Homemade 35mm Adapter

[James Hurd]

Ben,
Go to a sign shop or Fastsigns and ask for some Etched Looking vinyl. Slap some of that on the disk and you'll save hours of grinding.

[Bob Hart]

Ben.

The 5 micron grit I refer to is aluminium oxide. Francis Lord Optics supplied me with some when they sent me the prisms. You would not need much at all. The AO5 is preferable with glass as it tends to pit rather than gouge the glass as silicon carbide does. However with the plastic clear CD-R, silicon carbide may be okay.

Silicon carbide can be found from optical workshops, gemstone or lapidary suppliers. I think the silicon carbide grade would be 1200 but I am not sure. In the lapidary scheme of things you would need the last finest grade of silicon carbide used for gemstone tumbling before the tin oxide for polishing stones.

There is another way you can get a groundglass finish on a plastic disk with control over scratching and that is by using a wet and dry silicon carbide paper of about 600 grade which is coarser than ideal. You place your clear CD-R disk on a firm flat smooth surface with a piece of thin soft cloth under it to protect the clear side.

Place the wet and dry paper grit-side down on top of that. Find a hard object about the size and shape of the blunt end of a pencil. The shiny end of a pushrod from a car engine would be okay. Otherwise get hold of a used alternator bearing from the local auto-electrician or any other bearing you can find.

If you have the pencil-like object, rub this over the back of the wet and dry paper firmly to press the grit hard against the CD-R. Take care not to let the paper skid across the disk otherwise it scratches instead of stamping pits into the CD-R.

If you have the bearing, roll this across the paper with the bearing tilted slightly to pressure one corner. This is an easier and better method and does not tear the paper to pieces like the pencil object does.

Frequently move the paper by picking it up and not sliding it, so that variations in the grit density on the paper are randomised across the disk.

This is a tedious and tiring process due to the pressure you have to exert on the roller or pencil shaped object and takes about an hour or two to do completely across all the area of the disk.

James' suggestion about the stick-on material is probably better.

If you end up using grit in water, the solution should be a thin watery slurry not muddy. The motion should be small circular orbits. To avoid localised overpressures on the CD-R disk which bends, use a piece of white foam or a block of wood to move the disk around in the slurry on a sheet of glass. Don't let the slurry get dry. don't use pressure on a plastic CD-R when dressing it this way.

[Bob Hart]

A furthur shot of a test pattern through the prism version yielded slightly better than the stated resolution of a PD150P at 530 TV lines. The optical resolution appears to crash at about 670 TV lines although this is more or less theoretical as digital artifacting is also occurring at about 710 TV lines on the test pattern.

The optical path was :

50mm f1.8 Nikkor prime lens wide open > GG (same oharadisk dismantled out of the non-erecting version) > 2x 40mm x 40mm x 56mm x 40mm common thickness right-angle prisms > Century Optics 7+ achromatic diopter > camcorder.

Lighting conditions were one overhead household 75watt incandescent globe in frosted ceiling dome. Camcorder was on full auto.

[Ben Gurvich]

Are you supposed to film everything at 1.8? If i go to 4 on my Nikkor, the lens size gets about 50% smaller meaning i have to zoom in more and lose more light.

Also I went to my local signs shop and they gave me a free offcut of the ethec looking vinyl, however, my camera could not see thru it as it were too dense. Im glad i didnt have to pay for it.

Cheers,
Ben

[Brett Erskine]

Ben-
To fix the image from getting smaller either get a (better) plano-convex lens(es) next to that GG or do what P+S Technik did - add a secondary iris. Good luck.

[Bob Hart]

Ben.

Second Brett's suggestion. As an alternative or addition to a secondary iris, if you need to control light into the camcorder to force the shutter speed down, neutral density filter material between the gg and lens helps. I'd suggest 1x f-stop pieces of gel filter which is like thin plastic and can be cut up with scizzors. If you want more than 1x f-stop you can add layers.

The secondary iris Brett refers to is I think in the relay stage between the groundglass and the camcorder. Brett can correct me on this.

As well as the secondary iris, P+S Technik apparently do something before the groundglass with a number of 1x f-stop neutral density filter increments selectable by a control.

f1.8 is better for depth-of-field effects, however, not all lenses are equal and some fail for resolution when set wide-open.

How large is the image frame you are taking off the groundglass? The still-camera frame will give you some problems unless you do as Brett suggests. The smaller movie frame, which sits across the film, not along it like the still-camera frame should be more manageable. Brett's test chart is handy if you can print it to be exactly 24mm wide by 18mm high.

Otherwise, hand-draw a target to this size and place it where your groundglass image is and see how much of your image is covered by this target. Ideally, for faithful replication of the 4:3 movie frame, it should be tight in your camcorder frame edge-to-edge.

How thick is the vinyl? If you take your lens outside and project an image onto the registration sticker on your car windscreen, this should give you some idea of what to expect from the vinyl. If the vinyl isn't the same then it likely won't work.

[James Hurd]

Ben,
There are several thicknesses available for the vinyl. The stuff I have is very thin.

Good luck.

[Ben Gurvich]

<<<-- Originally posted by Bob Hart :

f1.8 is better for depth-of-field effects, however, not all lenses are equal and some fail for resolution when set wide-open.<<<--

I am using a Nikon Nikkor 1.8 50mm Lens,

<<<-- How large is the image frame you are taking off the groundglass? The still-camera frame will give you some problems unless you do as Brett suggests. The smaller movie frame, which sits across the film, not along it like the still-camera frame should be more manageable. Brett's test chart is handy if you can print it to be exactly 24mm wide by 18mm high.

Otherwise, hand-draw a target to this size and place it where your groundglass image is and see how much of your image is covered by this target. Ideally, for faithful replication of the 4:3 movie frame, it should be tight in your camcorder frame edge-to-edge. <<<--

Im not really sure what thise target is for, is it a reference point on how far to zoom in before its too far?

<<<--How thick is the vinyl? If you take your lens outside and project an image onto the registration sticker on your car windscreen, this should give you some idea of what to expect from the vinyl. If the vinyl isn't the same then it likely won't work. -->>>
The vinyl i got was the thinnest vinyl they had at the sign shop, its a brand called avery and on the swatch chart it was the lightest stuff they had.

I just went into the shop and asked for etched vinyl but if anyone has a product code or anyhting might be abit more help.

Thanks again guys for the info.

Cheers,
Ben

[Bob Hart]

Ben.

Please forgive my tendency to get caught up in my own jargon.

I'm using the same lens.

By "target" and "4:3 movie frame, I mean a rectangle area of 24mm across by 18mm high. This is as much of the image projected on the groundglass by your Nikkor lens to use if you want to faithfully represent the 35mm motion picture image.

The f1.8 is actually going to give you a larger image almost 37mm across but you only want the centre out of that because the image will be getting darker towards the outer edges.

To get the correct image, the mount face or flange of your Nikkor lens has to be 46.5mm away from the groundglass surface of your disk.

In most Agus35 builds, the groundglass surface is facing the camcorder so you will have to measure from the back of the disk to the mount face of the Nikkor lens to set rough back-focus.

I have sent the groundglass plastic disk, some unpublished "how-to" info as .pdf files on another disk and some demo footage as DVD-Video written onto a DVD+R disk which should play in more recent DVD players.

If you are designing your own AGUS35 entirely, before you build, you need to mount your camcorder on a tripod or a piece of wood, attach your close-up lens, then find the closest place in front of the camera, where the 24mm x 18mm rectangle drawn on a piece of card is framed tight in the viewfinder and is sharp.

It is better if the zoom does not have to be fully extended to get the frame and sharp focus because most camcorder zooms lose light performance towards the narrow end of their range. With the PD150 and a 7+ close-up lens and the 24mm x 18mm rectangle about 5 inches away, about 3/4 of the zoom range has to be used to get it framed tightly in the viewfinder view.

If you are building a prism version, a +7 close-up lens is as powerful as you can use. Any higher power and the groundglass has to be too close to the front of camcorder for the prisms to fit in.

[Bob Hart]

The AGUS35, the Australia Plumbers Version prism version got a work-out in a sort-of production situation tonight and was found wanting.

For wide-angle master shots, there was a problem getting a sharp image with the 28mm lens. A separate monitor is desirable.

This version mounts directly to the camcorder lens hood bayonet mount. Hand held it is fine but when the whole combination is put on a robust tripod, in ths case a wooden legged fluid head Miller, then vibration from the glass disk becomes an issue.

Because the tripod doesn't move, so the camcorder structure itself deforms, not much but enough for an intermittant one TV line vertical jitter.

The prism version is heavier than the non-erecting version and the center of forces created by the disk sits about 78mm outboard of the front of the camcorder compared to the 18mm of the non-erecting version.

This is enough to set it bouncing minutely from the midpoint of the camcorder casing forward which affects the stability of the image falling on the CCDs and probably doesn't serve the expected life of the camcorder all that well.

So it would seem that baseplate support direct to the AGUS35 cannot be avoided so this endorses the fully supported designs which have featured here.

[James Hurd]

Hey Bob,
I think I found some prisms. They're 38x54x38. Do you think they would work the way you have yours arranged?

http://www.edmundoptics.com/onlinecatalog/displayproduct.cfm?productID=2388&search=1

Thanks for any help!

[Bob Hart]

James.

I haven't looked at the edmund optics site as my browser keeps chasing its own tail when downloading so I can only go by your description.

The prisms as I understand them should work providing the common thickness across all faces is no less than the 38mm of the short sides.

The image path might be a little tighter and require more precise alignment of the prisms in their mount and the optical centers of the camcorder and the front lens.

For the standard movie 4:3 frame, this will only leave you with a 1.5mm margin either side of the groundglass frame into the front face of the rear prism if you mount it close to the groundglass as I have.

I have sent an email to Chris Hurd with a request to post a montage of the test in a production environment. This file will have the name "grabmon4.jpg" when it gets to www.dvinfo.net/media/hart.

The info on this test is :-

Lenses, Sigma f1.8 28mm, Nikon f 1.8 50mm.

Relay path, ao5 glass oharadisk @ 1500 rpm, 2 x 90degree prism 90degree opposed, Century Optics 7+ achromatic diopter, Sony DSR PD150P camcorder.

Lens settings wide-open.

Camcorder settings. Manual white-balance. 0db video gain, shutter 1/50 second, f setting at f3.4.

Lighting.

Environmental. 18watt overhead installed flouro in diffuser cover.

Added in key/fill configuration. 1 x 240watts equivalent edison screw flouro in old Photoflood portrait reflector lamp, 1 x 100watts equivalent edison screw flouro in old Photoflood portait reflector lamp.

I could not get the proper lamps for the Photofloods which were the only lamps I had which would not burn out actor's highlights. I would like to have had a small backlight which was definitely needed in the two-shot.

I dropped the Agus on site which jolted the backfocus out of alignment on the left side. This didn't help. Fortunately, the glass disk did not break.

[James Hurd]

Thanks Bob.

Does anyone no a good achromat for the DVX100A? The one I'm using now (Hoya 55mm with a 72mm step-up) is destorting the edges and causing the blue halos.

Thanks Guys!

[Bob Hart]

James.

I understand Century Optics make an achromatic 7+ for the DVX-100. I think Panasonic may have one of their own too but I am not sure. They do have their own 16:9.

I don't know about the 72mm version but the 58mm version for the PD150 works fine. I think you will see a vast improvement over the Hoya 55mm which is okay for its intended purposes but being looked through by a larger diameter lens is not one of them.

[Bob Hart]

"grabmon4.jpg" is now posted on www.dvinfo.net/media/hart. It is about 7/8ths down the list.

I didn't realise at the time but the wide shot was a failed take - I had forgotten to turn the disk motor on.

The images look a bit better now after some colour and brightness correction. Does anyone know how to get rid of vertical bars which appear in the image when it is severely adjusted as this had to be.

For credits where credit is due, the project was an extract from a local screenplay "Roo-dog and the pull of the moon".

Actors are :-

Paul Booth.
Alison Roberts.

Should any production entity seek to cast in Western Australia they can contact Annie Murtagh-Monks and Associates or Perth Actors' Collective, who hosted and assisted the workshop.