DIY adapters for RB645 and 35mm lenses

[Ted Ramasola]

Im posting this to share my experience with other DIYers on the pros and cons of each setup.

I first made the mid format adapter and with the help of people in this forum was actually able to have it work. I have used it for corporate work and for a TV spot. The image on the GG is big and manageable to "hit" to avoid vignette/falloff.

image on GG is 36mm x 24mm.

After several real world uses i found the limited availability and variety of lenses as its major con.

I then proceeded to fabricate a 35mm adapter for the nikon mount.
Lenses are easily available. I used a different and finer GG this time.

The adapter size is much smaller because of the small diameter of the GG.

I can use our photo department's wide array of 35mm lenses especially our old manual Tokinas, Sigmas, Tamrons and Vivitars.

The available accesories allowed me to use a nikon m2 extension tube as a lens mount and a gutted polarizing filter as a locking mount that engages the achromat at the rear.

Im "cheating' on image size on gg at 30.5mm x 18mm to maximize image size (concentrating on optimum image that falls within the TV safe area).
below is a side by side pic of the two adapters for comparison. and grabs of a bokeh test using the 35mm adapter 50mm nikon lens at f1.8 and a macro shot using 105mm f2.8. a wide shot using a sigma 20mm f2.8.


Ted

[Bob Hart]

Ted.


What power close-up lens (achromatic dioptre) are you using?

I found that the 7+ was best for a compact handheld portable adaptor but the best results I get are from using a 4+ and sitting furthur back off the groundglass.

The camcorder zoom for the 4+ has to be all the way tele to 54mm and this loses a little more light than the 7+ arrangement which requires the zoom to be no more than 45mm.

The "Tears in the Rain" behind-the-scenes clip on YouTube was shot with the longer setback 4+ arrangement. It is a backbreaking swine to handhold though.

Your edge softness I suspect may relate to the relay path - close-up lens and camcorder lens combination. It may be helpful to position the close-up lens closer to the front element of the camcorder lens if you can achieve this.

I have used my "Sony" evolved adaptor on a JVC HD100 via its standard Fujinon lens.

I actually fastened the close-up lens on front of the standard Fujinon focus barrel itself rather than use the 82mm filter mount as this remains in a fixed position relative to the lens itself which moves forward or backwards with focus movements.

Mounting the close-up lens on the filter mount does not allow the close-up lens to be close enough to the Fujinon's front element to work properly.

The 7+ dioptre I used was a 58mm filter diameter size threaded mount inside a 72mm sleeve. This was crudely attached to the front of the Fujinon's moving focus barrel.

I did not see any edge softness with it but I did not test extensively as the JVC HD100 owner has the real thing from P+S and does not need to use my adaptor.

If you can get your dioptre closer to the Fujinon, I think your edge softness may go away.

You could probably afford to give a little of the image area away, maybe come down to about 28mm wide which is what the 4+ and full zoom does for me with the Z1 and still keep an adequate area for the TV safe area.

This might require you to go to a finer grind on your disk. I note a few arcs on the pliers and spanner image which suggest the grade is yet a little coarse.

Otherwise nice dense colours and contrast.

[Ted Ramasola]

Bob

thank you again for the great assistance you provide.

the achromat is from dennis wood, i believe he told me its a 10+, it has a 72mm thread.

The suggestion you gave me provided significant improvement on the soft edges. Im attaching two charts.
after removing the achromat from the stepdown ring i attached it directly to the adapter's rear via a series of empty filter rings. setting it farther from the gg at 78mm or a notch more than 3in.

sharpness improved at the sides even beyond the TV safe areas while still giving me 29mm image width.

the moving of the achromat deep into the focus barrel also diminished significantly falloff.

Many thanks for your critical inputs Bob.

Ted

[Bob Hart]

Ted.

You have a significant improvement and maybe as good as it will get with the 10+ which might be a bit strong for the Funjinon unless you can get it a little closer inside the front of the Fujinon lens. The 10+ is a necessity otherwise the non-flip adaptor-camera combination becomes unworkably long.

With the Century 7+ I actually had the 72mm threads bedded into one of the ridges of the anti-reflective plastic area in the rotating focus barrel.

If you buy in Dennis Woods' flip enclosure you may be able to come down in power of the achromatic dioptre which may help, inf fact you may have to if the path becomes too long. My home-made is a flip using prisms which is why I had to stay in the 4+ to 7+ range which is roughly 8 inches and 5 inches setback, physically shorter with the prism foldback of course.

Dennis may also have a condenser in there between the achromat and the groundglass. You would need to ask him about that. P+S Technik also seem to have come down in favour of lower power and longer setback for their direct relay versions. I have not seen their setup for lens-in-camera styles.

The Pro35 has an optic behind the groundglass. I don't know if it is there in the Mini35, but the setup is similar.

Still, within the TV safe area, what you have is useful and careful composition and framing should taske care of the corners most times.

[Ted Ramasola]

Bob,

My first point of reference in my setup was the chart you posted indicating diopter power --distance from gg---camcorder zoom. You indicated 3in distance from GG. After applying the setback of the diopter by removing it from the step down ring and placing it inside snuggly against the fujinon focus barrel it changed the distance from gg significantly.

The previous chart i posted still has soft sides and corners which is enhanced when using wide lenses. By shortening the GG to diopter rim distance to 70mm or 2 3/4 inch i was able to produce sharper edges and corners.
The 70mm distance is based from the front(matte) side of the gg to the rim of the 10+ achromat.
Otherwise,depending on gg thickness, in my case its 1mm, its 69mm distance from the back surface to diopter rim.

Image captured on GG increased slightly to 32mm x 18.5 mm

Im attaching grabs and a chart, this time i use a white backed chart to better aid me in avoiding falloff.

Ted

[Bob Hart]

Like all R&D, things get better in smaller increments beyond the first giant leap.

You still have some soft corners but this may be related to the larger 32mm image width.

I have about run out of ideas. The only last thing I can suggest is to perhaps to to set the "groundglass to front of lens" distance so that the Fujinon lens is as close to infinity focus as you can get it but still keep just a little for final focus trim with the lens barrel focus adjustment.

Don't use macro-focus or the backfocus on the Fujinon lens for the final trim. Those functions both use the same retrofocus element.

It might be worthwhile emailing Dennis for his "groundglass to front of lens" distance. The elusive last few percentage points of image quality may be related to adding a condenser element, if his optical path was engineered and not a mix and match of available trade parts.

Beyond that, you may have to zoom in that little bit closer and shoot a smaller groundlgass image. Zooming furthur in brings the front element of the internal travelling group that bit closer to the achromatic dioptre which might help with the corner softness as well. At 28mm width, you will still be 2mm either side wider than the standard academy frame.

I never tested my arrangement to a 32mm width as my prism path comes in on the sides before then. I have hobbled myself with cost cutting by using identical prisms of only just enough image width to clear 28mm.

In an ideal world, I would spend on two prisms of different sizes or a a prism-mirror hybrid with larger mirror pair a la Mini35. Unfortunately, the ideal world and my financial means are not in harmony.

Out of interest, in your kit of stills lenses, do you have a 55mm f3.5 Micro-Nikkor. The Micro-Nikkor can focus to 8.5 inches. But as a non flip direct relay, it would be a bazooka with no real advantage over the Fujinon/achromat for overall length except for possibly sharper image and better edge to edge sharpness. You would lose some light performance as well.

On another thread for Letus, a user has fitted another Nikon lens with macro to his JVC via a lens adaptor to use as a direct relay like Quyen does with the Minolta prime and reports good results.

His lens may be a zoom which might lose a bit of resolution and some light but this might be offset by the smaller 1/3" CCD shooting through the centre. Hopefully he will publish more details and some grabs.

That may be an alternative option for you to examine.

[Ted Ramasola]

Bob,

I did some tests with the fujinon settings and found out how much the iris settings on the 16x can affect the image from the adapter.

Speaking of the letus extreme, Luis Caffesse over at another forum
posted his tests with a res chart. Here are the links to his test with and without the letus using the HVX and zeiss 50mm t1.4.

http://www.pitchproductions.com/pics/HVX-RES.tif

http://www.pitchproductions.com/pics/LETUS-RES.tif

I then took out a similar chart and framed my tests same as his and found some interesting results.

My DIY adapter was able to come close to the letus especially in the corners.
The soft falloff on the letus is wider.

Using the same opening at 1.4 on the 50mm i had I achieved a sharper corner image(take note of the '16:9' text on the chart corners) when i stopped down the jvc to f11.

at f2.8 on the nikon and 5.6 on the jvc is similar to the letus+hvx chart.

Pushing the fujinon for comparison, at f5.6 nikon and f1.4 on the jvc the image fell apart.

In my set up, macro is fully engaged, and focus is at infinity. I use the zoom ring to achieve focus on the GG.

Im attaching the charts below to campare with the charts on the links.

Regarding arcs on the image. Heres my test results.
Lenses 20-40mm f2.8, 50mm f1.4, 85mm f2 dont exhibit the arcs up f4.
Lenses 105mm f2.8, 200mm and above exhibit the arcs.

Ted

[Bob Hart]

Well what is there to say. There's no disputing the test charts.

That's the popular notion of f5.6 or thereabouts camcorder iris being the sweet spot in groundglass relay, shot up the butt like a garden thrush, leastways in this particular lens/achromat combination.

The edges seem square and correct so the spacing between the Fujinon and achromat seems about right. What zoom have you got the Fujinon set to? I think it is a rectilinear lens which means there will be diagonal stretch in the corners to keep straight lines straight anywhere in the frame. If there is any softness in the corners, a rectilinear lens will amplify it.

I could not get the .tif files to come down the pipe, probably a bandwidth thing this end.

[Ted Ramasola]

Bob,

Zoom setting on the Fujinon is slightly above the 20 mark. Say the white line points at the rear end of the zero.

A funny thing i discovered about the nature of lenses is that all our long lenses from 105mm up to the 300mm seem to show the arcs. Is there something with the optics in them that makes this happen? Or is the narrow light path the culprit?

But anyhow, knowing the parameters i can work with through these tests will allow me to rest easy on an actual shoot.

Ted

[Bob Hart]

20. That's wide compared to my setup.

I did observe the Fujinon seemed to require a slghtly closer GG position than the Sony Z1 for the same zoom position, in my case closer to 50.

This may be due to different specification for the "on-chip" lenses CCD cameras apparently have between different manufacturers. It is tech info I do not know at all.

It is a contrary business. The wider lenses provoked the groundglass artifacts on my setup. If there is any oil gets on a Mini35 GG, the wide lenses provoke that. Yet on yours it is the long lenses.

Your groundglass textured surface. Is it on front face of the disk closest to the SLR lens or the rear face of the disk? Mine is on the rear face.

[Ted Ramasola]

The textured Surface of the GG faced the SLR lens. You think i'll have better results if i invert it? Conventional knowledge do dictate the matte surface faces the light source.

[Bob Hart]

I think the jury is still pretty much out on which way is best for the groundglass. My personal preference is to have the groundglass texture facing the camera.

Film emulsions of course face the lens.

I have no scientific reasons to back it up other than to hope multiple generations of diffusion and reflection within the groundglass itself are less evident as a halation artifact.

My half-baked theory is that the groundglass texture itself might hide from the camera's clear optical view, half a generation of any clearly defined reflection within the groundglass back onto the texture.

As to your arcs. They are quite sharply defined. My guess is that individual large spots or large pits might be making them.

It might be worth experimenting with pinpoint spots of water ink, diluted down to about a mid grey with a bit of white PVC paint or some typewriter correcting fluid darkened to a neutral grey.

Dip into it with a tooth pick and spot just into the pit itself without going over onto the adjacent area of the glass.

If you are using a plastic disk, this might make trouble for you as plastic is harder to clean off without damaging the texture so you might leave best be and not try to change things.

[Ted Ramasola]

Bob,

I opened my adapter cleaned the GG, experimented inverting it.

I found some big specs on it and blew them out but small arcs remain at small apertures using long lenses.

The inverted GG produced same results as far as resolution and gg induced grain goes. It might introduce different flare characteristics when light hits at an angle, didn't test that and inverted it back.

In relation to the corners improving when the relay lens is stepped down, i'm attaching a "chart", of the tests results on corners after i recalibrated my focus on the gg. I actually gained a little bit on f5.6 from my previous test, but f8 and smaller is best.

I hope this will also help those who are new at using adapters since i think the principle also applies to the "branded" adapters.

Ted

[Ted Ramasola]

Bob,

Quoting your earlier post;
"Like all R&D, things get better in smaller increments beyond the first giant leap."

I made another small increment and this seems to confirm what you said that I should zoom in more on the gg.

I hesitated doing so since I thought I got the optimum balance in edge to edge sharpness and light falloff. However I discovered darker corners outside the TV safe area which flows into the viewable area when images are subjected to extreme contrast adjustments in post and when using very wide lenses.

Im posting a group of images below and the settings on the fujinon for you to see if i'm nearing the settings that you expect in this set up.

I also suggest to DIY'ers out there to do a "plain wall" test to manage falloff.

Your help and comments Bob really show your experience and altruism in these forums.

Best wishes for the new year.

Ted

[Bob Hart]

Ted.

Looking good. You may have also recovered a little more corner sharpness along the way.

You should be able to do chroma key work with it now.

Does the option for small adjustments with the zoom for composition or effect remain? This is one feature I like about relay via the camcorder's own lens, which compensates for the awkardness of the longer arrangement.

All the best for the coming New Year.