Collimation on RED ONE

[Emanuel A.]

Actually, I'm not the right person to ask this. Although I'd say that Stephen is the right person to answer on this.

I've been in touch with a very cool person from the industry in hollywood. He has worked with names like Spielberg, etc. I just won't disclose his name here though I can tell you he's the most well knowledgeable person that I could know on these days. Same level than our Stephen.

And he asked me a relevant question:

How would I collimate using a digital sensor camera since it's not the same than a film capture device?

 

[Kevin Halverson]

I guess I don't really understand your concern. Setting back focus is a common task and really not difficult at all. In the case of the RED, it might be based upon shims, or it might be a bit easier, I don't know that these details have been released yet. I am confident that they have designed a scheme that won't be excessively complicated.

 

[Brook Willard]

It could just be that this person isn't 100% sure of how the sensor works. People who are used to 3CCD cameras often don't realize that a Bayer sensor can be effectively treated just like a film camera with regards to collimation. The PL-mount flange focal distance is 52.00mm... just collimate like you would with a film camera. No magic 3CCD trickery required.

That said, I'm sure the cameras will be collimated from the factory.

 

[Emanuel A.]

Thanks for the answers. As I said, I'm not the best man for this doubt. Actually, I've suggested to him in order to put both (plus Stephen) in touch. My concern is regarding to the changing mounts. I can't invest on the Birger's mount, for example, if I haven't a back guarantee about avoiding future problems on field.

As I said, this person is very knowledgeable about every issues on the motion picture technologies related. I have learnt a lot with him and he has been very helpful in order to give me some lights on my glass purchases.

I couldn't imagine the secrets beyond the PL compatibility as far as the PL mount adapters is concerned. BE CAREFUL with this! NOT ALL the lenses are compatible with a PL mount. No matters if you have or not the right PL mount adapter to the job...

I believe this question is so valid that this thread is open. As matter of fact, I invited him to our forum. This would be a great resource for this community. I hope he may accept it. We wait and see.

Other thoughts?

 

[Emanuel A.]

Here's morning. I could check my email and had got mail sent six hours ago from the same person. Very nice and friendly human being. I'd add, already a friend. I had asked him his input in order to proceed with my inquiry.

Here goes his question:

«HI Emanuel, the question I have is, how do you use the auto collimator on the Red One Camera? When you use an Auto Collimator with a film camera the procedure is generally as follows:



1) A precision first surface mirror is fit onto the film camera so that its mirrored surface is resting flush and flat on the side rails that the film emulsion would ride on when the film is being exposed in the camera.

2) A taking lens is fit on the lens mount of the film camera, the focus ring on the lens is set to infinity, the iris ring is set in the full open position and the camera shutter is moved so that it is not in the light path between the lens and the film plain.

3) The auto collimator is placed in front of the taking lens so that the reticule image that the collimator generates is projected through the taking lens and is then reflected off the first surface mirror that was installed in the camera aperture and back into the collimator.

4) If the lens is calibrated correctly at infinity, and the cameras flange focal depth (the distance from the lens mounting surface to the film plain) is correct, the collimator reticule image should be sharp when seen through the collimators eye piece.

5) See attached sketch.





This test tells you three possible things.



1) All is adjusted correctly.

2) The lens needs adjusting.

3) The flange focal depth is incorrect.



So my question on the Red One camera is how you put a precision reflective surface at the image sensor position in order to check the collimation. Can you put a first surface mirror in the image sensors position? Is the front surface of the sensor flat enough and reflective enough to reflect the collimators reticule image back to the collimator? Are there any optical surfaces between the lens mount and the image sensor that would interfere with the collimation of the optical system? I’m sure someone will have the answer to my question.



Best Regards,

RB»

 

[Stephen Williams]

Hi Emmanuel,

I asked the same question in October 2006.
http://www.cinematography.com/forum2004/index.php?showtopic=17540

Stephen

 

[Mr. Paul White]

Alright but what's the conclusion?

 

[Kevin Halverson]

Again, I still don't see the concern. There is no practical difference if you use an optical device to set the FFD, or any other method, as long as the method results in a correct distance between the rear element and the sensor surface. That is all that is being adjusted and more importantly, all that matters.

 

[Don King]

??

Sorry, this is very confusing. Two professionals with film background are doing the same question so...

On the other hand, Mr. Williams has been very proactive about collimation so he should have had any conclusion or am I missing something there too?

 

[Stephen Williams]

Again, I still don't see the concern. There is no practical difference if you use an optical device to set the FFD, or any other method, as long as the method results in a correct distance between the rear element and the sensor surface. That is all that is being adjusted and more importantly, all that matters.

Hi,

The only question is will it work with a Red one? I haven't myself had a chance to test it so......

Stephen

 

[Kevin Halverson]

Hi,

The only question is will it work with a Red one? I haven't myself had a chance to test it so......

Stephen

Will what "work"?

Are you asking about using an optical collimator or a depth gage?

The reflective properties of the imager surface should result in an 'OK' image for optical collimation, but since no one has had access to a RED One, we can't be certain. Using a depth gage might result in a correct distance, but it seems that you really risk damaging the imager, so I don't "think" this would be a good approach.

I can think of a number of other approaches that would be fine too. One might be a situation where you have a lens that is known to be correct, then it could be used to confirm the flange to imager distance by comparison to a know correctly setup camera. Verification of the focus distance against a know correct lens is another valid approach. Really, there are a number of ways to accomplish the task.

What matters is getting the distance correct, anything that results in that goal should be just fine.

 

[Stephen Williams]

??

Sorry, this is very confusing. Two professionals with film background are doing the same question so...

On the other hand, Mr. Williams has been very proactive about collimation so he should have had any conclusion or am I missing something there too?

Hi,

Does any body know how Panavision Collimates the Genesis?

A collimator will provide an image at infinity, by enlarging the output of a Red one can see if a lens set at infinity is in correct focus, however how do you calculate how many microns to adjust if it not, experience will put you in the ballpark. An autocollimator allows you to look back through the lens onto the film plane seeing the image from the collimator, the 'auto' part may or may not help, until somebody tries with a Red One we are guessing!

Stephen

 

[Stephen Williams]

Will what "work"?

Are you asking about using an optical collimator or a depth gage?

The reflective properties of the imager surface should result in an 'OK' image for optical collimation,

What matters is getting the distance correct, anything that results in that goal should be just fine.

Hi,

Is the 'OK' image that you see in the right place, or could it be wrong by a few microns? I think you can forget a depth gauge!

Stephen

 

[Kevin Halverson]

Hi,

Is the 'OK' image that you see in the right place, or could it be wrong by a few microns? I think you can forget a depth gauge!

Stephen

If you think about it, a front surface mirror doesn't precisely emulate the properties of pan chromatic film by much more than a few microns! All these methods only get "close" in reality. How close is close enough is really a bit subjective.

I do fully agree, I ain't letting anyone with a depth gage anywhere near one of these cameras!

 

[Dexter Gregoire]

What he's talking about is the Collimation do not lend themselves to exact theoretical understanding, because they are composed of many flange focal depths but but are not large enough to be accurately described as dynamic. They are held together by strong interactions like the lens, sensor, etc... and are on the order of 1 % of their relativistic rest with errors usually smaller than those from other approximations.:clown2:

The conclusion, being in focus is very important and I believe the REDONE being like a digital DSLR has it's own way of telling you that you are in focus.

The REDONE is proving itself and I'm sure the Birger mount will do the same-(fingers crossed).

 

[Stephen Williams]

If you think about it, a front surface mirror doesn't precisely emulate the properties of pan chromatic film by much more than a few microns! All these methods only get "close" in reality. How close is close enough is really a bit subjective.

Hi,

When using a autocollimator with a film camera, I don't use a mirror, I use film and have the camera running. That gives me more information than a static image, also the changes that occur when the camera runs at different speeds .

Stephen

 

[Stephen Williams]

What he's talking about is the Collimation do not lend themselves to exact theoretical understanding, because they are composed of many flange focal depths but but are not large enough to be accurately described as dynamic. They are held together by strong interactions like the lens, sensor, etc... and are on the order of 1 % of their relativistic rest with errors usually smaller than those from other approximations.:clown2:

The conclusion, being in focus is very important and I believe the REDONE being like a digital DSLR has it's own way of telling you that you are in focus.

The REDONE is proving itself and I'm sure the Birger mount will do the same-(fingers crossed).

Hi,

Having an image focused at infinity is useful to check lenses. Focus pullers have to use the marks on lenses, by the time you see the image is soft on monitor it's too late!

Stephen

 

[Kevin Halverson]

Hi,

When using a autocollimator with a film camera, I don't use a mirror, I use film and have the camera running. That gives me more information than a static image, also the changes that occur when the camera runs at different speeds .

Stephen

Even this approach is still subjective and only very "close" as the image that you see is from the boundary between the anti-halation layer and the emulsion layer's and is therefore not "exactly" representative. Also consider the effect of different wavelength of light.

At some point in time, we are getting down to point where the ability to "observe" a process without influencing it in some, unintended way, becomes a factor. Luckily, this point is beyond what we need to achieve for good results.

 

[Stephen Williams]

Even this approach is still subjective and only very "close" as the image that you see is from the boundary between the anti-halation layer and the emulsion layer's and is therefore not "exactly" representative. Also consider the effect of different wavelength of light.

Hi,

Thats why one normally uses a depth gauge to measure the FDD, & put lenses on a collimator. Using just an autocollimator is not accurate enough as several errors could cancel out each other. Quite what effect this will have when changing Red lens mounts in a hotel room or worse on location remains to be seen!

Stephen

 

[Kevin Halverson]

Hi,

Thats why one normally uses a depth gauge to measure the FDD, & put lenses on a collimator. Using just an autocollimator is not accurate enough as several errors could cancel out each other. Quite what effect this will have when changing Red lens mounts in a hotel room or worse on location remains to be seen!

Stephen

I couldn't agree more.

Really, unless we are talking about PL mount glass, I don't think this applies. Individuals that try to use still camera mount versions can't expect repeatability (lens to lens) that justifies this level of concern. Even with just one of these types of lenses I doubt you could mount and then remount it twice and achieve the same results. Again, its all speculation until we see what provisions RED has made for this.