RED One and Master Primes issues

[Robert Horwell]

looks like we were all posting together, i was trying to make exactly the same point as Frederik. But in a much more long winded way...lol

 

[Mitch Gross]

Correct depth is an absolute measurement. It is a standard. It is never correct to set a lens based on a given mount or to set a mount based on a given lens. It is only correct to set either based on the standard.

Every device has some offset to the standard. This is known as its tolerance. A calibrating device such as the NULL Lens or a $60,000 Chrosziel bench collimator has an incredibly narrow tolerance because they are engineered to specifically to hold tolerance to the detriment of anything else. In the NULL that means that there is no focus ring so that every element of the unit stays locked in place.

A standard lens, even a really good one, has some slop in it. The fact that a focus ring can rotate means that it has a bit of front/back movement along the threads, otherwise they would bind. That slop creates a greater range of tolerance.

At Abel Cine Tech, we always set our lenses to within five microns of absolute depth. We do the same with our cameras. We do this using collimation tools with the narrowest of tolerances possible in the world. Let a room heat up or change humidity and that tolerance will shift.

Pawel, your assumption appears to be that if you set a lens based on a measuring device, then that lens can now be used as a measuring device. This is incorrect as it does not account for the tolerance of the lens, which must be added to the tolerance of the original calibration tool. This is a phenomena known as Tolerance Stacking, which basically menas that with every step removed from the original reference standard the imprecision of the devices involved reduces the accuracy of the measurement. When measuring in microns it is vital to stay as close to the original reference standard as possible.

For fun with statistics and math, I'm sure Wikipedia has some great references on Tolerance Stacking. Think of it as taking a photograph and then photocopying it repeatedly. How many layers of copies can you make before the image is barely recognizable?

 

[Matthew Duclos]

Correct depth is an absolute measurement. It is a standard. It is never correct to set a lens based on a given mount or to set a mount based on a given lens. It is only correct to set either based on the standard.

THANK YOU! Well said, my friend

 

[Sanjin Jukic]

Thanks Mitch.

Excellent explanation and great help for all of us.

 

[Robert Horwell]

Correct depth is an absolute measurement. It is a standard. It is never correct to set a lens based on a given mount or to set a mount based on a given lens. It is only correct to set either based on the standard.

Every device has some offset to the standard. This is known as its tolerance. A calibrating device such as the NULL Lens or a $60,000 Chrosziel bench collimator has an incredibly narrow tolerance because they are engineered to specifically to hold tolerance to the detriment of anything else. In the NULL that means that there is no focus ring so that every element of the unit stays locked in place.

A standard lens, even a really good one, has some slop in it. The fact that a focus ring can rotate means that it has a bit of front/back movement along the threads, otherwise they would bind. That slop creates a greater range of tolerance.

At Abel Cine Tech, we always set our lenses to within five microns of absolute depth. We do the same with our cameras. We do this using collimation tools with the narrowest of tolerances possible in the world. Let a room heat up or change humidity and that tolerance will shift.

Pawel, your assumption appears to be that if you set a lens based on a measuring device, then that lens can now be used as a measuring device. This is incorrect as it does not account for the tolerance of the lens, which must be added to the tolerance of the original calibration tool. This is a phenomena known as Tolerance Stacking, which basically menas that with every step removed from the original reference standard the imprecision of the devices involved reduces the accuracy of the measurement. When measuring in microns it is vital to stay as close to the original reference standard as possible.

For fun with statistics and math, I'm sure Wikipedia has some great references on Tolerance Stacking. Think of it as taking a photograph and then photocopying it repeatedly. How many layers of copies can you make before the image is barely recognizable?

Thanks mitch, i was beginning to think i was going mad!..but your way of putting is much more direct. Creative minds love analogies.

 

[Pawel Achtel]

Pawel, your assumption appears to be that if you set a lens based on a measuring device, then that lens can now be used as a measuring device. This is incorrect...

