FF35 Cine Glass: Get Ready

[Pawel Achtel]

Thanks to Pawel Awel for the Master Primes image circle dimension.

Awesome! Thanks for this, Emmanuel. I like this teamwork.

I also wanted to add that the 31mm was a usable and sharp image circle. The acctual image circle was slightly larger, with some brightness and sharpness fall-off.

 

[Tom Lowe]

Pawel, was that image circle for all MPs?

How does that image circle compare to other PL-mount cinema 35mm and S35mm glass? I wonder if Jim had Master Primes in mind when he made Epic's S35 30mm?

Since the S35 cut-off is 30mm, then probably even with MPs productions will shoot S35, which is apparently 5K on the S35 Epic. (Do we know what the resolution would be for S35 windowed on the FF35 Monstro?) That might be enough for "True 4K," though I'm a big believer in massive oversampling, especially in low light.

 

[David Mullen ASC]

I see there is a 5K mode listed for the 6K FF35 Monstro EPIC -- since it is probably a windowed mode, there's a chance that the Master Primes would just barely cover that area, since 6K = 36mm wide, then 5K = 30mm wide. Of course, the lenses would all behave a bit more wide-angle if you're trying to get a 30mm wide image out of a lens meant for a 24mm wide format.

 

[Pawel Achtel]

Pawel, was that image circle for all MPs?

No, 31mm was the smallest and measured for the 35mm and the 14mm. The 75mm had larger image circle (38mm) and the Red 300mm covered the entire FF35.

That might be enough for "True 4K," though I'm a big believer in massive oversampling, especially in low light.

It would cover around 5k in 2.40:1 on all MPs. What is appealing is that the MTF of MPs at the resolution limit of the sensor is still quite high, it is across the entire field of view and it is available at f/1.2. And this is what swings my decision to go with MP/S35 format rather than Still glass/FF35. I would still use Still glass/FF35 combination, but only for specialty shots, such as night time lapse. Just about anything else calls for MP/S35. Even day-light time lapse will benefit from MPs due to their incredible resistance to flaring and accurate focusing scales.

 

[Fredrik Callinggard]

A friend of mine guessed that you most probably would "kind of" cover it around the 50mm Master Prime but we're going to do tests.

 

[Emmanuel Cambier]

Since the S35 cut-off is 30mm, then probably even with MPs productions will shoot S35, which is apparently 5K on the S35 Epic. (Do we know what the resolution would be for S35 windowed on the FF35 Monstro?) That might be enough for "True 4K," though I'm a big believer in massive oversampling, especially in low light.

A 2.40:1 image from the windowed FF35 using the MPs will roughly be 4760px X 1983px

Emmanuel

 

[Tom Lowe]

Sounds like that 75mm MP might be a popular item!!

Pawel, I'm surprised that you are not planning to get FF35 with its increased DR? At least then, you would have the option of throwing on some FF35 glass for certain specific projects.

 

[Emmanuel Cambier]

I see there is a 5K mode listed for the 6K FF35 Monstro EPIC -- since it is probably a windowed mode, there's a chance that the Master Primes would just barely cover that area, since 6K = 36mm wide, then 5K = 30mm wide. Of course, the lenses would all behave a bit more wide-angle if you're trying to get a 30mm wide image out of a lens meant for a 24mm wide format.

When I was puting up the diagram, I got a shock when I realized the MPs whouldn't cover the 30mm X 15mm S35 sensor. As you can see the corners will vignette, this is due to the 2:1 ratio of this sensor.
The widest you will get is around 4760px same as with the FF35 windowed.
So their dosen't seem to have any advantage going with a S35 sized sensor.

Emmanuel

 

[Pawel Achtel]

Sounds like that 75mm MP might be a popular item!!

My understanding is that all longer focal length MPs cover 38mm image circle, not just the 75mm.

Pawel, I'm surprised that you are not planning to get FF35 with its increased DR? At least then, you would have the option of throwing on some FF35 glass for certain specific projects.

Oh, I will get a FF35 for exactly the reasons you mentioned. It is just that I will be mostly shooting S35 with PL lenses.

