Is the future FF35?

[Roberto Lequeux]

The glass will be there if people start buying FF35 Epics with a set of Red FF35 lenses and it sounds like a helluva lot of you are planning to do so. The competition won't be happy sitting on the sidelines, so soon enough thereafter there will be more than just the Zeiss Compacts and the Red FF35 glass.

Sure, a lot of those owners would use their FF35 Epics in S35 mode and be more than happy with their investment, given the higher bit-rate and DR you'd get from it. But what is going to be there to stop you guys from shooting as big as possible when the sun is up?

I think the question is really just about how much of a pain it will be to shoot FF35 at night, or inside. Wouldn't you agree?

And the big question regarding the camera specs would is then how much more sensitive that FF35 sensor will be. Right?

I asked before and someone was kind enough to explain, but I forgot already... Sorry. Could someone please tell me how much more sensitive the sensor would need to be to approximate the lighting and focus pulling that we are used to now with S35 if we began shooting FF35?

 

[Patrick Scheller]

I'm coming from still photography and used to shoot with Pentax 645 and Hasselblad (yes 6cm x 6cm on the negative!) and I got addicted to those big formats.

Most of you don't have much experience with those big formats as I see. Numbers of resolution and DR and DOF are relevant, I totally agree but what blows your mind (like mentioned before from Andrew Walker on post #18) is the impression of the images in a bigger format! There are no numbers to express that. Pictures just feel more brilliant, more palpable, more sharp and you can feel the depth of a room much better which is NOT due to the DOF!

Just have a look at that pic from Buzz Aldrin taken 40 years ago with a Hasselblad on the moon and maybe then you can understand what I'm talking about!
And remember that technology of lenses, coating and film stock was not nearly as good as it is today!!
:emote_hippie:

http://upload.wikimedia.org/wikiped...EdwinAldrin_big.jpg/601px-EdwinAldrin_big.jpg

 

[david farland]

Could someone please tell me how much more sensitive the sensor would need to be to approximate the lighting and focus pulling that we are used to now with S35 if we began shooting FF35?

Here....

Depending on your Coc for s35/FF35 lenses, say the FF35 sensor is 1 stop faster.
So the equations you're likely to have are:
S35 (Coc=.017) x 50mm x f4 x 4 metres = .88m DOF
FF35 (Coc=.025) x 75mm x f5.6 x 4 metres = .80 DOF

~1 stop if you are comparing similar setups/frame ratios.

Use this for comparison...http://johndudak.com/photographers/dof.php

Dave

ps: nice Apollo pic..

 

[Scott C.]

The glass will be there.... I think the question is really just about how much of a pain it will be to shoot FF35 at night, or inside. Wouldn't you agree?

I don't. If you agree that the glass will be available then what is the problem?

 

[Roberto Lequeux]

Well in my ignorance I was expecting it to be far more than just 1 stop. So I was expecting focus pulling to be much more of an issue. If I am not mistaken then 640ISO is what would be necessary to counter that. So maybe we'll be about half way there if we get ISO500?

In that case, if I understand correctly, if Epic FF35 were rated at ISO500 that would mean that focus pulling would be close to what it has been with Red One with a little more light. And with a significant amount of additional exposure you could close that gap.

Maybe we'll be even more surprised... Could Epic FF35 with the larger photosites give us clean 640ISO or is that too much to expect by any measure?

 

[Scott C.]

Well in my ignorance I was expecting it to be far more than just 1 stop. So I was expecting focus pulling to be much more of an issue. If I am not mistaken then 640ISO is what would be necessary to counter that. So maybe we'll be about half way there if we get ISO500?

I was thinking that way too initially but it is more simple than that once you concede that the optics will be available.

Consider a Red One with a 50mm RPP shooting 10 ft. to subject, exposing properly at f5. Dof is just over a half foot (.542 ft.), no? Now imagine a sensor of the same sensitivity at FF size with a Canon 50mm 2.8L... You could shoot that at f3.5 for approximately the same DoF (.55 ft.).

