the same as mine - to treat the LCD as a CRT projector - then the two prime reasons to get the rays going parallel are:
1) Maximise throughput through the LCD;
2) Stress the fresnel as little as possible by using it the way it was intended - to converge roughly parallel beams to a focal point.
This approach should maximise brightness and clarity. Your original suggestions for the placement of the fresnel & primary objective should work optimally.
I would use as long a focal length fresnel as possible, placed as close to the LCD as will avoid moire patterns, and then choose the best objective lens you have with a focal length that gives tha appropriate zoom for your room when combined with the fresnel.
The closer you get the fresnel to the panel, the less the fresnel optics will be stressed: the majority of the light from a given pixel will only be deflected by a few of local points on the fresnel, rather than its whole surface.
In support of this, the more parallel you can get the light beams entering the LCD, the better.
I think that with a more diffuse light source, you may be able to put the fresnel closer without moire being a pain, but I don't think this would give optimal throughput for the LCD, and the emerging beams might not be good for the fresnel quality.
The natural slight diffusion through the LCD panels will be enough for a good fresnel, and will reduce moire.
eebasist,
I think that he bulbs, if suitably arranged, will not produce hotspots. If the LCD is being treated as a giant CRT projector, then you could live with the natural slight diverging beam that any non-perfect parabolic reflector would produce, and place the lights such that the beam patterns overlap enough to give uniform illumination.
For multiple small bulbs, the best pattern is hexagonal - it is the closest way to pack circular objects, and would reduce the intensity differences between the center of the beam, and the gaps between them.
Bill.
Invictum,
You can download Conexant WDM drivers from here, but your system should be 98SE, ME W2000, or XP
http://www.iulabs.com/download/848wdm_iu.zip
xblocker
You can download Conexant WDM drivers from here, but your system should be 98SE, ME W2000, or XP
http://www.iulabs.com/download/848wdm_iu.zip
xblocker
Gunawan,
Two extra thoughts - related to protecting your LCD:
1) Make sure any bulbs you use are UV filtered;
2) If you are using halogens, would it be worth trying that IR-blocking film posted earlier?
Bill.
Two extra thoughts - related to protecting your LCD:
1) Make sure any bulbs you use are UV filtered;
2) If you are using halogens, would it be worth trying that IR-blocking film posted earlier?
Bill.
Bill, the problem i see is that the analogy to the multiple bulbs of a crt projector is slightly incorrect. I have a crt projector right now. (well 3 of them). The beam patterns overlap when you manually change where on the crt face the electron gun fires the image. Since he will be placing the bulbs with no real adjustments, i still believe that hotspots will occur.
hi
got my mini website up and running.
if you click on MY projects, you will learn how to make a DIY projection screen for $27.
http://www.geocities.com/predator3do/index.htm
ok, check it out and tell me what you think
aleksey
got my mini website up and running.
if you click on MY projects, you will learn how to make a DIY projection screen for $27.
http://www.geocities.com/predator3do/index.htm
ok, check it out and tell me what you think
aleksey
Hi eebasist,
What I was actually thinking of was that the circles of light produced by the beams from the bulbs hitting the back of the LCD panel would overlap slightly - creating a fairly uniform illumination of the back of the panel.
My reference to a CRT projector is to the concept of treating the illuminated LCD panel as a light source - similar to a single tube from a CRT projector (in this case, a very large RGB tube...)
In the same way that a CRT emits diffuse light which is captured by a large lens, and projected onto the screen, the LCD panel will emit a (very) slightly diffuse light when driven by Gunawan's multi-bulb arrangement, which would be captured by a fresnel and fed into a more normal objective to be projected onto a screen.
This arrangement makes much less strenuous demands on the optical point-source quality of the light source/reflector, but, if correctly set-up, with a decent fresnel, gives a high brightness, high quality image.
I did not mean that the bulbs would each project their own portion of the image. 🙁
Sorry for the confusion - as usual with me: more words, less meaning... 🙄
Bill.
What I was actually thinking of was that the circles of light produced by the beams from the bulbs hitting the back of the LCD panel would overlap slightly - creating a fairly uniform illumination of the back of the panel.
My reference to a CRT projector is to the concept of treating the illuminated LCD panel as a light source - similar to a single tube from a CRT projector (in this case, a very large RGB tube...)
In the same way that a CRT emits diffuse light which is captured by a large lens, and projected onto the screen, the LCD panel will emit a (very) slightly diffuse light when driven by Gunawan's multi-bulb arrangement, which would be captured by a fresnel and fed into a more normal objective to be projected onto a screen.
This arrangement makes much less strenuous demands on the optical point-source quality of the light source/reflector, but, if correctly set-up, with a decent fresnel, gives a high brightness, high quality image.
I did not mean that the bulbs would each project their own portion of the image. 🙁
Sorry for the confusion - as usual with me: more words, less meaning... 🙄
Bill.
Guys if you want to know more of HTPC (home theater personal computer) setup, please have a look this site, you can find and learn from simple HTPC to sophisticated HTPC.
Woneill,
the purpose using convex lens in front of each lamp array is to produce parallel light rays.If you only use parabolic reflector, the parallel rays only come from the reflector, direct rays from lamp will spread wide in front of the reflector and wasted, unless you cover the front half of the lamp to bounch the rays back to the reflector (same as car headlamp).
All parallel rays will pasthru the second fresnel and hit in the same point. (=fl of the fresnel). If we put objective lens at that point, we will produce sharp projected image (with good obj. lens of course)!
Diffuse light will produce blurry projected image.
Simple test can be done with cheap laser pointer to show it.
Please see attached sketch below:
eebasist,
IMO hotspot will generated by light passthru LCD panel not at 90 degrees to the panel, it depend on how strong the light intensity, LCD type and how much non perpendicular light hit the LCD (placed the spotlight near the side of the LCD panel and spot the LCD in 30 degrees or more, you will see very dimmed LCD screen compared with if you spot in 90 degrees angle).
This link about LCD, maybe good as a reference.
seeyou.
Woneill,
the purpose using convex lens in front of each lamp array is to produce parallel light rays.If you only use parabolic reflector, the parallel rays only come from the reflector, direct rays from lamp will spread wide in front of the reflector and wasted, unless you cover the front half of the lamp to bounch the rays back to the reflector (same as car headlamp).
All parallel rays will pasthru the second fresnel and hit in the same point. (=fl of the fresnel). If we put objective lens at that point, we will produce sharp projected image (with good obj. lens of course)!
Diffuse light will produce blurry projected image.
Simple test can be done with cheap laser pointer to show it.
Please see attached sketch below:
eebasist,
IMO hotspot will generated by light passthru LCD panel not at 90 degrees to the panel, it depend on how strong the light intensity, LCD type and how much non perpendicular light hit the LCD (placed the spotlight near the side of the LCD panel and spot the LCD in 30 degrees or more, you will see very dimmed LCD screen compared with if you spot in 90 degrees angle).
This link about LCD, maybe good as a reference.
seeyou.
Attachments
Hi Gunawan,
Your diagram is correct as far as it goes - but you stop at the objective, and didn't follow through to the projected image. (I am being lazy myself because I am terrible at drawing diagrams...)
You are absolutely correct that light coming directly from the bulb will be diverging. If you use the MR11/MR16 bulbs with built-in reflectors, then they have transverse filaments, as do some car headlamp bulbs H3 etc. and there will be much light coming from the front of the bulb.
If you use the bulbs with longitudinal filaments, then this is much less of a problem. The question you must ask yourself is whether the light from the filament is more important than the light from the reflector: which plays the biggest part in producing the image?
You could use a diverging fresnel lens within the reflector itself to correct this and produce a single point source, but it gets very messy!
Within limits, though, the simple arrangement should not be a problem. My reasoning is this: If you treat the LCD as a light source (the CRT projector concept), then with a single objective lens, the second of your lens diagrams on your website applies. (Multiple paths of light from the same point on the light source get focussed by the lens to the same point on the image.)
When you put a fresnel panel between the LCD and the main objective, you are not really changing anything from an optical standpoint - the fresnel and primary objective can be treated mathematically as a single compound lens. So again, the second lens diagram on your website applies. The main reason for using the fresnel here is to ensure that as much of the light from the LCD reaches the main objective as possible.
Because of limitations of the way fresnels operate, and because of the small size of the primary objective when compared to the LCD/fresnel, the effect you illustrate on your diagram here is important - wildly diverging rays coming through the LCD will never reach the primary objective, and their contribution to the final image will be lost. This could be due to scattering by the fresnel, or could be that the rays are just too far off the mark.
Luckily, any significant divergence can be limited by the reflector/focussing arrangement on your light source. The LCD will also introduce a little divergence of its own: subsequent research with a microscope on a couple of my panels has indicated that diffraction effects and surface unevenness within each shutter cell are playing a much bigger role than I ever expected. (Despite my prior claims, MLAs are NOT used on the panels I have investigated so far!!! My bad!!! 🙁 There is still a diffusion effect, though... More on this, later...)
Anyway, ideally, due to the nature of LCDs and the limitations of fresnel lenses, the light coming trouogh the LCD should be as parallel as possible, and perpendicular to both the LCD and fresnel.
This will get the maximum amount of light to the primary objective lens without wasting too much, either lost in the LCD, scattered by the fresnel, or plain missing the target...
Diffuse rays from the LCD that DO hit the objective, will not necessarily cause a blurry image, because those rays that make it through the main objective lens SHOULD be focussed by the fresnel/objective combination to the same point on the final image. (Ok, fresnels are not perfect, but if aligned per your original specs, the result is not bad).
This is why I was refering to the CRT projector concept: the diffuse light output from a CRT tube is focussed through a large lens into a sharp image. This is because while the light emitted from the CRT is very diffuse, the image on the CRT which is emitting the light, is sharp. This diffuse emission is why CRTs have a wide viewing angle, and LCDs dont. It is also the reason why you can actually watch a CRT screen - some of the diffuse light from the screen is projected onto the back of your eye into a sharp image.
The fundamental limit on this concept is the quality of the fresnel. The more diffuse the light coming from the LCD, the more work the fresnel has to do in resolving it into a sharp image, and the more its limitations affect output quality.
As you have seen, though, when treated nicely, the fresnel can cause more benefit than harm.
Bill.
Your diagram is correct as far as it goes - but you stop at the objective, and didn't follow through to the projected image. (I am being lazy myself because I am terrible at drawing diagrams...)
You are absolutely correct that light coming directly from the bulb will be diverging. If you use the MR11/MR16 bulbs with built-in reflectors, then they have transverse filaments, as do some car headlamp bulbs H3 etc. and there will be much light coming from the front of the bulb.
If you use the bulbs with longitudinal filaments, then this is much less of a problem. The question you must ask yourself is whether the light from the filament is more important than the light from the reflector: which plays the biggest part in producing the image?
You could use a diverging fresnel lens within the reflector itself to correct this and produce a single point source, but it gets very messy!
Within limits, though, the simple arrangement should not be a problem. My reasoning is this: If you treat the LCD as a light source (the CRT projector concept), then with a single objective lens, the second of your lens diagrams on your website applies. (Multiple paths of light from the same point on the light source get focussed by the lens to the same point on the image.)
When you put a fresnel panel between the LCD and the main objective, you are not really changing anything from an optical standpoint - the fresnel and primary objective can be treated mathematically as a single compound lens. So again, the second lens diagram on your website applies. The main reason for using the fresnel here is to ensure that as much of the light from the LCD reaches the main objective as possible.
Because of limitations of the way fresnels operate, and because of the small size of the primary objective when compared to the LCD/fresnel, the effect you illustrate on your diagram here is important - wildly diverging rays coming through the LCD will never reach the primary objective, and their contribution to the final image will be lost. This could be due to scattering by the fresnel, or could be that the rays are just too far off the mark.
Luckily, any significant divergence can be limited by the reflector/focussing arrangement on your light source. The LCD will also introduce a little divergence of its own: subsequent research with a microscope on a couple of my panels has indicated that diffraction effects and surface unevenness within each shutter cell are playing a much bigger role than I ever expected. (Despite my prior claims, MLAs are NOT used on the panels I have investigated so far!!! My bad!!! 🙁 There is still a diffusion effect, though... More on this, later...)
Anyway, ideally, due to the nature of LCDs and the limitations of fresnel lenses, the light coming trouogh the LCD should be as parallel as possible, and perpendicular to both the LCD and fresnel.
This will get the maximum amount of light to the primary objective lens without wasting too much, either lost in the LCD, scattered by the fresnel, or plain missing the target...
Diffuse rays from the LCD that DO hit the objective, will not necessarily cause a blurry image, because those rays that make it through the main objective lens SHOULD be focussed by the fresnel/objective combination to the same point on the final image. (Ok, fresnels are not perfect, but if aligned per your original specs, the result is not bad).
This is why I was refering to the CRT projector concept: the diffuse light output from a CRT tube is focussed through a large lens into a sharp image. This is because while the light emitted from the CRT is very diffuse, the image on the CRT which is emitting the light, is sharp. This diffuse emission is why CRTs have a wide viewing angle, and LCDs dont. It is also the reason why you can actually watch a CRT screen - some of the diffuse light from the screen is projected onto the back of your eye into a sharp image.
The fundamental limit on this concept is the quality of the fresnel. The more diffuse the light coming from the LCD, the more work the fresnel has to do in resolving it into a sharp image, and the more its limitations affect output quality.
As you have seen, though, when treated nicely, the fresnel can cause more benefit than harm.
Bill.
please let me know
DOES ANYONE HAVE ANY IDEAS ON THE LCD PANEL THAT WE MIGHT BE ABLE TO USE IN OUT PROJECTORS???
please let me know...
aleksey
DOES ANYONE HAVE ANY IDEAS ON THE LCD PANEL THAT WE MIGHT BE ABLE TO USE IN OUT PROJECTORS???
please let me know...
aleksey
Woneill,
the confusion is still here!
For communication there should be clearity about terms we use.
A lens between an object and projection lens in optics usually is called a 'field lens'.
LCD isn't a diffuse lightsource like a CRT! You can treat it like that, but the results are poor! You said it yourself, viewing angle is not the same as CRT. Even if a diffuse backlight is attached, e.g. laptops. As we know, view is much brighter on the optical axis. I don't say, there is no diffusion of the panel, but we have the fact, that there is a lot more transmission than diffusion. And transmission is the more important part of projection! That's also the reason, why panels work much better with direct light than diffuse light. To much light would be wasted, or you would need an overkill of light.
And, if LCDs were as diffuse as CRTs, why aren't field lenses in CRT projectors? They then should do the same improvement there as in LCD systems. But they don't exist!
Also there is a change of optical system with a field lens. If the FL of fresnel is to short, the objective could run into troubles not to refract properly the 'cone' angles of the lightbeam.
The task for every 'projectorist' remains, to get out most efficient light of these damned panels!
xblocker
the confusion is still here!
For communication there should be clearity about terms we use.
A lens between an object and projection lens in optics usually is called a 'field lens'.
LCD isn't a diffuse lightsource like a CRT! You can treat it like that, but the results are poor! You said it yourself, viewing angle is not the same as CRT. Even if a diffuse backlight is attached, e.g. laptops. As we know, view is much brighter on the optical axis. I don't say, there is no diffusion of the panel, but we have the fact, that there is a lot more transmission than diffusion. And transmission is the more important part of projection! That's also the reason, why panels work much better with direct light than diffuse light. To much light would be wasted, or you would need an overkill of light.
And, if LCDs were as diffuse as CRTs, why aren't field lenses in CRT projectors? They then should do the same improvement there as in LCD systems. But they don't exist!
Also there is a change of optical system with a field lens. If the FL of fresnel is to short, the objective could run into troubles not to refract properly the 'cone' angles of the lightbeam.
The task for every 'projectorist' remains, to get out most efficient light of these damned panels!
xblocker
Hi Xblocker,
I will try to be more precise.
CRT projectors don't need a field lens because their objectives are generally large compared to the image they are projecting, and from what I have seen, they use a compound objective with multiple lens elements anyway.
I actually viewed the fresnel field lens as part of such a compound objective (but with a much larger aperture.)
I agree that a LCD is nowhere near as diffuse as a CRT.
My choice to treat it like a CRT is similar to the philosophical choice between front and rear-wheel drive. They both work in their own way, and both have their own limitations and compromises.
For me, the choice of putting a fresnel between the light source and the LCD is equivalent to the rear wheel drive choice: it involves hard work in producing an efficient point source of light with the right beam characteristics. Usually, the beam is not completely uniform, with greater intensity towards the middle of the beam. This beam is "pushed" by the fresnels through the LCD in a conical beam towards the objective. In this arrangement, the further away the light rays are from the center of the fresnel, the more shallow the angle of the light passing through the LCD. The more shallow the angle, the less efficient the LCD becomes - reducing the intensity of the image away from the centre even more. The LCD is slightly diffusive (though, I agree, more transmissive) and some of the light gets scattered. Any light that does not hit the objective is lost to the image. This arrangement works, and is used in commercial OHPs and is used by OHP LCD panels!
The other arrangement, with the fresnel between the LCD and the objective, I liken to front wheel drive: Much less strenuous demands are placed on the light source conforming to a single point source with the right beam profile. Instead, the emphasis is on producing an even illumination of the back of the panel with the beam (or beams) of light having "roughly" parallel characteristics. Some slight divergence allows beams from individual bulbs to be overlapped to reduce hotspots in a multi-bulb arrangement. The light emerges through the other side of the LCD, filtered to favour those rays roughly perpendicular to its surface (hence the reason for a roughly parallel beam). The viewing angle is a few degrees in all directions, so, while it is not vital to get the rays dead straight, a diffuse backlight will not be very efficient. On emerging from the LCD, the roughly parallel rays are "pulled" by a fresnel field lens into a cone of light that beams towards the primary objective (or second lens in a compound objective). This large fresnel captures virtually all the light that emerges roughly perpendicular to the LCD (they are very compatible technologies in this regard). Light that is very non-perpendicular to the LCD is totally lost to scattering within the fresnel. The fresnel has very restricted operating parameters, and must be placed very correctly to reduce horrific distortion, and also limits the flexibility of where the main objective can be placed. In order not to over-stretch the ability of the objective to work within its limits, and to reduce distortion due to the fresnel, the longest focal length possible should be chosen for the fresnel. However, the light cone produced by this arrangement should be similar in characteristics to that produced with the fresnel in the more normal OHP configuration. Thus, any objective that would normally be used for OHP projection should not have any problems here. This arrangement also works, and was used in commercial large panel projectors during the mid 90's.
For me, it is easier to place the fresnel at the optimum position, and worry less about the light source than it is to get the light source perfect, and worry less about the fresnel.
I agree that the task to get the most out of these panels is still in front of us!
Any clearer? 🙁
Bill.
I will try to be more precise.
CRT projectors don't need a field lens because their objectives are generally large compared to the image they are projecting, and from what I have seen, they use a compound objective with multiple lens elements anyway.
I actually viewed the fresnel field lens as part of such a compound objective (but with a much larger aperture.)
I agree that a LCD is nowhere near as diffuse as a CRT.
My choice to treat it like a CRT is similar to the philosophical choice between front and rear-wheel drive. They both work in their own way, and both have their own limitations and compromises.
For me, the choice of putting a fresnel between the light source and the LCD is equivalent to the rear wheel drive choice: it involves hard work in producing an efficient point source of light with the right beam characteristics. Usually, the beam is not completely uniform, with greater intensity towards the middle of the beam. This beam is "pushed" by the fresnels through the LCD in a conical beam towards the objective. In this arrangement, the further away the light rays are from the center of the fresnel, the more shallow the angle of the light passing through the LCD. The more shallow the angle, the less efficient the LCD becomes - reducing the intensity of the image away from the centre even more. The LCD is slightly diffusive (though, I agree, more transmissive) and some of the light gets scattered. Any light that does not hit the objective is lost to the image. This arrangement works, and is used in commercial OHPs and is used by OHP LCD panels!
The other arrangement, with the fresnel between the LCD and the objective, I liken to front wheel drive: Much less strenuous demands are placed on the light source conforming to a single point source with the right beam profile. Instead, the emphasis is on producing an even illumination of the back of the panel with the beam (or beams) of light having "roughly" parallel characteristics. Some slight divergence allows beams from individual bulbs to be overlapped to reduce hotspots in a multi-bulb arrangement. The light emerges through the other side of the LCD, filtered to favour those rays roughly perpendicular to its surface (hence the reason for a roughly parallel beam). The viewing angle is a few degrees in all directions, so, while it is not vital to get the rays dead straight, a diffuse backlight will not be very efficient. On emerging from the LCD, the roughly parallel rays are "pulled" by a fresnel field lens into a cone of light that beams towards the primary objective (or second lens in a compound objective). This large fresnel captures virtually all the light that emerges roughly perpendicular to the LCD (they are very compatible technologies in this regard). Light that is very non-perpendicular to the LCD is totally lost to scattering within the fresnel. The fresnel has very restricted operating parameters, and must be placed very correctly to reduce horrific distortion, and also limits the flexibility of where the main objective can be placed. In order not to over-stretch the ability of the objective to work within its limits, and to reduce distortion due to the fresnel, the longest focal length possible should be chosen for the fresnel. However, the light cone produced by this arrangement should be similar in characteristics to that produced with the fresnel in the more normal OHP configuration. Thus, any objective that would normally be used for OHP projection should not have any problems here. This arrangement also works, and was used in commercial large panel projectors during the mid 90's.
For me, it is easier to place the fresnel at the optimum position, and worry less about the light source than it is to get the light source perfect, and worry less about the fresnel.
I agree that the task to get the most out of these panels is still in front of us!
Any clearer? 🙁
Bill.
Woneill,
thanks for further explanation! Field lenses and CRT: Once upon a time...there where CRT projectors on the market, which only used a single color CRT, similar to this '100"TV' thing. The ojectives were large, but nevertheless smaller than CRT size. I had such a monstrum and tried always to increase brightness. Also had the idea with a field fresnel lens. But no improvement of brightness. I came to the conclusion, that only micro lenses attached directly in front of the phosphor layer of the screen could send all the light into one direction. But i never saw a patent or development on this.
Look at the 100"TV crap. If there would be optically a way to improve brightness with a field lens, they could easily give a second lens to the package..
Back to business: The light loss in parallaxial areas of lighting systems is a common fact, if parallel light or not, which is due to optical properties of real lenses. Even with modern MLAs there isn't a 100% uniform illumination!We could be happy if we reach 85% with DIY resources.
I agree, that it would be better to send light parallel through LCD, i'm only a little sceptical to bring a fresnel into the image path. Fresnels are good in lighting path, but in the image path, image can't be better than fresnels quality, although knowing not to focus on it!
Cya
xblocker
thanks for further explanation! Field lenses and CRT: Once upon a time...there where CRT projectors on the market, which only used a single color CRT, similar to this '100"TV' thing. The ojectives were large, but nevertheless smaller than CRT size. I had such a monstrum and tried always to increase brightness. Also had the idea with a field fresnel lens. But no improvement of brightness. I came to the conclusion, that only micro lenses attached directly in front of the phosphor layer of the screen could send all the light into one direction. But i never saw a patent or development on this.
Look at the 100"TV crap. If there would be optically a way to improve brightness with a field lens, they could easily give a second lens to the package..
Back to business: The light loss in parallaxial areas of lighting systems is a common fact, if parallel light or not, which is due to optical properties of real lenses. Even with modern MLAs there isn't a 100% uniform illumination!We could be happy if we reach 85% with DIY resources.
I agree, that it would be better to send light parallel through LCD, i'm only a little sceptical to bring a fresnel into the image path. Fresnels are good in lighting path, but in the image path, image can't be better than fresnels quality, although knowing not to focus on it!
Cya
xblocker
Projection system for sale!!!!
🙂 To all who might be intersted.
my friend has a projection system combo for sale.
Sharp QA-2500 Projection Panel & Dukane Quantum 680 Overhead Projector
these are top of the line products that provide excellent image quality if combined.
The set is for $430. This may seem a lot but it is well worth it. You get an XGA projection system that can take various video inputs.
Please contact me if you need any information....
aleksey😎
🙂 To all who might be intersted.
my friend has a projection system combo for sale.
Sharp QA-2500 Projection Panel & Dukane Quantum 680 Overhead Projector
these are top of the line products that provide excellent image quality if combined.
The set is for $430. This may seem a lot but it is well worth it. You get an XGA projection system that can take various video inputs.
Please contact me if you need any information....
aleksey😎
Hi Xblocker,
I agree that putting a fresnel in the image path is a scary prospect, and that they are definitely not up to being the main projection lens.
However, if you use a fresnel in a strictly controlled way within its "magnifier" range, then it doesn't do too much harm to the image.
To be more precise: the light source is placed closer to the fresnel than the fresnel's focal point, and the viewer (eye/objective) is also placed close to the fresnel's focal point on the other side of the fresnel.
As with any magnifier, the closer the object being magnified is to the lens, the less it gets magnified. Also, the less it gets magnified, the less it gets distorted.
For a fresnel, being used to bring the parallel beams of light from the LCD into a cone, the least distortion would result from placing the fresnel against the LCD itself. Unfortunately, then you get moire effects.
If the light emerging from the LCD is slightly non parallel, then moving the fresnel away by about an inch or so, reduces moire (light from a single pixel is then being bent by more than one point on the fresnel). This is just enough to remove moire without destroying image quality. If the light were perfectly parallel, then distance between the fresnel and the LCD would not matter - there would always be moire effects. (This is not intuitive, but it IS true!)
(The fresnel is effectively a set of concentric rings of micro prisms, and the moire effects result from light from a single pixel being scattered as it passes through a boundary between two of these rings.)
If the objective lens is placed at the apex of the cone produced by this arrangement, then it will have all the benefits of a normal OHP projection setup, with only tiny distortions introduced by the fresnel, but with a couple of additional cool benefits:
1) The light source does not need to be emulating a point source (hard to do efficiently with the big MH bulbs);
2) Any distortions introduced by the LCD will not affect image quality as they would in the OHP setup, because you are focussing on, and not through the LCD;
3) The image intensity can be made much more even.
Of course, the fresnel needs to be good, with lots of grooves per inch, and preferably with a longer focal length.
But, it DOES definitely work!!! My prime unit is now that modified DP9100 I keep talking about, and it uses this arrangement. Its internal optics generate an almost parallel beam from a 575 Watt MH bulb (the type used in the Dukane 680, and Elmo 305), they beam that onto the LCD, and use the exact arrangement I have been describing.
The image quality is beautiful, even with the reduced resolution of the Spectra C that was used to replace the original dead electronics (1024x768).
With an 8' wide image, I can focus it so clearly that the boundaries between the individual RGB shutters is CRISP.
This is not just another one of my whacky theories!!! 😉
I can be an obnoxious pain at times, but I am genuinely trying to contribute to a project whose success has knocked my socks off!!!
Bill.
I agree that putting a fresnel in the image path is a scary prospect, and that they are definitely not up to being the main projection lens.
However, if you use a fresnel in a strictly controlled way within its "magnifier" range, then it doesn't do too much harm to the image.
To be more precise: the light source is placed closer to the fresnel than the fresnel's focal point, and the viewer (eye/objective) is also placed close to the fresnel's focal point on the other side of the fresnel.
As with any magnifier, the closer the object being magnified is to the lens, the less it gets magnified. Also, the less it gets magnified, the less it gets distorted.
For a fresnel, being used to bring the parallel beams of light from the LCD into a cone, the least distortion would result from placing the fresnel against the LCD itself. Unfortunately, then you get moire effects.
If the light emerging from the LCD is slightly non parallel, then moving the fresnel away by about an inch or so, reduces moire (light from a single pixel is then being bent by more than one point on the fresnel). This is just enough to remove moire without destroying image quality. If the light were perfectly parallel, then distance between the fresnel and the LCD would not matter - there would always be moire effects. (This is not intuitive, but it IS true!)
(The fresnel is effectively a set of concentric rings of micro prisms, and the moire effects result from light from a single pixel being scattered as it passes through a boundary between two of these rings.)
If the objective lens is placed at the apex of the cone produced by this arrangement, then it will have all the benefits of a normal OHP projection setup, with only tiny distortions introduced by the fresnel, but with a couple of additional cool benefits:
1) The light source does not need to be emulating a point source (hard to do efficiently with the big MH bulbs);
2) Any distortions introduced by the LCD will not affect image quality as they would in the OHP setup, because you are focussing on, and not through the LCD;
3) The image intensity can be made much more even.
Of course, the fresnel needs to be good, with lots of grooves per inch, and preferably with a longer focal length.
But, it DOES definitely work!!! My prime unit is now that modified DP9100 I keep talking about, and it uses this arrangement. Its internal optics generate an almost parallel beam from a 575 Watt MH bulb (the type used in the Dukane 680, and Elmo 305), they beam that onto the LCD, and use the exact arrangement I have been describing.
The image quality is beautiful, even with the reduced resolution of the Spectra C that was used to replace the original dead electronics (1024x768).
With an 8' wide image, I can focus it so clearly that the boundaries between the individual RGB shutters is CRISP.
This is not just another one of my whacky theories!!! 😉
I can be an obnoxious pain at times, but I am genuinely trying to contribute to a project whose success has knocked my socks off!!!
Bill.
Re: Projection system for sale!!!!
hmmm, how do we know it's not yours, and you're trying to sell it?
;-) hehehe just messin around.
again, I'm curious, but what is the ideal amount of lumens people here are aiming for? 60k, 30k? lower, higher?
prjctr_builder said:
hmmm, how do we know it's not yours, and you're trying to sell it?
;-) hehehe just messin around.
again, I'm curious, but what is the ideal amount of lumens people here are aiming for? 60k, 30k? lower, higher?
Lumen measurement?
If a body happens to have a plain old-fashioned light meter, could the lumens of our setups be measured? If we could start doing some measurements and share them, we might be able to make our efforts more scientific and productive rather than subjective.
multiplexor said:
If a body happens to have a plain old-fashioned light meter, could the lumens of our setups be measured? If we could start doing some measurements and share them, we might be able to make our efforts more scientific and productive rather than subjective.