Not quite. What I say is that, at the end of the day, one must finally and ultimately check the camera's FFD with the taking lens because, well, you are filming and focusing with that very lens. If for any reason, like temperature or tolerance, the lens's FFD is not exact, your shots will be still in focus when you use that lens.

What is the point of having shots out of focus and correct FFD = 52.000mm?

 

[Mitch Gross]

But Pawel, you often cannot tell if it is correct with your len or not. You may not know until late in post. That's why it is important to have your lenses periodically checked as well. There is room in the tolerance if the camera is checked against a standard and the lens is checked against a standard. A lens that will not stay within tolerance, well it's time to get rid of that lens. And if you try doing this wth a zoom lens you may find that you cannot hold focus durng a focal length change.

Sure, when you're out in the field you do what you gotta do to get things to work. But bailing wire & chewing gum should not be the standard method of keeping one's car bumper on. With the use of proper tools you will not have to continually check each lens for accuracy every time you change from one to another. You may not change lenses much with the RED in an underwater housing, but I dare say that the vast ajority of users change focal lengths frequently during the course of the day.

 

[Pawel Achtel]

Mitch,

But, if you do not check your camera's FFD with the lens then IF either lens's FFD or camera's FFD changed for whatever reason, you are going to ruin the footage. If you check with the taking lens, then at least your footage will be fine.

The NULL is to set the camera's standard FFD. That's fine, but I would alsways check the camera's FFD on set with the taking lens. You mentioned a zoom lens. There is a simple procedure to adjust the FFD of a zoom that is routinely done on every shoot (or even take) on ENG cameras. I used to do it five times a day. It ensures that the particuilar lens is matched with the camera, not the standard.

I'm not saying that the NULL is completely useless, but certainly not as useful as some people hope it to be. It should not make you to assume that focusing scales are accurate just because you have checked camera's FFD with the NULL.

 

[Jarred Land]

im siding with Pawel on this one.. I think what he is saying is being twisted a bit, what he is actually saying is good ( i think )

If you have one camera, and only one lens (even just for a given shoot) setting the backfocus to that lens is not a bad thing to do. If the lens is out, or the camera is out, it is the only way most people that don't have access to dedicated tools can bring them both back together. Remember, tolerance stacking works both ways. If your lens is out, setting the Flange on the camera to a standard isn't going to do much, perhaps even make things worse.

Of course it is great to set both to a standard.. if you play by the book thats the way to go. but alot of people don't have the ability/resources of doing that every shoot.

 

[Sanjin Jukic]

I would need a precise instrument to measure all that

just because I work with shims to collimate my R1 IMS mount.

Optitek's Optimator looks pretty attractive to me:

Optimator on RED ONE.

"The OptiTek Optimator is a high-precision instrument
that assures precise sensor location checking and
adjustment of camera back focus and sensor flange depth.

Traditional methods of back focus checking that have long
been used with cine cameras are not possible with digital
cinema cameras because of sensor design and the possibility
of damaging it. Yet the extremely high resolution of
digital cinema cameras amplifies the effect of any sensor
geometry or back focus location errors. The Optimator is
an easy to use, cost effective solution for owner/operators
and rental houses alike.

This precision device simply mounts in place of the lens
on any single sensor, PL mount digital cinematography
camera such as the Sony F35 and the Red One, and
projects a high-resolution test chart onto the sensor.
When the camera’s video output of the Optimator’s test
chart is viewed on an instrument-grade monitor, the
chart becomes clearest when the lens mount is precisely
positioned relative to back focus. To achieve optimum
back focus positioning, lens mount adjustment can be
accomplished with the Optimator in place.

The Optimator’s high resolution optics are capable of
resolving in excess of 5,000 lines per inch (196.8 lines
per millimeter), which is several times more than the
resolution of any current digital cinema camera. And
because the Optimator’s extremely flat field displays the
whole frame accurately, it makes sensor tilt errors easy
to spot. The test image coverage area is 34mm, which
exceeds the imaging area of a 5K sensor.

Optimator’s test image is projected by a white LED backlight
with a power supply that is precisely calibrated to
assure identical light intensity on each step, and on each
Optimator. A built-in rechargeable battery and on-board
battery charger are provided to power the LED.

The system’s lens mount adapters are precision ground
stainless steel to minimize thermal expansion issues.
They provide for quick remounting on various systems
such as PL, Panavision, Nikon, etc."

READ MORE>>>

 

[Mitch Gross]

I think that Pawel is right in that it is always smart to double check the camera with the lens. But it is not the one and only way i would ever say to set your lenses or mount. This is coming from a guy who was a DP for several decades and owned his own camera packages for many years. In theory, my lens mount and my lenses should have never moved but in reality over time they would. So as smart maintenance practice I would send them over to my local camera house (at the time I was a loyal client to, you guessed it, Abel Cine Tech) and had them check the collimation on my lenses and camera. While of course I would do a double check using my eye, the camera and a given lens there was absolutely no way i would rely on this as my only verification of proper collimation.

 

[Jochen Schmidt-Hambrock]

Pawel has Master Primes and he doesn't seem to even look at the LCD when focusing. Just shoots the laser and dials it up on the MPs. Smack on every time! I'm pretty sure he said he didn't even have to adjust the back focus, it was dead on from the factory.
:001_tongue:

Jim :biggrinjester:

My experience exactly.

Jochen

 

[Stephen Strangways]

In an ideal world, owner/operators would all regularly collimate our cameras to 52mm, and shim all our lenses (as needed) to 52mm as well, ensuring everything matches.

If you own a camera and want to rent lenses from a good camera rental house that has collimated all of their lenses to 52mm, it makes sense to collimate your camera to the same standard.

However... if you own camera and lenses, then purchase a collimator to adjust your camera's FFD, but you do not have the ability to test your lenses and shim as necessary, you are not really do anything useful.
In that case, you are better off collimating your camera, which you can adjust easily, to your lenses, which you cannot adjust.

 

[Jacek Zakowicz]

R1 as Collimator

R1 as Collimator

In an ideal world, owner/operators would all regularly collimate our cameras to 52mm, and shim all our lenses (as needed) to 52mm as well, ensuring everything matches.

If you own a camera and want to rent lenses from a good camera rental house that has collimated all of their lenses to 52mm, it makes sense to collimate your camera to the same standard.

However... if you own camera and lenses, then purchase a collimator to adjust your camera's FFD, but you do not have the ability to test your lenses and shim as necessary, you are not really do anything useful.
In that case, you are better off collimating your camera, which you can adjust easily, to your lenses, which you cannot adjust.

You can certainly adjust BF on any lens. Once the camera is set precisely you can adjust lenses WITH it just like you would on a collimator or projector.
While it's not a "cave man" job anyone with average technical understanding and some experience/training will handle it just fine. I'm planning a simple accessory to be supplied with my Optimator that will make that procedure very transparent and straight forward. I tried it this weekend and it's doing what it's supposed to do- consistant readings especially on some still zooms.
It will work on all zooms and primes, still or PL under one condition- your camera has to be adjusted correctly.

 

[Sanjin Jukic]

You can certainly adjust BF on any lens. Once the camera is set precisely you can adjust lenses WITH it just like you would on a collimator or projector.
While it's not a "cave man" job anyone with average technical understanding and some experience/training will handle it just fine. I'm planning a simple accessory to be supplied with my Optimator that will make that procedure very transparent and straight forward. I tried it this weekend and it's doing what it's supposed to do- consistant readings especially on some still zooms.
It will work on all zooms and primes, still or PL under one condition- your camera has to be adjusted correctly.

Jacek,

great news.:hurray:

Looking forward to get more about it.