 

[Pawel Achtel]

According to my calculations (based on Brook's sensor chart: http://www.reduser.net/forum/showthread.php?t=23770), the Mysterium-X would allow for 4860x2048 (2.37:1) assuming 30mmx15mm (5120x2700), pixel size 0.0059mm. I think 4860x2048 would be very compelling acqusition format for 4k finish. It would also allow for some look around area.

The FF35 Monstro would allow for just a whisker less resolution with its slightly larger 0.006mm photosite size.

 

[Joel Kaye]

Otherwise, I think most would rather the sensor stayed S35 in size and somehow we managed to get that extra quality. It just may not be possible to improve things dramatically without resorting to physically larger sensors.

OK, I get that idea. IF a larger sensor is required in order to get more dynamic range then people would want it. I'm kinda hoping they can pack the dynamic range into s35mm because I'm already happy with the resolution for the kinds of things I shoot. Maybe that's not possible. I haven't been keeping up with all the specs being tossed around. I kinda just want to see the footage vs. RED when possible.

 

[Tom Lowe]

FF35's 30mm S35 mode might be cutting it very close for a "True 4K" finish. The spatial efficiency needed for that (85%) is actually beyond what most of us consider possible from Bayer-filtered cameras. I think it's even beyond what Jim and Graeme have stated is possible (as of right now). Many people say that 70% spatial efficiency is much more realistic, especially for low-light shooting.

The IQ gained by shooting full 6K might be significant vs 4.7K origination. We'll have to wait and see how all this plays out.

 

[Antoine Baumann]

When I was puting up the diagram, I got a shock when I realized the MPs whouldn't cover the 30mm X 15mm S35 sensor. As you can see the corners will vignette, this is due to the 2:1 ratio of this sensor.
The widest you will get is around 4760px same as with the FF35 windowed.
So their dosen't seem to have any advantage going with a S35 sized sensor.

Emmanuel

Don't forget there will probably be some pixels left for "view around", so I guess, as on the redone, you will not use the full 30x15 to record but somewhere around 28x14.

I am Still for a 4 perf 35mm sized monstro sensor for epic.

ciaow,
antoine.

 

[Michael Lindsay]

...Pawel, I'm surprised that you are not planning to get FF35 with its increased DR? ...

I've only shot under water a couple of times (once with the Red) and the last thing I needed was less contrast!

I can understand why underwater guys need the sharpest, most contrasty and fastest lenses..

Red's 5000k native must work well with all that red light loss?

Pawel?

regards

Michael L

 

[Pawel Achtel]

I've only shot under water a couple of times (once with the Red) and the last thing I needed was less contrast!

I can understand why underwater guys need the sharpest, most contrasty and fastest lenses..

Red's 5000k native must work well with all that red light loss?

Pawel?

regards

Michael L

Yes, the native 5000K colour temperature is very nice.

 

[Daniel Browning]

FF35 creates all sorts of challenges and problems...Decreased DOF and difficulty in pulling focus

I don't think most filmmakers are dying to use larger sensors with the attendant depth of field challenges...

I would like to point out that the DOF challenges will not exist in daylight, when one can stop down and compensate with ND.

In artificial or low light, it will be a little more difficult, but not much. If the FF35 sensor has the same performance per area as the S35 sensor, but twice the area, then one may underexpose by over a stop and still collect the same amount of light (less intensity per area, but spread out over a larger area).

Increase gain in post and you'll end up with the same brightness, noise, etc. as the S35 image.

Typically, smaller sensors have better performance per area than larger ones. On top of that, underexposure is much harder on compression.

With those factors considered, I expect that FF35 will have only a 1/3-stop loss in control over DOF at the deep end (for a given image quality / noise level), and not the 1.3 stops that one would get with film (due to non-linear response preventing underexposure).

 

[Pawel Achtel]

So, what you are saying is, if suitable lenses were available, one could use say f/2.0 FF35 lens and achieve similar performance and DOF as f/1.2 S35 lens, except there will be more headroom for highlights, right? Interesting point.

