DIY Anamorphic Lenses Using Trophy Prisms

Very nice work. So far, this is really smooth. I am pretty compulsive and this is looking like something I might consider doing (althought I usually just read these threads vicariously). I am looking forward to the rest of the build.

Where do you get alcohol based opaquing fluid?

Eroc

Completely eliminated or "mostly" eliminated?

Do you plan to use them in a rotating "stretch-pass" config or a movable sled config? If sled config have you checked for vertical image shift introduced by the lens?

Thanks!

I have just knocked my lens up in a box and it looks awsome

I can't believe something like this is available to the average joe with under £50 to spare.

That price includes the screen btw. I bought some timber, metal mounts, corner fixings, BOC and black masking material for about £30. I've not put the screen together yet but holding it up it's far better than even the InFocus screen I'm using now! The picture looks at least 20% brighter

Oh, and it.. is... HUGE

-Joel

Welcome to the club! Yes, there are pretty expensive alternatives that have correction features and optical coatings, etc. But the price difference for what I believe are pretty incremental advantages is simply amazing.

I don't think I'd go that far. I'd call the difference between the top of the line Prismasonic and the ISCO III incremental. The differences between either of those and a DIY or low cost 2 prism solution are night and day.

True, it is satisfying to many and that is great, but to call the advantages of the very good lenses "incremental" is a gross misrepresentation.

I actually think you are possibly over stating that a bit. I can't see any flaws in my setup at all (maybe it could look a bit tidier?) but now that I have my screen up I can honestly say I have no desire to swap it out for a commercial lens and see no real reason to.

I had just one reflection on the screen caused by light reflecting back into the projector lens but one small strip of masking tape later and the screen is reflection free.

I also think the focus is pretty sharp too which I was expecting to go right out.

I'm by no means a video projection engineer but I have been in this game for long enough to know this simple solution, if executed properly, is a great alternative to commercial products and can produce fantastic results.

-Joel

Any progress BungalowFilms?

I completely agree. This is certainly true in many areas.

I don't know what your anamorphic solutions is (PC trophies or what), but have someone bring over an ISCO III and compare them. There is NO way you won't see a difference.

I didn't say that a DIY or simple 2 prisms is not decent or not good enough for the vast majority of people, I just said that the difference between an ISCO III and a pair of trophy prisms is not "incremental". If you really believe that, you have not really compared them.

BTW, here is my simple rule: if my wife can't see an improvement that I can see--it is incremental. If she can see a difference--it is larger than incremental.

Definitely curious to know what you find. Also, are you still using the Panny 720p LCD or a higher CR machine? I think with the low CR of the 720p LCDs, reflections are harder to see. Definatley try some bright white credits on a black background.

I'm very curious to hear what you find.

Well that I can agree with

-Joel

Dave,

Also, I will reiterate my request here again, which I already posted in the "Prism Sources" thread:

Could you please comment on picture sharpness while using the prisms? with panny's "smoothscreen" technology, the picture is somewhat soft compared to pj's without it, and now with added prisms, the projected image would be softer than before. A few screenshots with and without the prisms would be helpful. Thanks.

I use the same criteria, unfortunately my wife seldom notices an improvement - or else she won't admit to seeing a difference so I won't spend any more money

When my wife does she a difference, she always caveats it with, "but not enough to justify the expense."

This just only reinforces a point I've recently started to talk about on AVS - optical surface finish. Even though these prisms used "appear" optical grade, the thuth is they are not as they were never designed to be optical components, but trophies.

The glass type may very well be BK7 and finshed to an optical finish, but with out knowing the actual specs, the claim is loose at best and if your seeing noticeable change, then sorry they are not optical grade.

CA is of course another issue and only correctable with the addition of extra elements - and no you can not just use 4 prisms together as you need to use not only a different glass type and bond them correctly, but also have to know the precise angles to make it work. This then of course takes this well away from the DIY sector.

Your comparison is very good. Now I would like to see this done with a propper lens...

Mark

Dave, different prisms have different amounts of CA. Also, the CA is dependant upon careful alignment, if the PJ's lens is not perfectly center and using no horizontal shift and the prisms are not aligned perfectly, you will get worse CA.

The CA on the right side of your image is excessive even for trophy prisms. I suspect you could get it better--but it won't be great. As an aside, with VERY careful alignment and set up, I've got an HTB lens set up so that the worst CA at the worst side of the screen is about as bad as your "S" in "Pirates". Nothing even close to being as bad as what it is on your "pearl".

Lastly, I assume you are using your PJ to do the vertical stretch. I've not looked at the capabilities of the Panny there to see how good the scaling algorithm is. A 2.35:1 movie vertically stretched on your 720p without the lens should be more detailed than it shown in letterbox on the 720p. If it is not, it is because they are not doing the scaling correct--you should look at that.

The reason being, most scope movies have a vertical resolution of the active image area of 800-817 pixels. Scaling that down to 720 should result in very little loss of detail. If however, they are downscaling to 500-something then upscaling for the stretch, you'll loose quality.

I'm currently using a 720p projector. The scaling is being performed by a Lumagen which takes a 1080p/24 signal, and scales to 720p/72 (locked) which is then sent to the projector with 1:1 pixel mapping. Given that horizontal resolution is less important to human perception than vertical, and downscaling 800 to 720 results in about the same loss of detail as upscaling 800 to 1080, my 720p actually does scope content pretty impressively (many of those in my neighborhood think it looks better on scope content than one of the other neighbors Panny AE2000).

I would say the best thing to do is use an external processor for the high quality scaling (and because it can also correct for any picture differences introduced by the lens), and try some other solutions. Solutions like the HTB or Mark's are not going to give you ISCO III performance, but for a reasonable price, they can give you much better performance that your screenshots especially if paired with quality scaling.

I have found the ghosting with the "French" prisms to be bad, but their focus preservation was pretty good--since the native panel CR of your current PJ is low, they may work for you.

The current HTB prisms actually have pretty good focus preservation. They have a 15 day return policy if you try and don't like.

Which trophy prisms are you using? I've got the 4"x6" versions from Precision Crystal and I'm not seeing any vertical shift at all. With my setup I do get some pretty bad ghosting, though, even after having masked the edges of my prisms with black electrical tape. I also painted the inside of my lens casing black, and went so far as to try sticking black velvet on the inside walls of the thing, too, just to see if that worked any better than the black paint. It didn't -- at least not that I could tell. I plan to keep monkeying with it to see if I can make things any better, but for me, at least, it's still nowhere near bad enough to just scrap the whole project.

Yes.


very common with prisms. Basically, you need to try them in every combination of orientation with regard to each other. Swap the front and back. Turn one over 180 degrees, etc. If you are lucky, you'll find a combination with little to no vertical shift. Although more often then not, there will still be a little, so you will have to use your PJs vertical shift to fine tune if you are using them with a movable lens.

Part of the problem with ghosting is that we are not all comparing apples to apples. On a 720p LCD (which has a poor native panel CR), the black is so elevated, ghosting does not look nearly as bad as it does on a PJ with a native CR of even 4000:1 to say nothing of something like an RS2. Elevated blacks hide or diminish the impact of ghosting.

I've only got a calibrated 3800:1 on my current PJ and I can't stand any ghosting. I did scrap the DIY prisms due to the issues.

To BungalowFilms - Great comparison! I really appreciate that you to the time and effort to make a well controlled and objective comparison.


When can you start?

What do you mean by this? Was one of the faces of the prism parallel to your PJ's lens? If so, it should have been at an angle to the PJ lens.

Yes, sharpness is an issue with the trophy prisms. The "French" prisms are better at preserving sharpness.