Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

ultraprimes and focal length for 3D

From a conceptual standpoint, 3D is similar to trends in the 1950's like Cinerama, meant to create an immersive experience that resembles human perspective, and wider to medium focal lengths work best for that. Cinerama had a particularly wide-angle view.

I would guess that a lot of 3D work is in the 18mm-50mm range.
 
David is correct,

3D imagery is best captured with wide angles, however too wide of an angle on a Human subject will deeply exaggerate facial features,
as it did in Avatar, when Sam Worthington is recording the daily notes.

But the wider the focal the stronger the 3D effect is, especially beautiful in Underwater footage.

If I was to choose, I would start as wide as I could, but some Mirror Boxes have limitation on how wide you can go.

So far on ATOM, with Pedro and Tonaci, we have seen the 17mm to be right were it is at.
 
  • Thread starter
  • Banned
  • #4
Thank you David and Ketch.
I should be more specific in my question.
I am ordering a set of 5 lenses (ultraprimes) for my personal use (mostly 2d), but also having in mind to use them on other 3d rigs.
Out of two set I am getting a quote for, what is the focal length that is mostly used on 3d sets today (ultrprimes)
16mm
24mm
32mm
50mm
85mm

or

16mm
28mm
40mm
65mm
85mm

I want to stay compatible with preexisting most common setups.
As the wait time is long and I have everything in place, I need to make a decision fast.
Thank you again for help.
Albert
 
albert rudnicki said:
Thank you David and Ketch.
I should be more specific in my question.
I am ordering a set of 5 lenses (ultraprimes) for my personal use (mostly 2d), but also having in mind to use them on other 3d rigs.
Out of two set I am getting a quote for, what is the focal length that is mostly used on 3d sets today (ultrprimes)
16mm
24mm
32mm
50mm
85mm

or

16mm
28mm
40mm
65mm
85mm

I want to stay compatible with preexisting most common setups.
As the wait time is long and I have everything in place, I need to make a decision fast.
Thank you again for help.
Albert
Click to expand...

There are no specific focal lengths preferred for 3D, but the first set has a more useful range in my opinion -- 16mm to 28mm is too big a jump in the second set.
 
lens set

lens set

I just bought a (5) lens set of Ultra's just recently and spent some time on this very issue. Mr Mullen correctly points out a gap in one of your set options.

If you line up all the Ultra primes available and choose every other lens as a set, leaving out the extra wide and very long options, you end up with these two (5) lens sets.

(A set) 16mm, 24mm, 32mm, 50mm, 85mm (the more common 5 set)

(B set) 20mm, 28mm, 40mm, 65mm, 100mm

Now if you count the span between each lens this is the result

(A) 8, 8, 18, 35
(B) 8, 12, 25, 35

If the gap transition between each lens is important (yes it should be) then the clear winner between these two (5) lens set choices, is set B because the span between focal lengths is a much smoother progression. The A set replicates the 8mm focal difference twice then takes a larger step between the 32 and 50 to make up for the lag behind in what should be 12mm diff.

For this very reason we chose the B set to purchase. If we decide to add a 6th lens at some point it would be the 16mm, which actually fits the mathematical progression almost perfectly.
 
KETCH ROSSI said:
If I was to choose, I would start as wide as I could, but some Mirror Boxes have limitation on how wide you can go.

So far on ATOM, with Pedro and Tonaci, we have seen the 17mm to be right were it is at.
Click to expand...



what's the widest lens you can use with Epic 5K on a mirror rig? With some drawings I have been doing for a rig I'm designing, it seems like it has to have a vertical angle of view of 70 degrees or less or you would need a huge mirror - but I don't know what focal length that would be on 5K.
 
The question "what are the most common" depends entirely on the DP. I agree fully with David's assessment that wider is better but it depends on what you want to do with these lenses. Do you want to rent them to productions with DPs who are used to shooting 2D and are getting into 3D? Do you plan to be the only DP/stereographer using them? I cringe when focal lengths beyond 85mm go up, but sometimes that's what the DP wants and you have to make it work. Truthfully any focal length can be made to work in some way, just don't try to cut between a 24mm and a 135mm and expect those shots to match in 3D. They key is understanding the phenomenon of "cardboarding" in 3D and choosing when you want that and when you don't and to what degree. If it were me, I might go:
16mm
24mm
32mm
40mm
65mm
 
Thanks for all valuable suggestions.
The set will be mostly used for my regular 2d productions, thus I need to cover my standard f.lenght 16-85 (there is no 18 in ultras)
I may go 3d at one point, or may rent them for that purpose and simply want to be compatible with what's currently most popular configuration used.
I guess 50mm is more common then 65mm
32mm more common then 28mm and so...
I just want to make sure I am getting something that can be easily matched with other sets. I doubt I will ever buy another set:)

Probably one of the first jobs I want to try in 3d is live shows.
http://www.vimeo.com/7102148

Based on what I used 18-85 was perfect for more then 90%

thanks again.
 
i wouldn't assume your sets will match with other sets...

you could easily end up buying UPs that are very unususal in terms of specific focal lenght or focus scales.(there are obvously other factors which should be considered when matching lenses)..

If you can specify middle letter focus scales that might help a little but definitly don't assume.

good luck

Michael
 
Michael Lindsay said:
i wouldn't assume your sets will match with other sets...

you could easily end up buying UPs that are very unususal in terms of specific focal lenght or focus scales.(there are obvously other factors which should be considered when matching lenses)..

If you can specify middle letter focus scales that might help a little but definitly don't assume.

good luck

Michael
Click to expand...

Michael
Do you say they have to be factory matched in order to work well on the 3d rigs?
 
albert rudnicki said:
Michael
Do you say they have to be factory matched in order to work well on the 3d rigs?
Click to expand...

I think it depends on the project whether they "have" to be matched but any variation makes things more difficult in 3d. From a post in another thread about one of the 3d majors in production it spoke of having their lens matched by the tech at the rental house so it appears you can adjust them to match sufficiently but probably not feasible for random rentals. Can I offer an alternative? I can probably get you 2 sets of illuminaS35s for less than a set of ultras and they are very modern new design, Schott glass, lighter, and smaller which is significant with mirror rigs AND they are T1.3! Contact me off list for more info.

We will be testing these this week on ketch's Epic-M.
 
David Collard said:
Albert,
Save your money.
Check out: www.uniqoptics.net
Click to expand...

David
I did consider a purchase of Uniqs...
They are great, I am sure, yet the size&weight is where they fall short on my list.
Same with RPs
Another problem is that those are new lenses not widely recognized and available.
As an investment and resale value zeiss will stand ground for years to come

I know you own a set and Uniqs, would love to take a look at them.
Best

Albert
 
Michael Hastings said:
I think it depends on the project whether they "have" to be matched but any variation makes things more difficult in 3d. From a post in another thread about one of the 3d majors in production it spoke of having their lens matched by the tech at the rental house so it appears you can adjust them to match sufficiently but probably not feasible for random rentals. Can I offer an alternative? I can probably get you 2 sets of illuminaS35s for less than a set of ultras and they are very modern new design, Schott glass, lighter, and smaller which is significant with mirror rigs AND they are T1.3! Contact me off list for more info.

We will be testing these this week on ketch's Epic-M.
Click to expand...

Do you have the updated version?
I send you an email.
thanks
 
Actually the choice is not that of the DP alone, as it is the Stereographer that must have a said, as DP's might not have the
knowledge of what effect will be achieved by a said lens.

In 3D, the wider Focal, creates deeper and a more exagerated 3D effect, so it is not just a matter of field of view as in standard
monoscopic 2D, personally I would never allow the DPO to make that decision, that would be my decision as the Director, after
consulting with the Stereographer, which in this case that would also be me, he he... ;)

Widest on ET's Mirror Box on Atom and Epic 5k is right in the middle of 16/17mm measured with both the Angenioux and the Ruby
with Pedro and Tonaci at ET few days ago.

All glas must be matched, as any small lens disparity in 3D is exagerated, and so makes things more difficult in Post,
so matching glass is a must, on set calibration a given.
 
KETCH ROSSI said:
Actually the choice is not that of the DP alone, as it is the Stereographer that must have a said, as DP's might not have the
knowledge of what effect will be achieved by a said lens.

In 3D, the wider Focal, creates deeper and a more exagerated 3D effect, so it is not just a matter of field of view as in standard
monoscopic 2D, personally I would never allow the DPO to make that decision, that would be my decision as the Director, after
consulting with the Stereographer, which in this case that would also be me, he he... ;)
Click to expand...

Ketch, I'd have to say that as a DP, I would never work with a director who shut the DP out of discussions on what lens to shoot a scene with! Doesn't matter if it's a 3D movie or not. That's like shutting the costumer out of involvement in what fabric to sew a costume out of. DP means "Director of Photography" and if lenses don't fall under the category of photography, then nothing does. What's next, are you going to work directly with the Gaffer and shut the DP out of lighting decisions as well?
 
Last edited:
Albert,

What cameras do you think you will be using for s-3d?

I design a lot of different stereoscopic systems and the first thing I consider is to match the focal length to the sensor size.


Roughly speaking my own very rough rule of thumb to get good “rounded” stereo (but not go over the top) is to use a focal length that is near to or a bit smaller than the width of the sensor. If distortion is well controlled by whatever means then for really quite immersive stereo you can go to a focal length that matches the height of the sensor, but generally not smaller than. So I like to work with one or two focal lengths in between the two dimensions. Focal lengths about 20% longer than the sensor width start to look less rounded and start heading in to “cardboard town” (I think, but everybody else has a different way of working).


The cameras I am building and designing around are precision metric focusable instruments. In such a case I tend to steer clear of retro-focal lenses and try to go for a lens that is of a symmetric or Biogon type (just so radial lens distortion is MUCH lower and also more accurate to calibrate and correct for), as in (our digital systems) we don’t need that retro focal throw as there are no mirrors or prisms in between and we don’t need to physically “look” through the lens. The problem is that Biogon type lenses are not that fast compared to cinematic primes. I would really like to see FAST symmetric lenses being specially built for s-3d.


Out of interest does anyone have their hands on distortion curve data for the Ultra-Primes?


AND does anyone have actual experience verifying that the Helicoids in the Ultra primes are truly zero back-lash? Just wondered what the repeatability was in microns or whether they tend to internally mismatch and slide a bit between two lenses when focusings adjustment are made?

Cheers,

Eric
 
You must log in or register to reply here.
Back
Top