The question is then, where do we get high-MTF FF35 cine lenses? The Red 300mm is certainly a good one. Not as sharp as MPs, but certainly sharp enough. I like it. Maybe Red can step in and provide bigger selection at wider angles? This would be a very interesting approach.

Can someone report on the image circle of the Red 18-85 Zoom, please? If it covers more than 38mm image circle at wide end, I think I would kill for one. :biggrin:

At the moment still lenses are very limited due to their inferior mechanics. Also, most wide angle still lenses have sharp falloff in MTF as you go off centre. Also, finding a wide still lens that doesn't flare is next to impossible. I guess the availability of glass will be the key.

 

[Daniel Browning]

So, what you are saying is, if suitable lenses were available, one could use say f/2.0 FF35 lens and achieve similar performance and DOF as f/1.2 S35 lens, except there will be more headroom for highlights, right? Interesting point.

Exactly right. Basically, what it all comes down to is the aperture. Not the f/number (which is "relative" aperture), but the the physical aperture, absolute opening, or "hole in space". The physical aperture of 32mm f/1.2 is 26.6mm. The physical aperture of 50mm f/2.0 is 25mm. They both provide the same FOV on S35 and FF35, respectively, and the DOF is the same because of the same physical aperture (~25mm).

This extends to *all* formats. For any given AOV, subject distance, and absolute aperture (not f/number) the DOF is the same for lenses on S35, FF35, Scarlet, and even 617. With larger formats, it's the same because of longer focal lengths and narrower f/numbers (ignoring bellows factor). For smaller formats it's the same because of shorter focal lengths and wider f/numbers. For example, with a 16.4-foot subject distance, all of the following camera/lens combinations will have the same 40 degree horizntal AOV and 6.7 feet DOF:

Scarlet 2/3":     14mm f/0.9
4/3" DSLR:        24mm f/1.6
Super35:          34mm f/2.2
Still FF35:       49mm f/3.2
645:              76mm f/5.0
617:              220mm f/14

They all have approximately 15mm physical aperture despite the 7-stop difference in f/numbers.

[At 16.4 foot subject distance, bellows factor is not very significant, even for the 617 format lens. I used h/CoC=1200 for all of the above to arrive at 6.7 feet DOF. The actual DOF will of course vary by the resolution, post-processing, and display environment, but I'm not trying to address those.]

Incidentally, all the lenses have same amount of light falling on the sensor as well (assuming the same transmissivity). Furthermore, the lenses even have the same issues with diffraction: if the sensors all had the same total resolution (e.g. 3K in 617), they would all experience the same amount of diffraction at the same physical aperture (not f/number). (RED is planning the larger ones to have more resolution, which is good of course.)

So physical aperture is a big factor in a lot of things. Many photographers tend to focus on "relative" aperture, whereas other fields (e.g. astronomy) have a greater connection to the physical aperture (which is useful for understanding things about different sensor sizes, especially).

The question is then, where do we get high-MTF FF35 cine lenses?

That's a key question, maybe even the biggest one, and there are undoubtedly many other challenges as well.

 

[Pawel Achtel]

That's an excellent point you've made, Daniel. Thanks for that!

 

[Pawel Achtel]

But wait, there is diffraction.

But wait, there is diffraction.

OK, let's take the consideration one step further. Let's consider diffraction limit of two lens systems from two extreme examples Daniel provided:

Scarlet 2/3": 14mm f/0.9
617: 220mm f/14

They both have approximately 15mm physical aperture and similar angle of view. Is the resolution affected equally by diffraction? Well, the answer is, suprisingly, NO!

The optical system with the larger sensor will be more affected by diffraction. Why?

This is because the distance to the focal plane also increases with focal length, and so the airy disk diverges more over this greater distance. As a result, the two effects of physical aperture size and focal length cancel out. The 617 will be affected by diffraction at f/14 (absolute 15mm) much more than the 2/3"at f/0.9 (also 15mm absolute aperture) .

This is also why diffraction limit of a short focal length lens is about the same as that of a long focal length for the same relative aperture (f stop), but not for the same physical aperture.

Now, not so clear cut once we consider diffraction limit :whistling: