VistaVision and Monstro and cinematography?

[Erich Roland]

How many days was it gone if I may ask?

Waiting for mine to come back.

We just received our 2nd upgrade back and it was just over 2 weeks away. The first Monstro took 3 weeks.

 

[Tom Gleeson]

How many days was it gone if I may ask?

Waiting for mine to come back.

Björn,

Around ten days including travel.

 

[Marcus Friedlander]

Björn,

Around ten days including travel.

Can we all take a second and appreciate how impressive that is?

Oh how times have changed (for the better :rofl

 

[Tom Gleeson]

Just a quick update that with the ACS we shot some "basic principle" tests looking at VistaVision sensor versus S35mm sensor. Collating results at the moment but using Resolve and FCPX is a slow and arduous process as I am only a part time user of post software. There is much cussing and gnashing of teeth. Its also school holidays and Summer break so this will also slow the process but it is happening. Looking back at the test footage I will also admit I suck at the scientific process.

 

[Hrvoje Simic]

If you use a longer lens in VistaVision to match the field of view of a shorter lens on a Super-35 camera, there is no "flattening" effect... the perspective, the view, is the same -- the only difference will be depth of field unless you compensate.

I am confused, it must be the language and terminology.

My tests with S35 and FF from the same position and compensating focal length to match FOV showed more natural volume representation and less object enlargement towards the camera on a larger format. The larger the format, the less size discrepancy in z-axis. Like the shot in Steve Yedlins test, where in a wide shot IMAX format shot shows more natural proportions of a man sitting in a chair with legs closer to camera, while on Alexa's shot his legs are disproportional to body. Same with portraits and face depth proportions in general. The shot is not "flatter" in terms of inter-object relations, as when switching to a longer lens on a same format, but the objects are less distorted through scene depth.

 

[Michael Tiemann]

I am confused, it must be the language and terminology.

My tests with S35 and FF from the same position and compensating focal length to match FOV showed more natural volume representation and less object enlargement towards the camera on a larger format. The larger the format, the less size discrepancy in z-axis. Like the shot in Steve Yedlins test, where in a wide shot IMAX format shot shows more natural proportions of a man sitting in a chair with legs closer to camera, while on Alexa's shot his legs are disproportional to body. Same with portraits and face depth proportions in general. The shot is not "flatter" in terms of inter-object relations, as when switching to a longer lens on a same format, but the objects are less distorted through scene depth.

Are you keeping the nodal points of the two systems the same? It's not the distance to the film plane you have to keep constant, but the distance to each lens's nodal points that must be constant.

 

[Tom Gleeson]

I am confused, it must be the language and terminology.

My tests with S35 and FF from the same position and compensating focal length to match FOV showed more natural volume representation and less object enlargement towards the camera on a larger format. The larger the format, the less size discrepancy in z-axis. Like the shot in Steve Yedlins test, where in a wide shot IMAX format shot shows more natural proportions of a man sitting in a chair with legs closer to camera, while on Alexa's shot his legs are disproportional to body. Same with portraits and face depth proportions in general. The shot is not "flatter" in terms of inter-object relations, as when switching to a longer lens on a same format, but the objects are less distorted through scene depth.

David's statement "If you use a longer lens in VistaVision to match the field of view of a shorter lens on a Super-35 camera, there is no "flattening" effect... the perspective, the view, is the same -- the only difference will be depth of field unless you compensate." is correct with perfect lenses. The perspective is locked by the geography and position of the camera. The "confusion" comes from the design of the lenses themselves and the difficulty of creating lenses that have no distortion which gets more and more difficult as focal length decreases. In our tests we used the Zeiss CZ.2 28-80mm zoom and we compared a 28mm 35mm(4.5K) shot to a 40mm (8K) and the pin barrel distortion at 28mm changes objects position and size compared to 40mm. I think a major advantage of large formats is wide FOV shots without using short focal lengths with all their limitations. Dan Sasaki of Panavision talks about this issue that a lens designer has to make compromises and to correct all the geometry in a wide angle lens they sacrifices their ability to add character to the lens.

Note: We also used FOV to match shots in our tests to take into account a moving nodal point on our zoom lens.

 

[Hrvoje Simic]

It's not the barrel distortion. Barrel distortion also distorts the image in X & Y axis.
Camera position, angle and distance to object are the same.

Objects closer to camera always get larger.
The larger the format the reduced discrepancy between distance/size and more natural depth proportions. That's why a large format intuitively feels better. More natural depiction of volume .






Toggle between Steve Yedlin's Alexa 35 and IMAX grabs and watch the facial proportions.
Head angle differs so a bit confusing but still noticeable.

 

[Michael Tiemann]

[...]
Camera position and angle is the same.

Objects closer to the lens always get larger. [...]

So you are not respecting the nodal points of the lenses. If you want to compare apples to apples, it's not about locking down the tripod position, the camera position, nor even the image sensor position. It's about locking down the nodal point of the lens and ensuring that the master ray (central ray of the subject) lands at the center of the image. Anything else and you are just comparing whether a subject is more flattered by pushing in some inches or pulling back some inches. Pushing in a few inches is not going to change the rendering (or the apparent size) of a distant background very much. But it can change the perspective on a close-in subject very much. Given that cinematography is, indeed, a game of inches, you don't want to let those inches confuse what you are measuring.

 

[Hrvoje Simic]

So you are not respecting the nodal points of the lenses. If you want to compare apples to apples, it's not about locking down the tripod position, the camera position, nor even the image sensor position. It's about locking down the nodal point of the lens and ensuring that the master ray (central ray of the subject) lands at the center of the image. Anything else and you are just comparing whether a subject is more flattered by pushing in some inches or pulling back some inches. Pushing in a few inches is not going to change the rendering (or the apparent size) of a distant background very much. But it can change the perspective on a close-in subject very much. Given that cinematography is, indeed, a game of inches, you don't want to let those inches confuse what you are measuring.

I've edited the upper post for clarity.

Same distance to object (not push in), same camera angle, same angle of view, different formats/focal length combos depict volume differently. Check the face in that example.

There is no way a portrait on a wide lens shot up close on a smaller format can look as natural as an equivalent angle of view lens on a large format. It matters less as distance increases and matters more the closer the camera gets. If the talent in that shot was even closer, framed for the face to fill the frame the effect would be more apparent and face more distorted in depth. Nose would look noticeably larger than it is for example.

 

[Michael Tiemann]

Not sure where this is heading so I'l try to clarify.

Same distance to object (not push in)[...]

Measured from what? The image plane? The front of the lens? The nodal point of the lens? My claim is that keeping the image planes the same may still provide different "depth" or "volume" or whatever other sensual term you'd like to use, unless the nodal points are also the same. It's the measurement from the nodal points that brings all the math into alignment.

 

[Björn Benckert]

It's not the barrel distortion. Barrel distortion also distorts the image in X & Y axis.
Camera position, angle and distance to object are the same.

Objects closer to camera always get larger.
The larger the format the reduced discrepancy between distance/size and more natural depth proportions. That's why a large format intuitively feels better. More natural depiction of volume .






Toggle between Steve Yedlin's Alexa 35 and IMAX grabs and watch the facial proportions.
Head angle differs so a bit confusing but still noticeable.

As long as you look at a human head as reference that is not even fixated you might as well use a rubber band to measure length...
Then you take different lens designs into account... sure a fisheye looks different than a perfectly rectlinear lens.

But the differnce of the guys head shape has to do with a few things, his position when he sits infront of camera differs. Lens design differ, nodal point position most likely differ. As the guy is close to camera the error in messurments become even more pronounced.

Mathematically Mullen is right. The rest is like saying one lens could look different from another. And yes they usually do.

That Yeldin test is so cluttered with misassumptions its almost scary to watch especially as it´s so formally well made.

 

[Hrvoje Simic]

But the differnce of the guys head shape has to do with a few things, his position when he sits infront of camera differs. Lens design differ, nodal point position most likely differ. As the guy is close to camera the error in messurments become even more pronounced.

And there is also....the format size/focal length combo affecting volume depiction.

Btw, head shape is not all what's happening here.
Please observe the head size compared to shelves behind.

Mathematically Mullen is right.

I didn't suggest David was wrong, just proposed one more factor so no need for polarisation.

The rest is like saying one lens could look different from another.

No, the rest is saying that format size does make a difference.

 

[Jacek Zakowicz]

I agree with Bjorn 100%
Lets get real here:
For IQ ,perspective and EVERY technical aspect of the (optical formed by the lens) image the format makes unequivocally, absolutely NOOOOO difference. This thread proves it once and for all. The framing differences are operator error and maybe , lens distortion characteristic difference, the overall sharpness difference between the two lenses (that DOF variance). That's ALL ;-)
Happy New Year!!!!!

 

[Ben Scott]

Can someone tell me what the nodal point of the lens is?

 

[Les Hillis]

Can someone tell me what the nodal point of the lens is?

The point within a lens at which the light rays it gathers in from different angles are made parallel for 'projection' onto the flat surface of the camera sensor.

http://www.hugha.co.uk/NodalPoint/Index.htm

 

[Björn Benckert]

The point within a lens at which the light rays it gathers in from different angles are made parallel for 'projection' onto the flat surface of the camera sensor.

http://www.hugha.co.uk/NodalPoint/Index.htm

The point does not necessarily need to be within the lens. Sometimes its even behind.

But yes, its the point which you can pan / tilt around with zero parallax. change.

 

[Les Hillis]

Thanks for the clarification.

 

[Ben Scott]

Thanks

 

[Peter Lyons Collister ASC]

So I am sorry but I may be confused. Is someone saying you can't design really fast lenses that cover VV?

I shot on a Monster equipped PV DXL last night on a 50 year old 50mm T/1.0 Panavision lens and it totally covered the VV sensor.