Lens Adapter for HD100 finished.

[Bob Hart]

You are definitely on a winner with your arrangement. With a 120 format sized image, you should be in front for resolution, even with a Maxell blank.

I second the achromatic dioptre suggestion. Your images are well sharp enough. The colours will probably look even better in a highly contrasted, finely textured image if you ca get rid of the CA.

Very well done.

[Ted Ramasola]

Thanks for the kind words, youre right Bob, but from where i am located its so hard to get hold of an achromat, I need a +10. Right now its a stack of 3 close ups which produce those CA at the sides. I know they are there, and I know that the close ups are causing it, but i have to shoot with what i have till i get a good close up cause my projects all piled up.
I tried removing one from the stack and the CA was negligible but i can see a vignette so i compromised.

Ted

[Ben Winter]

Is the grain from the focusing screen or gain on the camera? It'd be nice to see a better achromat on this thing, the edge-to-edge sharpness is painful.

[Ted Ramasola]

Ben, the grain is from the spinning GG. Yeah, I wish I had a good achromat too. The CA is agravated in areas of contrast near the sides.

[Ted Ramasola]

Ben,
Heres a frame from a well lit shot. Theres minimal grain in this shot.
I think grain is aggravated by lower exposures. But i never use gain with the adapter to minimize the grain.

[Jon Wolding]

Ted, cinevate.com is selling a nice +10 achromat. Talk to Dennis Wood about it.

That guitarist shot is so damn sharp! Get rid of the CA and figure out the bokeh, and you've got most of the adapter market beat!

[Bob Hart]

Ted.

The Maxell disk you use now, - is it opaque all the way through the thickness or opaque just on one surface only? If it is opaque all the way through, you might discover a little more sharpness and contrast with a disk which has a surface groundglass finish on one face.

To get the full wide image area without corner falloff in brightness with a surface finished groundglass, the grade of the surface finish might need to be a little coarser than for the 5 micron finish I use on my disks and other builders also use on 24mm wide movie frame versions.

Medium format is a frontier I haven't visited so don't pay much heed to my comments.

I believe Wayne Kinney found his disks for his SGPro at Knight Optical in the UK. Maybe he will add some info here on reading this.

[Ted Ramasola]

Bob, they are opaque only on one side. and its the one i faced towards the mid format lens.

[Ted Ramasola]

Thanks Jon, i'll check on it.

[Ted Ramasola]

I took Jons advice an got in touch of Dennis Wood of Cinevate, he shipped me his achromat and it solved a lot of my adapter's problems!

Im testing it today and below is a chart using 80mm mamiya 645.
f1.9 on the mamiya f2 on the hd100. image size on gg is still 45mm across and 25mm high.

[Jon Wolding]

If you're using a Cinevate achromat, then something else must be causing the chromatic aberration (jpeg compression? HDV? Camera settings?). I've never had that degree of CA in the center of the frame.

[Bob Hart]

Ted


To my eye, the fringing on the lines in centre looks a bit like a codec compression artifact although it may be the beginnings of a chomatic abberation.


Optically there seem to be three issues.


There is a small barrel distortion with sharpness falloff into the corners.

There is chromatic abberation towards the horizontal edge thirds of the image.

There is a third chroma issue which is not symmetrical across the image.


The centre fringing is not something I would get heartbroken over.


The barrel distortion is something I would expect to see. Medium format relay is uncharted territory.

Dennis's acromat will have been optimised for a smaller area image off the groundglass and your arrangement may be reaching into areas of the lens we would otherwise not see. I will comment furthur down in this message.

The chromatic abberation in the edge thirds is probably linked to the barrel distortion.


The third chroma issue is unusual. It may be simply mixed colour temps of separate lighting sources onto the chart, colour variance across the chart itself if printed with a colour printer on photo paper ( I had this issue myself )


If the chart and its lighting are okay, then it may be created by

axial offset of the optical centres of the lens on front of the adaptor and the camcorder optical centre,

angular misalignment of the lens on front of the adaptor.


Both of these individually or in combination would introduce an assymetrical defect.


Another possibility is offset of the achromat dioptre centre-axis relative to the camera optical centre axis.

Unless you have mounted the achromat separately to the relay lens and not by threaded step rings, I doubt this is an issue as the dioptre and camera optical centres are created by some precise machining and die making.


My first experiment would be to make a spare groundglass and install it, maybe lay a clear panel over the existing groundglass if you don't want to disturb it, view the projected image by eye and mark clearly on the spare groundglass, the outer edge of the image circle and the precise centre of the image circle.

Across this area I would also draw a checkerboard of crossed straight lines.

I would then fit the adaptor up to the camera without any lens fitted on front of the adaptor, light the groundglass through the lens hole and study the camcorder image of the lines and centre mark.

You will need to lock the groundglass disk from moving with a piece of adhesive tape on the clear face of the disk which you can clean afterwards without damage.

This will show if the barrel distortion and asymmetric colour shift is a charactistic of the front lens, and if it is correctly centred.

You could do this with the front lens mounted and imaging a plain white background but you won't be able to identify any added CA caused separately by the lens if it is attached.


My second experiment would be to move the adaptor furthur forward, almost to the distant or infinity end of the focus limit of the relay lens/achromat. This will result in a slightly larger area visible off the groundglass but might reduce the distortion.

If the defects are projected by the front lens onto the groundglass itself, then the defects will become more apparent.


I am unclear if you are using the JVC standard lens with the achromat on front for your relay.

If the Mamiya f2 you refer to is your actual relay lens used like Quyen's Minolta lens and has replaced the JVC standard lens then much of my following comment is disqualified.

This lens if used for relay, could cause you some edge falloff, chroma problems and distortion as it is not optimised for 3 x CCD imaging. Getting correct backfocus of this lens may also be a problem task if you have made a custom mount or spacer-adaptor ring to marry the lens to the JVC mount.

My third experiment would be to attempt to place Dennis's achromat closer to front of the relay lens if possible.

You might need to dismantle the optical element of Dennis's achromat from inside its ring to do this.

If you are using the JVC standard lens then you will discover that the lens focus movement changes the distance between an achromat mounted in the thread of the fixed lens hood mount and the front element of the standard lens so this experiment is worth trying.

I found it necessary to mount my own achromat directly to the JVC lens barrel itself to avoid this problem.

Dismounting Dennis's optical element from its ring could be risky so I don't encourage it.

If the barrel distortion becames worse the achromat may need to be placed furthur forward with a thin spacer washer on the threaded shoulder and maybe a spacer in the ring around the element in the acromat to take the element furthur forward.

Too far forward and the image may develop a pincushion distortion. all of this interacts in combination with variations in lens focus and variations in the distance between the groundglass and the front of the camcorder relay lens.

If you can successfully fine tune in this area, you may also pick up a bit more sharpness.

In the JVC menus, you may find a colour correction function which deals with broad area colour variation across the image caused by individual lens characteristics.

Many broadcast cameras have this adjustment and the setting is saveable as a scene file, personal preset or suchlike. The JVC KY-F50 camera head has this feature so your JVC may have it as well but I don't know what it will be called.

Your image as it is I could quite happily live with. I would tend to compose my shots to avoid areas of high contrast fine texture in the horizontal edge thirds of the image.

LATE FOOTNOTE:

I observe on yout previously posted image of the man and guitar, a darkened area on the right side of the image as viewed which suggests your front lens may be offcentre also to right of camcorder optical cetre axis if you are not flipping the image uprigt into the camera. I had not previously observed this.

You may need to consider using a slightly longer focal length prime lens for your relay lens and taking a smaller area off the groundglass to get inside the otspot and also the soft areas if they are relay lens related.

[Ted Ramasola]

Bob,

thanks for taking the time with your assesment. I continued with my tests and found out that the sweetspot for this setup, escpecially using the brevis achromat is much smaller than 45mm width of the image. The sweetspot without corner distortiona and cA is actually close to a 35mm. I am basing my test on the mamiya 45mm lens since wide lenses tend to exhibit more of the vignette and distortion on the corners.

i'l try to see how much i can get away with as far as edge distortion, i'll try to get a balance between clean edges and a wider gg image. Im hoping to achieve lesser grain by a larger image.

With this observation, does it mean that 35mm adapters are actually capturing less than 35mm image width on the gg? Or is it exactly that?

Ted

[Bob Hart]

Ted

You pretty much have it right in regard 35mm adaptors.

An image width of 22mm to 24mm is more typical and closer to the 35mm motion picture film image width. Some do wider. So in proportional terms, your arrangement seems to be performing much the same but yielding a larger therefore potentially sharper image.

48mm will be outside the medium format sweet spot of 36mm x 36mm for groundglass relay.

You can cheat a little and frame a bit of defective image into the lost area around the TV "safe area".

[Ted Ramasola]

Bob,

I did some realignment and found a lot of improvemnt in sharpness and vignetting. The alignment proces is on a per millimeter basis and is so tedious, this might never get perfect for me, i wanna stop at a point where all my lenses can be usable and without to much vignette and corner softening, especialy with the wider lenses. I moved back my jvc farther from the adapter, and still using the stock lens at a setting of around 20 and full macro.
The brevis achromat is still attached with step down ring.
I posted some grabs with subjects near the corners and top and lower edges to gauge the corner and edge areas areas.

First two images from mamiya 645 45mm f3.5 and the diagonal guitars by an 80mm f1.9.

Ted