SALT test...

[lensmaster]

Pawel,

Is this the "on-axis" MTF. What about off axis? I also expect a very decent difference at 2.8 and 4.5. At 4.5, these lenses should have a 4.5/2.8 larger airy disk. When slowed down so much, I bet the blur looksvery circular on the MP's

I have done alot of MTF testing in my life. I have a nice bench but it's not big enough to fit the MP;s

 

[Paris Remillard]

Hey Peter -

I'm not doing any custom work for a while (getting ready for a shoot in NYC) but I can tell you how to do it:

Get a PL receiver from Cinevate. I think theirs are the most available and affordable. You buy the PL adapter for their Brevis units and remove the PL receiver part.

Next, take the EF/EOS mount off the 5D. On Pawel's 50D it came off by removing the four phillips head screws. They can be hard to turn so get a good screwdriver and push down hard. Take your time.

Now, have a machinist make you a collar or stand off to go between the receiver and cam. You'll need to get four longer screws for this; I suggest McMaster Carr.

How thick should the collar be? Well, the EOS is 42mm and the PL is 52. So you do the math and make a drawing for your machinist, measuring the thickness of the Cinevate PL flange and the thickness of the EOS mount you just removed and accounting for them. A digital Mitutoyo caliper is nice to have here.

Now, the big question is how to remove the mirror and any other obstruction from inside the 5D so you can use all your PL mount lenses. For that I think the thing to do is find an experienced Canon tech who can pull that stuff out without harming the rest of the camera's functions. I don't know how to do this but I would think the best bet is to find that Canon tech, wherever s/he is.

For the SALT test, though, maybe RED can free up an EPIC???

Hey, Doug. Is there a reason that one couldn't make a collar or stand off with the female pl-mount on the front and a male EF mount on the back? So that you could just mount the PL-Mount adapter like an EF mount lens or teleconverter, etc... Is the opening in the male EF mount too narrow to allow for the size of the rear element of a PL-mount lens?

 

[Pawel Achtel]

Pawel,

Is this the "on-axis" MTF. What about off axis? I also expect a very decent difference at 2.8 and 4.5. At 4.5, these lenses should have a 4.5/2.8 larger airy disk. When slowed down so much, I bet the blur looksvery circular on the MP's

I have done alot of MTF testing in my life. I have a nice bench but it's not big enough to fit the MP;s

The MTF machine at ARRI measures MTF in the centre. You need to put the lens on a projector to see off axis performance. The most remarkable of all the lenses I tested was the 14mm with perfectly sharp and flat image edge to edge and projection showed easily distinguishable 100lp/mm pattern in the corners! Considering this is also the hardest focal length to achieve this, I have no idea how they did it, but they did.

The MTF of the MPs that I tested at f/2.8 seemed to be within 5-7% of the max MTF reached at around f/4.5 - quite remarkable. And, above f/5 all MPs were diffranction limited already.

Paris, it has been discussed that PL to EF adapter is not possible. There EF openning is too narrow and too shallow.

 

[Mitch Gross]

Gee, if only there was a Digital Cinema camera available that had a sensor large enough to show the larger image circles being discussed...

The Phantom HD has a square sensor, 2048x2048 with a photosite size of 12.5 microns. That's 25.6mm in height and width, making a diagonal of a whopping 36.2mm. Plenty big for the point of this conversation.

It's too bad that Abel only has more of these cameras than anyone else on the planet. ;-0

I can't promise that I can free up one of these cameras for SALT but once we lock down the date I'll be happy to see what I can do.

Of course there's also the Phantom 65. That's 4096x2440 at 12.5 micron pitch, which means a sensor 51.2mm x 30.5mm. That's a diagonal of 59.6mm. That should do the trick, no?

 

[Matthew Duclos]

Mitch, I like your style.

 

[david farland]

Makes me think of an episode of the the Simpsons when Homer is about to get written out of a TV program and asks why they can't do a live episode where he can protest about the treatment the station has been giving him. Reply was they probably could, but it'll be murder on the cartoonist's hand.

no a lot to do with the topic except Mitch's offer of using a Phantom sensor with huge 12um pixels made me think of large cartoon lines...ummmm.

But the question is what would be the difference in resolving power of the Phantom's 2K x 12um sensor compared to the Red's 4K @ 5.4um sensor...ie. pixels are 2x greater?

D

 

[Pawel Achtel]

Dave, the idea of large sensor is to check other properties, like shading, vignetting, sharpness across the image circle, CA, etc, etc... It's a good idea.

Of course, it won't give the MTF, which I will be flattered to have the opportunity to measure :smash:

 

[Douglas Underdahl]

Paris - yes, exactly as you stated, the EF mount is too narrow to fit the back of PL mounted lenses. I think that was why I removed it and made a new collar.

 

[lensmaster]

There are other machines that will measure off axis performance (like mine). Mine does infinity only but I have measured using off axis machines.

Using the Phantom with 12.5um pixels will not give the same appearance as the RED> The smaller pixels on the RED should be more demanding on the lenses (although the OLPF effectively doubles the pixel dimension in both directions.

I have seen the MP 14m and I agree it's one of the greatest lenses around. I was amazed at the sharpness. That's what aspheres and semiconductot optics tolerance will do.

 

[Matt Uhry]

The Phantom 65 / EOS 5d etc. would only in the interest of establishing coverage and quality at the far edges.

Primary comparative tests will be shot on the Red @4k, That's what we get paid to shoot with.

Everybody please stop spitting hairs about MTF and the like... This will primarily be a VISUAL test. We'll attempt to quantify how the lenses will do on future cameras.

If you want to buy a lens set based on MTF #'s A) Do your own tests. B) Seek filmmaking re-education.

Matt Uhry
www.mattuhry.com

 

[Larry Gebhardt]

Not to sound like I am looking for errors, but the Canon EF lens mount's flange focal distance is 44mm, not 42mm (that was the older FD mount's FFD).

I just now removed the metal lens mount on an older Canon 350D that I have and as I have been thinking about this for quite a while (making a PL mt. front end for a Canon camera to check lenses, etc.), again decided to look into making one of these and while I am at it, make it friendly to the wider lenses that extend well into the mount.

The metal mount on the 350D is 3.19mm thick and the collar should interface with the tabs (counterbore flats) that the original mounts sits on...thus the collar thickness (length) should be 52mm minus 44mm, plus 3.19mm, minus the thickness of the PL mount base that you are using. Mine is 5.67mm (PL mount left over from another project) giving me a collar length of 5.52mm. The ID of the collar should be a little larger than 54mm (OD of Arri mount lenses...I measured the PL mount I have and it is 54.04mm) allowing for a close, smooth fit with your lenses. The OD of the collar should be 65mm , just a bit smaller than the opening on the 350D...the collar will slip into this opening and be attached as the original mount was, with 4 screws. McMaster Carr is the place to find the longer screws, as Doug said.

Now the hard part....Zeiss wide angle primes extend well into the lens mount cavity, and on both the 350D and the 5D MKII I just bought, there is a flat area surrounding the mirror box. This is what is hitting the rear element cone on my wider lenses. I am going to go to a still camera repair shop near me tomorrow and get a guy I know there to take the 350D apart and see what I need to do to make clearance for the rear ends of the lenses I am familiar with. I have a bunch myself, but none of the newer MP's or UP's. I have the T2.1's and a bunch of zooms, but from my assortment, only the wide and normal angle primes will be affected (16, 24, 32mm). I will know more later. I may have to dedicate the 350D to this project (no big deal) as I may have to mill off some of the mirror box to make it work. The 5D2 does have more room inside the mount, but the piece that holds the lens electronic interface (the small brass pins) would have to be removed...and I sure don't want to do that to my brand new baby. Anyone have one (5D2) that could be donated to the pursuit of science? Haha! Tom...you listening?

I will keep the brothers here up to date on this project, but as it is not a priority, it may take a while.

Though this might help if anyone else is thinking about doing a similar thing. I don't think Nikon cameras will work as their lens mount and thus the internal clearance is smaller than Canon's...but I could be wrong.

Larry

 

[Tom Lowe]

Not to sound like I am looking for errors, but the Canon EF lens mount's flange focal distance is 44mm, not 42mm (that was the older FD mount's FFD).

I just now removed the metal lens mount on an older Canon 350D that I have and as I have been thinking about this for quite a while (making a PL mt. front end for a Canon camera to check lenses, etc.), again decided to look into making one of these and while I am at it, make it friendly to the wider lenses that extend well into the mount.

The metal mount on the 350D is 3.19mm thick and the collar should interface with the tabs (counterbore flats) that the original mounts sits on...thus the collar thickness (length) should be 52mm minus 44mm, plus 3.19mm, minus the thickness of the PL mount base that you are using. Mine is 5.67mm (PL mount left over from another project) giving me a collar length of 5.52mm. The ID of the collar should be a little larger than 54mm (OD of Arri mount lenses...I measured the PL mount I have and it is 54.04mm) allowing for a close, smooth fit with your lenses. The OD of the collar should be 65mm , just a bit smaller than the opening on the 350D...the collar will slip into this opening and be attached as the original mount was, with 4 screws. McMaster Carr is the place to find the longer screws, as Doug said.

Now the hard part....Zeiss wide angle primes extend well into the lens mount cavity, and on both the 350D and the 5D MKII I just bought, there is a flat area surrounding the mirror box. This is what is hitting the rear element cone on my wider lenses. I am going to go to a still camera repair shop near me tomorrow and get a guy I know there to take the 350D apart and see what I need to do to make clearance for the rear ends of the lenses I am familiar with. I have a bunch myself, but none of the newer MP's or UP's. I have the T2.1's and a bunch of zooms, but from my assortment, only the wide and normal angle primes will be affected (16, 24, 32mm). I will know more later. I may have to dedicate the 350D to this project (no big deal) as I may have to mill off some of the mirror box to make it work. The 5D2 does have more room inside the mount, but the piece that holds the lens electronic interface (the small brass pins) would have to be removed...and I sure don't want to do that to my brand new baby. Anyone have one (5D2) that could be donated to the pursuit of science? Haha! Tom...you listening?

I will keep the brothers here up to date on this project, but as it is not a priority, it may take a while.

Though this might help if anyone else is thinking about doing a similar thing. I don't think Nikon cameras will work as their lens mount and thus the internal clearance is smaller than Canon's...but I could be wrong.

Larry

Hahaha, HELL no, Larry! Your project sounds great, though.

Man, this is a great thread to read. I'm learning a lot.

The MTF machine at ARRI measures MTF in the centre. You need to put the lens on a projector to see off axis performance. The most remarkable of all the lenses I tested was the 14mm with perfectly sharp and flat image edge to edge and projection showed easily distinguishable 100lp/mm pattern in the corners! Considering this is also the hardest focal length to achieve this, I have no idea how they did it, but they did.

:yikes: Wow, I'd kill to get a chance to shoot some night stuff on that Franken-50D of yours with the 14mm if you can figure that mirror thing out. Wow! If you can do it, I would just pull the mirror out and toss it down the laundry chute. I never even use the mirror on 5D2, only LiveView. :beer: