Zeiss Supreme Prime series

[Björn Benckert]

http://www.fdtimes.com/pdfs/articles/zeiss/88FDTimes-ZEISS-Supremes-Report.pdf

According to Christian Bannert ,Vice President, Development & Manufactur-
ing, Consumer Products at ZEISS, when asked in a 44 page interview in Fdtimes. Page 11

Most ZEISS Cine lenses were done with ARRI. Why did you go
out on your own with the Supremes?

"ARRI and ZEISS both stand for high quality and reliability. Both
companies have been working very well and successfully together
for over 80 years.
As part of a joint effort with ZEISS in terms of design and devel-
opment, ARRI launched the Master Anamorphic, Master Prime
and Ultra Prime lens families. We received an excellent response
from our customers. We are delighted about this success. ARRI
will continue to offer these products and the service that goes
with them.
The partnership is not exclusive: ARRI and ZEISS have both pur-
sued their own projects concurrently. ARRI teamed up with other
partners to develop the Alura Zooms and the ARRI Signature
Primes. ZEISS launched the Compact Primes, Cinema Zooms
and the Lightweight Zoom on its own. The ZEISS Supreme Primes
are another lens range that ZEISS is pursuing independently.
ZEISS and ARRI will, of course, continue to pursue joint projects
together.
As a result, now ARRI have their Signature Primes. And we are
introducing our Supreme Primes. But let me repeat: we will still
work together with ARRI. Maybe in the future we’ll do another
new project together again. It really depends on the market"

That last line

"Maybe in the future we’ll do another
new project together again. It really depends on the market"

That sounds to me like, not eny time soon...

 

[rand thompson]

Bjorn,

I think you're right about that.

 

[rand thompson]

 

[Bob Gundu]

Wow, thats really bad lighting, and framing.

 

[David J. Buchanan]

I doubt Arri and Zeiss will just stop working together. You never know what the future holds. With that said, I'm surprised that they didnt go all out on the Supremes? Its not like cost is a thing at this kind of bracket.

 

[Josh Becker]

Wow, thats really bad lighting, and framing.

and jumping the eye line is a bit distracting... seems like a waste to be watching a talking head in a video about lenses!

 

[Aaron Lochert]

and jumping the eye line is a bit distracting... seems like a waste to be watching a talking head in a video about lenses!

Color is off between the angles as well. Skin becomes monochromatically orange on the close up -- he looks sickly.

 

[John David Pope]

Going by the photos of the Zeiss Supreme primes I thought they would be heavier than the Sigmas. But several of them are actually lighter in weight than the equivalent Sigma Cine prime focal lengths.

 

[rand thompson]

They Remind me of the Leica Summicron-C in relative size.

 

[Michael Lindsay]

In my hand they feel very similar in weight to the sumicron-c... a little more bulbous but with the addition of FF and lens encoders they are impressively manageable. As someone who once half dropped and miraculously caught (a foot from the ground) a MP I really appreciate one hand capable lenses..

 

[rand thompson]

Michael,

Great in size for a Red DSMC2 camera body. I wish leica would have chosen a similar approach with Thalia lenses, more of a compact summilux-c than a modified Leica-s lens formula.

 

[rand thompson]

Here is another example from Reduser member, John Marchant, of KipperTie called "Reliquaie" that he posted a link to earlier in this thread. Monstro VV with Zeiss Supreme Primes

 

[Michael Lindsay]

Michael,

Great in size for a Red DSMC2 camera body. I wish leica would have chosen a similar approach with Thalia lenses, more of a compact summilux-c than a modified Leica-s lens formula.

In a way Leica restarted the light small lens trend... sumilux-c lenses are just incredible.... but they are too hard to make and their current cost reflects that. Leica switched to Japan to make the sumicrons and they are great but I believe the Thalia's are made in Germany with mechanics more similar to sumilux-c (I may be wrong as I haven't peaked inside)... they don't have the unified 6ft to infinity thing but they are still impressive. I suspect if they built cine lenses from the ground up for medium format the cost would have been problematic for the market.

 

[rand thompson]

In a way Leica restarted the light small lens trend... sumilux-c lenses are just incredible.... but they are too hard to make and their current cost reflects that. Leica switched to Japan to make the sumicrons and they are great but I believe the Thalia's are made in Germany with mechanics more similar to sumilux-c (I may be wrong as I haven't peaked inside)... they don't have the unified 6ft to infinity thing but they are still impressive. I suspect if they built cine lenses from the ground up for medium format the cost would have been problematic for the market.

Yeah michael I hear you and what you said makes a lot of sense.

 

[rand thompson]

Serguei Saldivar, The Director and Cinematographer of "El Viaje", made available a 4K UHD version for download on Vimeo of Zeiss Supreme Prime lenses with the Red Monstro 8K VV camera.

 

[Ryan Patrick O'Hara]

Got to get some hands on experience with new Zeiss Supreme Primes at the ASC Clubhouse event yesterday. My initial thoughts: Zeiss went hard to hit some serious points while slightly neglecting others for weight and price point reasons.

Here is the TLDR: Supreme Primes are great size/weight you are used to. Image is neutral, nice, and when used on full frame sensor, they shine. Would probably not pick them for non-FF use though although they would perform just fine... it's just that there are too many great options in S35 world I already love and which look a little more like what I like out of the box... which is not typically neutral. Unless someone can make a PL-PL speed booster and then I'll reconsider .....

Zeiss made lots of great focal lengths to pick from, and you can't complain about aperture speed. Zeiss knocked focus breathing out of the park, but pinched their budget with non-linear iris assembly, non-cam driven manipulated focus scales, and presumably will allow a significant more amount of uneven field illumination when at wide open speeds on wide lenses than historically allowed for spec in s35mm terms. I walked away feeling like Zeiss has made a good new set of lenses that are better than Ultra Primes, but these are not going to be the 'Master Primes' of full frame. If anything these felt like much superior continuation of the CP.3's. Like what CP.4's would be, but instead of 2022, they are available today. haha.

Weight and Size: They are normal cinema lens sized. Heavier than Ultra Primes, but reasonably close in size. About the weight of Cooke S4's or Leica Summilux-C.... so these aren't SUPER light like CP's or Ultra Primes, but they are reasonably normal, unlike Cooke S7's or Master Primes which are heavier.

Mounts: These are designed for 52mm flange depth/backfocus. So although they will be offered in LPL and EF mount, these were not designed for LPL mount, so there is no benefit to using LPL. My personal opinion, but i'm not sure why anyone would order these in LPL because the Arri LPL to PL mount adapter is super solid and having the lenses in PL would make them backwards compatible to the countless PL cameras for the foreseeable future. All mount options have Cooke i/technology as well as the Xtended Data Zeiss is offering, which is distortion and uneven field illumination mapping. I've always been a huge i/data fan, and the Xtended data is amazing. I believe other i/data lens manufacturers are adapting XD as well?

Image. From what I could tell the lenses looked good. They don't ooze character, but I didn't expect them to. Zeiss claims the lenses are neutral and versatile, so don't look for too much of a 'look' in the glass except for clean nice shallow renderings. Zeiss reps claim they match the full frame CP-Zoom line rather well, but not 100% perfect. The lenses looked clean, pleasing, but nothing stood out as being unique regarding image, but I think that was the goal. Neutral, pleasing, but not clinically sharp. There was lots of talk about the focus fall off being a more gradual one, but honestly, I'd love to see some factual tests to see where the rubber meets the road. When looking at t/1.5 full frame shallow DoF, I can't tell with my eye if the focus fall off is any more gradual or not. For one that's just hard to see, second, I'm used to s35mm or slightly larger sensors, not Full Frame, so the comparison of DoF is just not juxtaposed easily in my mind. Is there a way to measure DOF fall off? Cooke has always been rumored to have more, and Zeiss is boasting that with these Supreme Speeds, but I'd love to see some sort of demonstration or comparison or measurable test.

One of the areas I think Zeiss has decided not to pursue in order to curb the price point and size of the lenses is in uneven field illumination. Granted the longer lenses showed very little, which is typical of lens design, however I did get to test the 25mm on a Red Monstro VV 8k Camera. I took the very center of the frame and I opened the lens to a T/1.5 and turned on false color. I changed my shutter and ISO until the very center of the camera just registered a red color as to indicate it was 100% blown out. I then panned/tilted the camera to put that exact spot in the extreme corner. The red, orange and yellow zones had all been eliminated due to the darker corner of the illumination field. Using ISO, I cranked up the exposure just until the red returned. I have to admit the Monstro 8K VV is larger than the other FF formats mentioned previously in attendance AND is not a real format aspect ratio (not 2.39:1, 1.85:1 or 16:9), so at the full 8k, it's punishing on image circles. Anyway, at the very most outer part of the edge, it looked to be a little over two stops of light loss from center to corner. If I were to estimate 5% safety or maybe 10% safety for a 16:9 image or a size roughly closer to the other FF formats from other manufacturers, the light loss was somewhere a little more than 1.5 stops.

I was expecting this because this is what most high end photography lenses do the same when they are full format coverage, wide focal length, and a fast aperture. You can design around this, but in my understanding the lens would be considerably larger in size, heavier in weight and more expensive. I think nearly two stops from center to corner is not very cinema-lens-esque in spec, but this would not be the first time a cinema lens does perform this way (looking at you Angie DP Rouge 16-42 among many others.) Maybe in an age or RAW capture and Zeiss's Xtended Data, the uneven field illumination mapping can be eliminated in real time on set, easily eliminated in post, and without artifacts. If so, perhaps the future of giving a damn about uneven field illumination will be dead and cinema lens designs can focus on other issues. Either way, I think cinematographers in the past have demanded a certain amount of even field illumination from corner to center, and Zeiss has felt that with full frame and high speed, they can let it go for the sake of size and cost.... while giving XD data to eliminate it easily in post. Found that interesting and predictable when I first looked at lens specs.

For those who do not know, a linear iris has an iris ring where every stop is evenly spaced. The distance from T/2-T/2.8 is the same as the distance you turn the ring from T/4-T/5.6... although this sounds 'normal' it's a more difficult and expensive implementation. A non-linear iris has a collapsing scale, like focus, where the wide open part of the iris needs more turning than the minimum aperture part of the iris. Zeiss, in what looks to be a clear effort to either use existing aperture assemblies from CP.3 lenses (both have T/1.5 14-blade iris assemblies... so maybe) or simply keep costs down on a new 14-blade iris assembly, have chosen non-linear iris scales. The person making the presentation tried to actually make that a positive by rationalizing that since most people use the more open side of the scale, the non-linear aspect of the scale allowed for more barrel rotation in those areas in which one could move the ring. I won't quote him but it was something along the lines of having more space to move around in that area of the scale. It was just a bizarre attempt to try to rationalize and sell a crowd on the cheaper, less preferred scaling. This was at the ASC clubhouse though and I'm not sure most DP's there know or care about the iris scale being linear or non, and as an AC/DP I can tell you this is really not a big deal, although I prefer linear without a doubt. There is no benefit to non linear except it's cheaper to implement into the lens. There are no in-the-field advantages, no matter what Zeiss sells you.

Mechanics of these lenses felt really good. As expected from a Zeiss Cinema lens, these felt well built. Super solid feeling, iris and focus rings turned very smooth. The focus ring felt smoother than Ultra Prime (always found those too dry and sticky) and CP.2's, but not as liquid smooth as those amazeballs MP's.

Another area it seems Zeiss decided to not focus their efforts on, in order to probably simplify and keep costs down, is the focus mechanics and the resulting scales. I didn't get a complete answer on how the focus mechanics are designed but from my observation it looks like the scales are not utilizing heavily manipulated cam driven mechanics, so some of the focus scales are not spaced as ideally for an AC as they could be. Don't get me wrong, they are completely usable, but unlike Leica Summilux C's and Cooke lenses which are highly manipulated, there is some spacing that could be desired on the Supreme's, especially because they are T/1.5 and FULL FRAME, lol. HOWEVER, let me say that these days, I almost only work with Preston, ArrI or Axis-1 wireless follow focuses. Preston and Arri allow for digital programing of lenses and then you can digitally remap the focus scale to an assigned and manipulated scale ring. What this means is that if using a WCU-4 or Preston digital hand unit, you can interpolate the helical (non-cam manipulated scale) to be a better scale. So just like with XD data and uneven field illumination, technology might be able to make these lenses excel in places they weren't optically or opto-mechanically designed to. An interesting world we live in. As an example, the 50mm has six witness mark distances from Infinity to 10'.... a Leica Summilux C 50mm has twelve, I believe and from memory those are in a little longer rotation. Point is a focus puller without a wireless follow focus that can remap a lens would have better time on a Summilux-C for the reasons of cam manipulated scales. Not a huge deal though. The scales aren't bad on Supreme's. Far worse have been passed off as cinema lens scales (remember Red Pro Primes?! RPP 50mm had like three marks from infinity to 10' in such a tight rotation)

 

[David J. Buchanan]

Got to get some hands on experience with new Zeiss Supreme Primes at the ASC Clubhouse event yesterday. My initial thoughts: Zeiss went hard to hit some serious points while slightly neglecting others for weight and price point reasons.

Thanks for the write-up!

 

[John David Pope]

The person making the presentation tried to actually make that a positive by rationalizing that since most people use the more open side of the scale, the non-linear aspect of the scale allowed for more barrel rotation in those areas in which one could move the ring. I won't quote him but it was something along the lines of having more space to move around in that area of the scale. It was just a bizarre attempt to try to rationalize and sell a crowd on the cheaper, less preferred scaling. This was at the ASC clubhouse though and I'm not sure most DP's there know or care about the iris scale being linear or non, and as an AC/DP I can tell you this is really not a big deal, although I prefer linear without a doubt. There is no benefit to non linear except it's cheaper to implement into the lens. There are no in-the-field advantages, no matter what Zeiss sells you.

I think the Zeiss rep has a valid point. I think most would agree that a barrel rotation is advantageous with a linear iris. That is, it is advantageous to have a large amount of barrel rotation between stops spaced in a linear manner. Barrel rotation is good. The more the better ( up to a point ). Thus, if a non-linear spacing makes for greater barrel rotation in the area where the iris is most likely to be set, that is a good thing and makes for "better" than a linear spacing that results in less barrel rotation in the range of stops where lenses are most used ( from T5.6 to wide open ). Imo the Zeiss rep has a valid point in that the non-linear design better facilitates incremental adjustment of the iris in the range of apertures most used.

The focus scale of lenses is never linear and AC's don't prefer that it be linear. The amount of barrel rotation between 5-10ft isn't the same as the amount of rotation between 205-210 ft. The amount of rotation is greater in the distances where the AC is likely to be performing incremental adjustment. So I don't see why an AC would prefer the aperture scale to be linear. As far as linear barrel rotation of the iris being something seen in more expensive lenses and not seen in "cheap" lenses, most low-cost still photography lenses use a linear iris scale. For example, Canon FD, Contax-Zeiss, and Nikon.

 

[Michael Lindsay]

Hi Ryan

Great detailed write up..

They are a new type of Helical and as you noted if they are not cam based heavy manipulation of focus scales is just not possible.... should they have then gone cam? I don't know as I have no idea about about much that would have added to the cost...but I do know they are a much better helical system than UltraPrimes... And if they are as reliable as UPs then fantastic!

Focus marks: What Leica acheived with Sumiliux-C is incredible but they are twice the cost and are still not FF... Interestingly Leica has produced 2 ranges of lenses since Sumiliux and have not attempted it again. As you note/implied I believe that Zeiss is just trying to hit size, weight and cost targets. A focus mapping LCS now costs less than 1/2 the price of one lens and works for everyhting. It is not as if Zeiss haven't been involved with Cam based focus scales so they can make it work.. but it seems not without a weight, cost or size penalty.

I was initially bothered by the non linear iris but after thinking about it, and asking a few assistants, I really don't care... Hawke don't linearise their irises either so maybe it is not only a cost restriction?.. Hawke lenses are not exactly cheap. However I would defer to others on this as the few times recently I have done Iris pulls they where remote and I worked off a monitor/WFM for timing and exposure neither type would have made any difference to me

Mounts: These are designed for 52mm flange depth/backfocus. So although they will be offered in LPL and EF mount, these were not designed for LPL mount, so there is no benefit to using LPL.

Possibly.... but I don't believe this to be necessarily true... Even looking at the different Bokah between the same lenses at the same stop on a LF, a Venice and a Monstro suggests to me there may be some differences created by the mount....

Re field ilumination I feel you are probably bang on... the world is changing and meta data in lenses could be used for getting around lots of optical artefacts! Uneven feild ilumination being the least of it. Also probably over 50% of what I have shot on MPs ended up with a tiny vignette on it anyway ;-)

For me the good news is outside of PV we now have some really good FF options....

PS As a long time user of Zeiss lenses I am pleased that they have more pleasant flares than I am used to..

 

[rand thompson]