Now, if I am understanding this correctly, that is a 1.5 stop advantage right there. If you then assume the Monstro will be 1.5x more sensitive, then we are talking about a 3 stop advantage over the Red One.

Apply that to our scenario and you could shoot the Monstro at the same distance, with the same lens at f6.3 and have just over 1 ft DoF. What is harder about that?

Am I getting this wrong?

 

[David Mullen ASC]

Yes and no - it's a 1.5-stop difference in DOF to compensate for, but if the Red One is 320 ASA and the Monstro is 500 ASA... then we're just back to the sensitivity / speed that most people are used to lighting to, 500 ASA, so they will still have to learn to light up to another 1.5-stops and shoot at an f/4-5.6 split rather than an f/2.8 if they want to keep the same DOF of shooting Super-35 at f/2.8. Sure, if they've been used to lighting for the Red all this time, and the Monstro is more sensitive, then it's not so hard... but honestly, a lot of modern cinematography is built around lighting interiors to f/2.8 so it will be hard to break people of that habit.

They'll have to think of it like shooting 35mm anamorphic where the old rule is to light to an f/4 whenever possible. At least FF35 at f/4 is not insanely shallow-focus.

But it does speak to the issue of whether FF35 lenses need to be faster than f/2.8 except for low-light situations.

 

[david farland]

David,
plugging in the figures I can't get 1.5 stops anymore. What am I doing wrong?

S35 (Coc=.017) x 50mm lens x f4 x 4 metres = .88m DOF
FF35 (Coc=.025) x 75mm lens x f5.6 x 4 metres = .80 DOF

Dave,

 

[Daniel Browning]

...the larger pixels in the Epic FF35 system should (hasn't actually been built!) have a higher dynamic range....and to me that's better.

I explained it all in the link. Larger pixels do not have higher dynamic range. The LX3 has two micron pixels, for example, with over a stop more dynamic range than the 5D2's 6.4 micron pixels (10 times more area).

I think the question is really just about how much of a pain it will be to shoot FF35 at night, or inside. Wouldn't you agree?

As explained elsewhere in the thread, thin DOF will not cause any additional pain because it will have the same noise level even when stopped down to get matching DOF. But if you're not going to use it in ample light or thinner DOF, you might as well save money and buy a smaller sensor.

And the big question regarding the camera specs would is then how much more sensitive that FF35 sensor will be. Right?

Kind of. The sensor performance is one of the questions. But I would argue that the amount of light is more important. A FF35 sensor with old technology (e.g. 5D1) can still easily beat a much newer sensor with more powerful technology (e.g. 50D). If technology is similar, of course, then they're not even close. Size matters. Bigger size = more light.

I asked before and someone was kind enough to explain, but I forgot already... Sorry. Could someone please tell me how much more sensitive the sensor would need to be to approximate the lighting and focus pulling that we are used to now with S35 if we began shooting FF35?

Simply divide the horizontal width of the part of the sensors you will use. If you will use only 24x13.5 from the S35 and 36x20.25 from the FF35, then the crop factor is 1.5X (36/24). (Using RED ONE recordable area gives 1.6X.)

On the other hand if you do the full S35 30x15 (including lookaround) compared to 36x18, the crop factor is only 1.2X.

To get the equivalent ISO, simply multiply (or divide) by the crop factor squared. For example, using 1.5X, ISO 1000 on FF35 needs to have the same noise as ISO 444 (1000/1.5^2) on S35. Using 1.2X, if ISO 320 on S35 has a certain amount of noise, FF35 will need to match it at ISO 460 in order to be capable of the same deep DOF.

Keep in mind that the ISO "rating" of the sensor has nothing to do with what we're talking. The rating is just the number that the manufacturer recommends for optimal results. Not for getting the *same* results with all cameras. I would guess that RED will recommend low-ish ISO ratings (500 or less) for all its cameras, no matter the size. For the discussion of DOF, we're talking about what ISO rating will be needed to get the same DOF among different cameras, which is different.

Maybe we'll be even more surprised... Could Epic FF35 with the larger photosites give us clean 640ISO or is that too much to expect by any measure?

I'm sorry for repeating myself, but it has nothing to do with the size of the pixels.

Consider a Red One with a 50mm RPP shooting 10 ft. to subject, exposing properly at f5. Dof is just over a half foot (.542 ft.), no? Now imagine a sensor of the same sensitivity at FF size with a Canon 50mm 2.8L... You could shoot that at f3.5 for approximately the same DoF (.55 ft.).

Now, if I am understanding this correctly, that is a 1.5 stop advantage right there. If you then assume the Monstro will be 1.5x more sensitive, then we are talking about a 3 stop advantage over the Red One.

Precisely. The sensor only needs to have the *same* sensitivity to get the same DOF (and same noise).

We *know* FF35 will be at least a stop better just based on size alone. Just as we know a 1000cc engine is very likely more powerful than a 500cc engine based on size alone. Technology will of course play a big part, but size (i.e. the amount of light) should form our base expectation.

plugging in the figures I can get 1.5 stops anymore. What am I doing wrong?

What's a third-stop between friends? CoC is based on sensor area (due to reproduction magnification), so you are assuming a different sensor size than David. Try comparing RED ONE recordable area (22.2x12.6) with FF35 (36x20.25), both 16:9.

 

[david farland]

I explained it all in the link.

Daniel, that link is too full of conjecture, half explanations and unanswered counter explanations and just very confusing to follow. I suggest you finish/clarify your points there, not here!
Dave,

 

[Daniel Browning]

Daniel, that link is too full of conjecture, half explanations and unanswered counter explanations and just very confusing to follow. I suggest you finish/clarify your points there, not here!
Dave,

Thanks for the feedback. EDIT: Can you be more specific?

How about this one?

http://photography-on-the.net/forum/showthread.php?t=706255

 

[Michael Hastings]

I don't. If you agree that the glass will be available then what is the problem?

If you agree that the glass will be available there is no problem, and of course the glass is already here in the SLR world from Canon, NIkon etc. I don't have a crystal ball but I do have 25 years of experience in and around the manufacturing of niche products and a bit of understanding of what determines the development cycle.

My post was just a suggestion that we may have to wait a very long time for FF35 lenses in true cinema style (i.e consistent t-stop, long throw focus gearing, consistent fronts, etc) and the trend of sensor technologies suggest that it may be never be worthwhile for a manufacturer to develop.

The only FF35 cinema lenses even announced as in development are by RED. Jim certainly has the financial resources to make it happen, but there is no timeline and since he is a practical guy he MAY decide (for the reasons I mentioned) that it doesn't make all that much sense.

Consider again that the RED Pro primes are basically a variation of known and existing designs i.e. cinema-style S35 lenses from Arri/Zeiss, and Cooke - with some compromises combined with higher volume production to get a better price point.

Full frame cine style lenses are an order of magnitude more difficult. When you add in the announced autofocus and image stabilization to compete in the Still world it becomes a huge deal. Just combining autofocus motors with the ability to go manual with the kind of long focus throw [LFT] and buttery smooth mechanics [BSM] of a cine lens (Hence to be known as an LFTBSM lens) is a very significant challenge that AFAIK has not been done before so expect a significant extended development time no matter how much money is thrown at it.

Add in Image stabilization - again AFAIK never done on an LFTBSM lens - and you add in another order of magnitude of difficulty, money and time for development.

Of course the sweet spot of developing that type of lens would be to use fly-by-wire focusing (i.e. the focus ring directs the motors) rather than gearing, as is done in many SLR lenses and prosumer video cameras. But if they go that route I think they will meet resistance from the highend DPs and/or simply an acknowledgment that users might as well just shift to a smart mount style (like the Birger EF mount) combined with off the shelf Canon and Nikon highend FF35 lenses.

I'm not suggesting that FF35 isn't the most desirable Epic (if it - or the 645 were available at the same time as the Epic X with the same accessory package I would probably spend the extra and get it as most of my shooting works fine with SLR lenses.) But I am suggesting that the lens world is likely to be stuck in S35 cine or adapted SLR lenses for quite a long time.

 

[Scott C.]

If you agree that the glass will be available there is no problem, and of course the glass is already here in the SLR world from Canon, NIkon etc.

Therein lies the $200,000 question: Will Canon L Series lenses, shot at 6k and down-sampled to 2k yield an image sharp enough and clean enough to be acceptable as a primary tool in a theatrical distribution chain? If it is, then I won't have any problem driving the Hyundai of cine lenses to work instead of some LFTBSM Mercedes.

The idea that an individual might be able to get a rig that is capable of producing theater worthy images for less than the cost of his house is mind blowing to me. It speaks to the kid in me who was inspired by Star Wars all of those years ago and dreamt he was making features when in reality he was animating Legos on 8mm. :tongue_smilie:

My mid-life crisis has a name. It is Red. My wife is just happy that it isn't Harley Davidson.

 

[Michael Hastings]

Therein lies the $200,000 question: Will Canon L Series lenses, shot at 6k and down-sampled to 2k yield an image sharp enough and clean enough to be acceptable as a primary tool in a theatrical distribution chain? If it is, then I won't have any problem driving the Hyundai of cine lenses to work instead of some LFTBSM Mercedes.

I believe Canon L series lenses, at any full frame resolution, are more than acceptable without down rezzing. Optically, (i.e. sharpness and cleanness) the current generation of Nikon and Canon lenses are as good, better, or at worst so close as to be indistinguishable from high end cine glass in a theatrical distribution chain. The only thing even a pro could likely notice visually MIGHT be breathing in certain shots with certain lenses.

I'm with you - I have no doubt that very high quality results can be had (via electronic control) without the LFTBSM and we will see examples over the next couple years. I suspect big budget pictures will stay with the LFTBSM - and why not since they can afford it and they more than pay for themselves if even a few retakes are avoided - when you are possibly paying hundreds of thousands per day for sets, cast, and crew.

Horses for courses...

 

[Evangelos Achillopoulos]

I'm thinging that I will go for Epic FF35...

But recently I have finished my SGblade-2/3 relay set up...

The DoF adapter is set up with my Viper and its on FF35...

What we realize is that, we need to step down to F5,6 to F8 in the lens to get some decent DoF larger than two cm... the DoF adapter is losing 1 stop... so to work with Viper's 320ASA/2 = 160ASA... we need to put huge amount of light to work with the FF35 apertures needed...

The EPIC FF35 has to have more than 800ASA normal sensitivity to allow us to work at least decent with FF35... I believe we will work always in the cropped mode of S35 most of the time...

You see DSLR's are not stacked to the 24Fps frame rate they can use 1600ASA noiseless... and have speeds of 1/10 of a second with no problem... you see they are not a motion picture cameras... they grabbing stills...

Its understood that I'm turning back to S35 my SGblade...

 

[david farland]

couple of things.....

- The FF35 design issue seems to come up at around 35mm and wider (equivalent to 25mm FOV).
Need to know why designer can happily produce still lenses under this and not cine grade lenses.

- Evangelos, your issue seems to arise when you use an 'effective' sensor speed of 80 asa (with adapter) whereas we're hoping the Epic FF35 will be around 450/500.

Thanks for the feedback.
EDIT: Can you be more specific? How about this one?
http://photography-on-the.net/forum/showthread.php?t=706255

I flicked through your link (and sub links) and noted there were over 50 pages of information plus literally hundreds of posts perhaps warranting further reading......phew Dan!
Pascal once apologised for writing a long letter as he didn't have the time to write a short one.
Perhaps if you feel so strongly about the topic and have spent so much time researching it you could write an original (& readable) paper on the topic that doesn't have some many overlapping topics which requires qualification or elimination.....hence all the conjecture!
Dave

 

[Daniel Browning]

Perhaps if you feel so strongly about the topic and have spent so much time researching it you could write an original (& readable) paper on the topic that doesn't have some many overlapping topics which requires qualification or elimination.....hence all the conjecture!

Thanks again for the feedback. In the future I'll try to improve the readability and replace the many links with original material.

 

[Shane Betts]

OK, drawing a target on my forehead. Fire at will. My theory is as follows.

Quality stills lenses are sharp enough, the issues are firstly mechanical and secondly about performance WFO. As sensors become faster (refer Canon 5DMkII) and with the shorter DoF with the bigger sensor, we should be shooting stopped down a little anyway, reducing the second problem. So how do they address the mechanical issues? I don't see stills manufacturers duplicating their lens range with cine-style lenses. And the cost and size of housing FF35 stills glass in cine alloy would be nuts. What I believe will happen is that someone like Red will solve this in software with cine-specific AF, making stills lenses work in cine applications. I know, I know, AF is bad for motion work. But, if someone smart (read, the Red team) actually put their mind to writing something programmable and designed for actual movie production (as distinct from shooting home movies), I have total faith that the current crop of stills lenses would work a treat. So we get bigger sensors and smaller cameras and lenses. That's what I call a win/win.

 

[Roberto B]

Is the future ff35? In a word: yes.

It seems very clear to me that cinema is moving to ff35. As we have been discussing here for more than a year, the one major issue is lenses. But that will be addressed by market forces. In the meantime , s35 lenses can be used on monstro ff35 once it comes out. For future glass purchases, though, it would be wise to keep an eye toward full frame options.

in a word: convergence.

- the gang

 

[Tom Lowe]

I'm coming from still photography and used to shoot with Pentax 645 and Hasselblad (yes 6cm x 6cm on the negative!) and I got addicted to those big formats.

Most of you don't have much experience with those big formats as I see. Numbers of resolution and DR and DOF are relevant, I totally agree but what blows your mind (like mentioned before from Andrew Walker on post #18) is the impression of the images in a bigger format! There are no numbers to express that. Pictures just feel more brilliant, more palpable, more sharp and you can feel the depth of a room much better which is NOT due to the DOF!

Just have a look at that pic from Buzz Aldrin taken 40 years ago with a Hasselblad on the moon and maybe then you can understand what I'm talking about!
And remember that technology of lenses, coating and film stock was not nearly as good as it is today!!
:emote_hippie:

http://upload.wikimedia.org/wikiped...EdwinAldrin_big.jpg/601px-EdwinAldrin_big.jpg

Great post, Patrick! You nailed it.

Therein lies the $200,000 question: Will Canon L Series lenses, shot at 6k and down-sampled to 2k yield an image sharp enough and clean enough to be acceptable as a primary tool in a theatrical distribution chain?

Well, I recently shot some 5.6K RAW 24p timelapse footage using several 5D Mark II bodies. One body had a Panavision 50mm anamorphic Primo, and the other two bodies had Canon EF 24mm and 50mm L-series lenses. Believe it or not, the $1,500 Canon still lenses easily beat the legendary Primo in terms of sharpness and overall image quality. And I've got the photos to prove it. :ihih:

Actually, the Canon EF 50mm f/1.2 was the real champion of this shoot. That lens is freekin amazing!

I think I come at this discussion from a different perspective than many. For the past 6 months, I have been busy shooting 5.6K RAW timelapse footage on a full-frame camera - the 5D2. And I can tell you right now that the results are absolutely astonishing. Completely spectacular. The quality of the images I am finishing at 4K are beyond mind-blowing. This is why there is absolutely no doubt whatsoever in my mind that FF35 RAW is the way of the future for cinema. It can be no other way! :ciappa: