Yes, I had to put the 90-deg sides facing in too because otherwise the lens was a bit too small for my projector too.
Ignore that first post where I talked about it being better with the 90-deg facing the projector and screen...I had the wrong prism layout for that first test.
However, read my web link again. In the 2-prism test you will see that I get 2 reflections, both caused by the small sides of the prisms that are easily removed with masking.
In the 4-prism lens, the reflection that I can't remove with masking is labelled as "D" and is on the far left side of the screen. So that *does* match what Bud and others are seeing. Depending upon the exact light path of the laser, I can get either one or two reflections on the left side of the screen.
Obviously with the laser I can't comment on the green-streak issue. All the laser test does is map out possible reflections from various light paths.
Anyway, take a look at the last laser diagram on the site again and I think you'll see what I mean.
What I conclude from the laser test is that a 2-prism lens should be fine (no reflections with proper taping of the prism sides). In a 4-prism lens, when there is bright text on the *right* side of the screen, you will get a small reflection on the far *left* side of the screen. And this seems to match exactly what was being mentioned with the Cars credits.
It this *is* the same reflection source, then I might be able to learn more about it from the laser, even though the chance of removing it seems unlikely.
Maybe it's back to figuring out how to reduce CA in the 2-prism lens.
Ignore that first post where I talked about it being better with the 90-deg facing the projector and screen...I had the wrong prism layout for that first test.
However, read my web link again. In the 2-prism test you will see that I get 2 reflections, both caused by the small sides of the prisms that are easily removed with masking.
In the 4-prism lens, the reflection that I can't remove with masking is labelled as "D" and is on the far left side of the screen. So that *does* match what Bud and others are seeing. Depending upon the exact light path of the laser, I can get either one or two reflections on the left side of the screen.
Obviously with the laser I can't comment on the green-streak issue. All the laser test does is map out possible reflections from various light paths.
Anyway, take a look at the last laser diagram on the site again and I think you'll see what I mean.
What I conclude from the laser test is that a 2-prism lens should be fine (no reflections with proper taping of the prism sides). In a 4-prism lens, when there is bright text on the *right* side of the screen, you will get a small reflection on the far *left* side of the screen. And this seems to match exactly what was being mentioned with the Cars credits.
It this *is* the same reflection source, then I might be able to learn more about it from the laser, even though the chance of removing it seems unlikely.
Maybe it's back to figuring out how to reduce CA in the 2-prism lens.
Don't take my comments as negitive. Your contributions are excellent. It is very refreshing to read someone else's posts that have as much enthusiasm as I have for this lens project to work.
I have just received an email about a possible fix for the reflections. Off to follow that up now...
Mark
I have just received an email about a possible fix for the reflections. Off to follow that up now...
Mark
Hey Mike, thanks for jumping in with both feet! The more working on this, the faster we'll fix it. Kewl idea about the laser, never thought of it. When you're doing it, look on the inside wall, the right side when looking in from the front and you'll see a great deal of reflection there. I believe THAT may be the culprit. When playing CARS(on my 4 prism setup) that area between lenses shows a couple of reflections on top of another. I'm trying to place a reflection absorbing material(black cloth) there to see if it helps.
Keep ya posted all!
Bud
Keep ya posted all!
Bud
Mark Techer said:
Mark I read this last week and think it is great explanation, easy to understand and well laid out. My only feedback would be the pictures that you used in your explanations.
I think the samples you chose for the pictures are quite dark, particularly the edges of some them, I think this reduces the impact of widescreen, when the edges of the screen tend to fall into the blackness of the room.
Minor I know, but if you could find a 'bright' scene that filled the screen, I think it may show better.
We need anti reflection coatings
Bud,
The reflection you see on the inside wall (I have one too), is it mostly green?
Also when I got my lens to work as a VC lens, there was a perfect "green" image on the roof...
Mark
Bud,
The reflection you see on the inside wall (I have one too), is it mostly green?
Also when I got my lens to work as a VC lens, there was a perfect "green" image on the roof...
Mark
Re: We need anti reflection coatings
Nope, no green. I think you may have one prism slightly off. Did you try swapping them all out?
Bud
Mark Techer said:
Nope, no green. I think you may have one prism slightly off. Did you try swapping them all out?
Bud
Re: Re: We need anti reflection coatings
Thanks roar,
I have also added some other images just tonight showing both 16:9 and 21:9. These images are very bright and clear. I also slightly changed the text which is the real reason I was asking for feedback...
I tried that. This image on the roof was so perfectly clear and green. Ever seen true RGB video if you pulled the red and blue leaving the syncs (H and V) connected?
Mark
roar said:
Thanks roar,
I have also added some other images just tonight showing both 16:9 and 21:9. These images are very bright and clear. I also slightly changed the text which is the real reason I was asking for feedback...
Bud Bray said:
I tried that. This image on the roof was so perfectly clear and green. Ever seen true RGB video if you pulled the red and blue leaving the syncs (H and V) connected?
Mark
Re: Re: Re: We need anti reflection coatings
I didn't think any of those were new pictures, there are just Cars pictures, 3 different ones I believe with the third one in 3 different ways. I did re-read it quickly... the last post says Nov 6th, is it a new post or an update to that one? I tried clearing my cache and refreshing, things must be slow making their way over the pacific this morning 🙂
Mark Techer said:
I didn't think any of those were new pictures, there are just Cars pictures, 3 different ones I believe with the third one in 3 different ways. I did re-read it quickly... the last post says Nov 6th, is it a new post or an update to that one? I tried clearing my cache and refreshing, things must be slow making their way over the pacific this morning 🙂
It is actually about the 4th re-do of the original November 6 post...
The two new images are from when Bud asked me to shoot the credits of CARS the other night, so I decide to take a whole heap of shots from that film. I just really liked the shot with Chick and Lightning...
I also needed a half decent 16:9 image because most of the prvious 16:9 images I have taken really suck, but this is OK, and it looks fairly straight.
The goal here is to try and show that CIH is about different ARs at projected at the same height, and also how to achieve it through scaling and optics.
I've kept the original 3 images because they show the screen's border, so is not just a case of taking a photo of the scope image which could be from a 16:9 screen if the blacks are deep enough...
I've also tried to keep the same angle and distance so that the shots are comparible...
Mark
The two new images are from when Bud asked me to shoot the credits of CARS the other night, so I decide to take a whole heap of shots from that film. I just really liked the shot with Chick and Lightning...
I also needed a half decent 16:9 image because most of the prvious 16:9 images I have taken really suck, but this is OK, and it looks fairly straight.
The goal here is to try and show that CIH is about different ARs at projected at the same height, and also how to achieve it through scaling and optics.
I've kept the original 3 images because they show the screen's border, so is not just a case of taking a photo of the scope image which could be from a 16:9 screen if the blacks are deep enough...
I've also tried to keep the same angle and distance so that the shots are comparible...
Mark
OK Mark, here's a question for ya. Maybe i've missed something along the way but...DaVinci Code is 2.40:1, Cars is 2.39:1 yet your set up shows no difference, not even a little?
I'm pretty sure (have to double check this tho) that since i'm set up for 2.37:1, a 2.40:1 does show a slight black bit top & bottom, maybe just an inch.
Bud
I'm pretty sure (have to double check this tho) that since i'm set up for 2.37:1, a 2.40:1 does show a slight black bit top & bottom, maybe just an inch.
Bud
Bud Bray said:
And it is a good question too, so I'll try and answer it best I can...
The native panel is 16:9 or 1.78:1 made from (in my case) just 480 vertical pixels. Some of the more popular panels are as follows -
854 x 480
1024 x 576
1280 x 720
1388 x 768
1920 x 1080
The anamorphic lens for video is 33% HE.
1.78:1 + 33% is 2.37:1.
The black bars are about 25% of the vertical rez.
In HD (1920 x 1080) standards that is the same as 1920 x 810. it would actually take a pixel structure of 2560 x 1080 squeezed sown into the 1920 x 1080 to be true anamorphic HD...
So if the image is squeezed or stretched by a factor of 33%, but the black bars only occupy 25%, there is going to be some room for error. I say 33% - 25% = 8%,so it is about 4% top and 4% bottom that gets clipped.
Also you state that CARS has an AR of 2.39:1. That may be the case of the R1, but the R4 is credited as 2.35:1. Combine this with the fact that my projector also has some over scan issues, and any slight differences like you have asked about will not be seen on my set up, so might explain why you also commented that I had "chopped off the ends" where I didn't do any such thing.
I think that I worked out that a film would have to be 2.55:1 or greater before I would see any letterboxing on my screen, so the differences between 2.35:1 and 2.40 are not noteable...
I just have to work with what I have got from this very ordinary projector until i can replace it (soon I hope)...
Mark
Wow, MikeP, this is really fascinating stuff. I've got to get back into this. I was going to last night, but we had a second night (and a second kid) vomitting through early evening to the wee hours of the morning. Don't these kids know I have important anamorphic lens experimenting to do???
Mark, I took a look at your blog (DIH Explained) and have just a couple comments. Overall I think it is very good. The text explaination is good.
I showed it to my wife to see if she could understand it (I know, that's evil). But actually she did! But she got a bit confused by the pictures and their captions. Sometimes you will be talking about the previous picture, but with the next picture right under the text, it's confusing which picture you are referring to.
I've run into this problem myself. What I've done is to put the bold "image caption" above the picture, rather than below it.
So, instead of:
(picture)
OPTICALLY STRETCHED IMAGE
(text)
you would have:
OPTICALLY STRETCHED IMAGE
(picture)
(text)
Putting the header before the picture makes it a bit clearer that you are starting a new section and that the picture belongs to the new section and not the previous one.
I also have to agree with the comment about the images. I think your first image with Chick and Lightning is great. But the image that you use to show the 2 steps of creating the CIH isn't as good because it's very dark across the top. So, in the first picture where you are talking about the black bars on the top and bottom of the image, it's harder to see the black bar along the top.
I haven't re-watched Cars in a while. I know it's always great to watch high quality images like Pixar produces. However, for a real show of how CIH works, it might be better to find a nice bright move with a scene that looks more obviously stretched vertically in step 2. I often describe the scalar test as stretching the image so that people look tall and skinny, so maybe an image with people might help. Or maybe an image with something circular so that you could see the circle becoming an elipse. Cars is great quality, but since it's all a bit "cartoony" it might be hard to recognize the vertical stretch for a common viewer.
Anyway, those were my minor comments. Overall it was very good and I think we need more of that kind of stuff to help convert the rest of the HT world 🙂 After all, we've got to get those projector manufacturers to start giving us the internal scalar modes that we need for CIH. I was very depressed last night when I learned that my new InFocus 7205 won't stretch 720p or 1080i material on the component inputs, which is where all my hidef material comes from. So until I get a HD cable box with DVI or HDMI and then buy a HDMI switch box (or spend tons of money on an external scalar) then I'm stuck with only watching normal DVDs in Scope format. I was really annoyed at InFocus for ignoring us like that.
I showed it to my wife to see if she could understand it (I know, that's evil). But actually she did! But she got a bit confused by the pictures and their captions. Sometimes you will be talking about the previous picture, but with the next picture right under the text, it's confusing which picture you are referring to.
I've run into this problem myself. What I've done is to put the bold "image caption" above the picture, rather than below it.
So, instead of:
(picture)
OPTICALLY STRETCHED IMAGE
(text)
you would have:
OPTICALLY STRETCHED IMAGE
(picture)
(text)
Putting the header before the picture makes it a bit clearer that you are starting a new section and that the picture belongs to the new section and not the previous one.
I also have to agree with the comment about the images. I think your first image with Chick and Lightning is great. But the image that you use to show the 2 steps of creating the CIH isn't as good because it's very dark across the top. So, in the first picture where you are talking about the black bars on the top and bottom of the image, it's harder to see the black bar along the top.
I haven't re-watched Cars in a while. I know it's always great to watch high quality images like Pixar produces. However, for a real show of how CIH works, it might be better to find a nice bright move with a scene that looks more obviously stretched vertically in step 2. I often describe the scalar test as stretching the image so that people look tall and skinny, so maybe an image with people might help. Or maybe an image with something circular so that you could see the circle becoming an elipse. Cars is great quality, but since it's all a bit "cartoony" it might be hard to recognize the vertical stretch for a common viewer.
Anyway, those were my minor comments. Overall it was very good and I think we need more of that kind of stuff to help convert the rest of the HT world 🙂 After all, we've got to get those projector manufacturers to start giving us the internal scalar modes that we need for CIH. I was very depressed last night when I learned that my new InFocus 7205 won't stretch 720p or 1080i material on the component inputs, which is where all my hidef material comes from. So until I get a HD cable box with DVI or HDMI and then buy a HDMI switch box (or spend tons of money on an external scalar) then I'm stuck with only watching normal DVDs in Scope format. I was really annoyed at InFocus for ignoring us like that.
MikeP said:
Mark: I also looked at your website, and MikeP summed up everything nicely for me as well.
The internal scaler will definitely be something that must be present in the next projector that I get. I do have it right now, interestingly, in my Benq 6100, which is SVGA, yet it allows me to stretch 2.35:1 to fill the entire panel, which is great.
I am still of the opinion that I can get the lenses to give me the 2.37:1 stretch from 1.33:1 source material...
MikeP said:
OK thank you MikeP 🙂 Now that you have pointed out the fact that the pictures need their titles above, it is simply a case of a another quick edit...But I'm just glad that is makes sense...
I see you point. Infaxct the othger day, I was showing a guy the how it works and removed the lens with Star Wars, and the most obvious change (len in Vs Lens out) was R2D2's dome head. I will look into taking some more shots...
Steve Scherrer said:
If your in the market for a new projector, why not just buy a 16:9?
How much are the BenQ7700(?). Sorry I don't have the model number, but they have 3 DLP 16:9 units where the two are 720 machines, and the one I am thinking of retails for just $2500, and are available way cheaper then that with the right connections...
Mark
Yes, fantastic posts, Mike -- I found them really helpful in ascertaining whether I want to try the 2 or 4 prism setup.. As I suspected, the 2 prism lens probably exploits the best these glass blanks have to offer. The 4 prism lens seems hamstrung by the 'diminishing returns' bell curve 🙂 ... but there may still be an affordable diy fix round the corner...
Mark: I'm actually not in the market for a new projector. I will keep this one for a while longer. I was just stating that the one I do have (SVGA) includes a full panel stretch--which is perfect for infinitely adjustable prisms...
Steve Scherrer said:
Cool, I just think that you need greater scaling options than just the 16:9 mode of a DVD player - HTPC or out board scaler will be required for a 4 x 3 conversion...
Still curious as to just how far these prisms will allow the light to stretch before becoming to distorted...
Mark
I played with the laser a bit more tonight to try and identify those two unknown reflections (reflections D on the 4-prism test on MikeP's Theater )
These two reflections are from the primary beam hitting the inner 2 surfaces of the outboard (closest to projector) prisms. If I spread these 2 prisms apart, I can move the reflections, but not get rid of them.
These reflections are not coming from any of the side or end surfaces. It's the main light path. So there is no way to mask them.
I'm going to play around with spreading these two prisms apart, or tilting them, to see what happens to the stretched image. I haven't decided if they really need to be butted together like in the diagrams.
Bud, have you tried playing with the spacing between the pairs of prisms?
I'll probably wait until my 7205 arrives on Sunday to really play with real images though. I need to play around with the prisms and then really decide if I'm going to do a 2-prism or 4-prism setup. I need to look at the CA from the 2-prism setup and see if I can live with it, and then compare it to the reflections in the 4-prism setup. Then design a nice housing. Lots to do. But heh, that's why this is a "hobby" right? Otherwise, people like Mark wouldn't have spent years on all of this 😉
These two reflections are from the primary beam hitting the inner 2 surfaces of the outboard (closest to projector) prisms. If I spread these 2 prisms apart, I can move the reflections, but not get rid of them.
These reflections are not coming from any of the side or end surfaces. It's the main light path. So there is no way to mask them.
I'm going to play around with spreading these two prisms apart, or tilting them, to see what happens to the stretched image. I haven't decided if they really need to be butted together like in the diagrams.
Bud, have you tried playing with the spacing between the pairs of prisms?
I'll probably wait until my 7205 arrives on Sunday to really play with real images though. I need to play around with the prisms and then really decide if I'm going to do a 2-prism or 4-prism setup. I need to look at the CA from the 2-prism setup and see if I can live with it, and then compare it to the reflections in the 4-prism setup. Then design a nice housing. Lots to do. But heh, that's why this is a "hobby" right? Otherwise, people like Mark wouldn't have spent years on all of this 😉