Making progress on DoF adapter at DVinfo.net

Paul Nixon · May 16th, 2008, 08:14 PM

I installed a new achromat - an L4673 from surplusshed.com. Seems to have helped a great deal - latest images:

http://i120.photobucket.com/albums/o...nstalled_1.jpg
http://i120.photobucket.com/albums/o...nstalled_2.jpg
http://i120.photobucket.com/albums/o...nstalled_3.jpg

Focus screen (still using salvaged screen for testing) is right-way-around in these shots. Achromat is mounted just ahead of the camcorder lens (fitted into bayonet adapter I "adapted" to mate the EOS tubes with the 37mm-52mm step up ring).

Lighting appears to be pretty even. Focus appears to be pretty even - probably isn't due to me placing the camera and shooting from the hip.

Wayne Kinney · May 17th, 2008, 03:17 PM

This looks very promising. The images seem clean, sharp and free of CA (bar GG artifacts). Keep at it!

Bob Hart · May 17th, 2008, 09:12 PM

Looks like you are well on track.

The third image suggests your centres are out a little :-

Assuming a 24mm x 18mm image and use of a lens of f1.8 aperture, the SLR lens centre axis relative to camcorder lens centre axis appears to be about 1.5mm left and about 3.5mm low.

If you are scanning a larger image area which I think you may be, those offsets will be greater. Getting the centres coincident will retrieve some darker areas mainly on apparent left of your image. This will bring in some dark in the opposite edges or corners.

If you can zoom in a little more you might consider doing this to get inside the dark corners if you don't lose too much resolution by doing this.

However, your first and second images, suggest my assumptions may be wrong or you have used different lens or lenses across the shots. There remains a slight apparent corner brightness falloff which I can and do live with on my own appliance. It becomes an issue if you ramp up contrast or colour in post.

The apparent centres of brightness appear to vary across the shots, which suggest your device may have some flex in its structure or its mounting to the camcorder between the shots can be a little off-axis. This is a relatively common occurance which even flex in the camcorder's own casework can contribute to.

Given that you are using machined tube and step-up rings, flex at the camcorder to adaptor joint is my assumption as to where centrelines are angularly deviating as the centre axes themselves should be automatically correct unless some very bad machining has taken place. The common solution for this is additional bridging between firmer parts of the structures of the camcorder and adaptor, usually by means of tripod base bridgeplate and rods supports to the adaptor and heavier lenses when used.

Camcorder imagers also often have their centres offset relative to the optical centre axes of the lenses which feed to them. Zooming in to the groundglass is pretty much the only solution to this problem if it brings in a dark area into an edge or two corners.

My comments are largely speculative and I am likely reprising research you have already done, so don't take my comments too seriously.

Paul Nixon · May 19th, 2008, 09:33 AM

Many thanks Bob,

This whole project has involved a great deal of head scratching, and one oddity I noticed early on was that the focus aid of the 35mm SLR lens was not centered in the camcorder view finder (I have the "Display" guide turned on - a grid of 9 squares appears on my camcorder viewfinder). At first I thought something was out of alignment, so I rotated the DoF assembly - the focus aid moved but not quite as I had expected (it did not reverse position).

And there is flex in the camcorder body. I notice the entire assembly has some play right in the focus ring area of the camcorder. I have the assembly mounted on a Manfrotto tele-lens support, but while this holds things steady during use, there's no guarantee of perfect alignment.

I will continue working on it.

Mike Dulay · May 19th, 2008, 05:00 PM

The HV20 sensor is offset. Mine is to the right so I can see the left edge of the focus screen more than the right. Also the OIS goes into action when the camera is moved so you might see the image shift slightly.

Your pics are looking good though. How did you mount the focus screen on your adapter?

Paul Nixon · May 22nd, 2008, 01:26 PM

Quote:

Originally Posted by Mike Dulay

The HV20 sensor is offset. Mine is to the right so I can see the left edge of the focus screen more than the right. Also the OIS goes into action when the camera is moved so you might see the image shift slightly.

Your pics are looking good though. How did you mount the focus screen on your adapter?

Thanks Mike,

Thus far I am using a 49mm filter (sans glass) attached to a piece of drain pipe that is then inserted into the EOS tube. To keep the focus screen in place I'm actually using the retaining ring that originally held the glass part of the filter in place. The old screen had a frame that made this really handy. The new focus screens didn't come with a frame (it was available for about $14) so it's a bit different, but for what I am doing at the moment it's okay.

I'm still getting beat up by off-axis blur (???) - I can get the center razor sharp but the focus goes out away from center. In my experiments I've noticed the effect seems to be LESS if I zoom in LESS. That seems counter to what I'd expect, but that's what I'm seeing.

Funny thing - I took my rig out in the field today for some "real life" testing. I was using it hand-held, and since I have the Manfrotto lens support I thought I'd try shooting upside down as I've seen others do. The HC1 flips the image somehow when the camera is inverted. Doh!

I've got two vibrator motors that I can use to convert my rig. I've got to get some small rods still, but I'm fairly confident I can pull it together. We'll see...

Mike Dulay · May 24th, 2008, 07:35 AM

Hi Paul,

Flipping the camera won't flip the image you see on the LCD, as you've noted, but at least you won't have to flip it in post. Over time I learned how to frame inverted.

As for the softness, you might want to check how much of it is from your camcorder to start with. I leave the zoom and focus locked, take out the adapter and point it at something with good contrast (like pillars or blinds). Then I bring the exposure down. After extracting the image/video you can play around with color curves and bring out the circle. Compare this against one you shoot at full wide. There's probably a better or more correct way to do this, but that's how I discovered it. My HV20 has a sweet spot just a tad before halfway where the entire frame is sharp. Maybe its similar for the HC1.

Rich Hibner · May 24th, 2008, 01:09 PM

flipping the camera will fllip the image on the lcd. but maybe it depends on the camera.

May 16th, 2008, 08:14 PM	#1
Paul Nixon Regular Crew Join Date: Feb 2007 Location: Mesa, AZ Posts: 141	Making progress on DoF adapter I installed a new achromat - an L4673 from surplusshed.com. Seems to have helped a great deal - latest images: http://i120.photobucket.com/albums/o...nstalled_1.jpg http://i120.photobucket.com/albums/o...nstalled_2.jpg http://i120.photobucket.com/albums/o...nstalled_3.jpg Focus screen (still using salvaged screen for testing) is right-way-around in these shots. Achromat is mounted just ahead of the camcorder lens (fitted into bayonet adapter I "adapted" to mate the EOS tubes with the 37mm-52mm step up ring). Lighting appears to be pretty even. Focus appears to be pretty even - probably isn't due to me placing the camera and shooting from the hip.

May 17th, 2008, 03:17 PM	#2
Wayne Kinney Major Player Join Date: Jun 2005 Location: Brighton, East Sussex, UK Posts: 938	This looks very promising. The images seem clean, sharp and free of CA (bar GG artifacts). Keep at it! __________________ Thanks, Wayne.

May 17th, 2008, 09:12 PM	#3
Bob Hart Inner Circle Join Date: Dec 2003 Location: PERTH. W.A. AUSTRALIA. Posts: 4,477	Looks like you are well on track. The third image suggests your centres are out a little :- Assuming a 24mm x 18mm image and use of a lens of f1.8 aperture, the SLR lens centre axis relative to camcorder lens centre axis appears to be about 1.5mm left and about 3.5mm low. If you are scanning a larger image area which I think you may be, those offsets will be greater. Getting the centres coincident will retrieve some darker areas mainly on apparent left of your image. This will bring in some dark in the opposite edges or corners. If you can zoom in a little more you might consider doing this to get inside the dark corners if you don't lose too much resolution by doing this. However, your first and second images, suggest my assumptions may be wrong or you have used different lens or lenses across the shots. There remains a slight apparent corner brightness falloff which I can and do live with on my own appliance. It becomes an issue if you ramp up contrast or colour in post. The apparent centres of brightness appear to vary across the shots, which suggest your device may have some flex in its structure or its mounting to the camcorder between the shots can be a little off-axis. This is a relatively common occurance which even flex in the camcorder's own casework can contribute to. Given that you are using machined tube and step-up rings, flex at the camcorder to adaptor joint is my assumption as to where centrelines are angularly deviating as the centre axes themselves should be automatically correct unless some very bad machining has taken place. The common solution for this is additional bridging between firmer parts of the structures of the camcorder and adaptor, usually by means of tripod base bridgeplate and rods supports to the adaptor and heavier lenses when used. Camcorder imagers also often have their centres offset relative to the optical centre axes of the lenses which feed to them. Zooming in to the groundglass is pretty much the only solution to this problem if it brings in a dark area into an edge or two corners. My comments are largely speculative and I am likely reprising research you have already done, so don't take my comments too seriously. Last edited by Bob Hart; May 17th, 2008 at 09:57 PM. Reason: errors

May 19th, 2008, 05:00 PM	#5
Mike Dulay Major Player Join Date: Mar 2007 Location: Somerville, NJ Posts: 304	The HV20 sensor is offset. Mine is to the right so I can see the left edge of the focus screen more than the right. Also the OIS goes into action when the camera is moved so you might see the image shift slightly. Your pics are looking good though. How did you mount the focus screen on your adapter? __________________ DIY, 35mm, HV20: http://www.primitivebuteffective.net

May 24th, 2008, 07:35 AM	#7
Mike Dulay Major Player Join Date: Mar 2007 Location: Somerville, NJ Posts: 304	Hi Paul, Flipping the camera won't flip the image you see on the LCD, as you've noted, but at least you won't have to flip it in post. Over time I learned how to frame inverted. As for the softness, you might want to check how much of it is from your camcorder to start with. I leave the zoom and focus locked, take out the adapter and point it at something with good contrast (like pillars or blinds). Then I bring the exposure down. After extracting the image/video you can play around with color curves and bring out the circle. Compare this against one you shoot at full wide. There's probably a better or more correct way to do this, but that's how I discovered it. My HV20 has a sweet spot just a tad before halfway where the entire frame is sharp. Maybe its similar for the HC1. __________________ DIY, 35mm, HV20: http://www.primitivebuteffective.net

May 19th, 2008, 09:33 AM	#4
Paul Nixon Regular Crew Join Date: Feb 2007 Location: Mesa, AZ Posts: 141	Many thanks Bob, This whole project has involved a great deal of head scratching, and one oddity I noticed early on was that the focus aid of the 35mm SLR lens was not centered in the camcorder view finder (I have the "Display" guide turned on - a grid of 9 squares appears on my camcorder viewfinder). At first I thought something was out of alignment, so I rotated the DoF assembly - the focus aid moved but not quite as I had expected (it did not reverse position). And there is flex in the camcorder body. I notice the entire assembly has some play right in the focus ring area of the camcorder. I have the assembly mounted on a Manfrotto tele-lens support, but while this holds things steady during use, there's no guarantee of perfect alignment. I will continue working on it.

May 24th, 2008, 01:09 PM	#8
Rich Hibner Major Player Join Date: Jan 2006 Posts: 218 Images: 1	flipping the camera will fllip the image on the lcd. but maybe it depends on the camera.