State of DIY anamorphic lens

If you try anything, let us know, I am interested in what you come up with. Do you have a source to get prisms cheap to experiment with?

Not in a DIY approach as no-one can DIY the optics that are required for CA and Astigmatism correction. This must be designed and built to exacting tollerances. The pro lenses certainly do allow full 1080 detail to be seen with no losses.

All lenses will allow you to recoupe the full 1080 lines as the image itself is scaled to fill the panel's 1080 rez prior to optical expansion.

The first 3 in that list are just 2 prisms lenses, so have no correction and in fact the Aussiemorphic 1 and 2 are discontinued because of that fact.

The next is a CA corrected 2 prisms (4 element) lens and the last is a fully corrected cylindrical lens which is in a different class altogether.

The key difference between cylindrical and prisms lenses is that cylindrical lenses like the ISCO can be brought into focus in both the horizontal and vertical plain at the same time. One of the limits of a prism lens is that they tend to only focus in one plain at a time, where the vertical rez (horizontal lines) are more important.

Without correction, the anamorphic lens will produce an image that is clear in the centre but gets progressively out of focus towards the edges. A huge portion of this "defocus" is CA, then there is anstigmatism grid distortion etc. It takes additional elements to correct for these abberations.

Surface quality is also important, where the "trophy" lense do not have a true optical coating. The MK3 is not a trophy lens BTW and it's prisms are finished to a true optical grade.

I am not sure if I can explaion this correctly, but given any fixed height of screen, the lens method will be denser than the zoom as it allow 100% of the vertical rez to be used where the zoom only uses 75%. The pixels themselves will become 33% larger for the zoom. There is only 810 lines used out of the 1080 for a Scope film. You can't get more lines that what is there, but the way in which we project them can make a difference.


Curved screens are usually used to correct abberations like pincushion, but in some cases, the curved screen is used to focus the light to a centre point for a brighter image. I personally would not use a curved screen without a lens simply because the optics in HT projectors are designed for a flat surface. Projecting onto a curved surface only adds problems.


Thank you. I only wish certain individuals would stop reporting it as SPAM and trying to have it deleted.

Having actually tested this, I can honestly answer this part. The cheap 2 prisms (trophy lens) can only resolve a portion of the pattern.

The image will be clear in the centre but get progressively softer towards the edges as both CA and Astigmatism come into effect. The other key point is that the prism lens can only be focused in one direction at one time, so whist you will be able to focus the horizontal lines (vertical rez) all the way to the edges, the vertical lines (horizontal rez) becomes quite blured (even to the point of smear) from about the 1.78:1 area. This is a very demanding test and only the best anamorphic lenses can resolve this anyway. In some cases, the projector's lens can add artifacts as well which only get magnified by the anamorphic lens. My BenQ for example has quite a bit of CA, so the anamorphic lens simply magnifies this as it is already in the light path.


ONLY a lens with fully adjustible astigmatism will allow true corner to corner focus on a curved screen. Allot of this actually is not the anamorphic lense's fault, but rather the fact the projector's lens is designed for a flat surface, but it can be made to work. HT projectors have their roots in data projectors where they are designed to work off a table at the base of the screen. Lens shift has allowed projectors to be moved closer to the centre of the screen, but their optics are not like film projectors that work off the centre.


Thank you for the moral support From what I know, CIH for video is about 7 years old now. This is the 3rd year for the Aussiemorphic Lens.

The DIY Lens thread was origianally started with liquid filled prisms for a VC lens and moved on to HE and pretty much stayed that once trophies were dicovered.


If you were to stay with a 2 element design, then yes, the VC has merrits, but...


...Allot also comes down to image size as well. The VC shrinks the 16:9 vertically to make a 2.37:1 image. The pixels are now compressed and abberations at that size are not very noticealble. It became a slightly different story when you blow that image up to the the same size as the image created by a HE lens. The difference now is that any CA, astigmatism etc will run horizontally not vertically and program pending, this may be better of worse. I found it to be worse and why I became interested in a CA corrected VC lens. I like the idea that the VC gives a longer throw, no pincushion etc so there are still many pluses for it.


I thought the "french prisms" were coated. Regardless, it is not a huge price to pay for an experiment that may prove a positive result.

The "French Prisms are small and when you rotate them for a VC lens leaves you with a very small apperture. You can get the trophies coated.

I've only seen a photo of the lens your talking about and the person that posted the shot seemed to put it in the too hard basket. The prisms looked small, but I think it would have worked if the throw was long enough.

To state this simply: There IS NO cheap soloution for CA correction as not only is the interaction of the two glass types critial, but so are the angles and the surfaces where the glass is cemented. Even the bonding process is crital as you can not have bubbles in the join either. If this is not 100% perfect, the new prism will not work.

Yes it is...

dougri

Did you ever try this experiment? Are you doing anything with the diy trophies?