SALT III- High Speed Lens Shootout "WFO" Results

[Matt Ryan]

http://matthayslett.com/salt-iii-high-speed-lens-shootout-wfo/

Welcome to the 2013 Southland Alternative Lens Shootout "Wide F%$@*^G Open" Edition. This was a coordinated effort between the following individuals: Will Keir, Phil Holland, Matt Uhry, Matthew Duclos, Jeff Whitehurst, Evin Grant, Ryan Patrick O'Hara, Luke Edwards, Charlie Pickle and myself. I'd like to begin by thanking our Sponsors who not only donated their time, but their facilities, insurance services, and their lens sets:

Lens Sets Tested: (Photos courtesy of Phil Holland)

1. Zeiss Master Primes

2. Zeiss CP.2 Super Speed

3. Leica Summilux-C

4. Luma Tech Illumina

5. Optica Elite S35

6. Canon K35

7. Zeiss Super Speed MKII

8. Canon CN-E

Mission Statement:

The purpose of this shootout was to directly compare lens sets on an identical setup. Our goal was not to "test" the lenses, rather provide you with all the info necessary to evaluate the lenses and decide what works for you. These lenses cover a wide range in affordability and availability. This shootout was never meant to prove which set is better than the other. We hope you will see that any of these lenses could be used to shoot quality material and have been for years. We attempted to get Cooke 5i's for the test but unfortunately none were available at the time. Some of the lenses have slightly varying focal lengths so understand field of view may appear slightly different (i.e. 24mm vs 25mm, 75mm vs. 85mm). Please also keep in mind the slight differences in exposure since we were working with lenses rated between T 1.3 and T 1.5. No matte box was used to insure there was no vignetting or concealing of flares.

We all know these lenses perform well closed down to a T 2.8, so this shootout was designed to push the lenses to their limits wide open. Most of the flaws visible wide open disappear once stopped down (examples will be available for download).


Projection:

The below results were attained using the lens projector at Duclos lenses that is set up in it's own room, specifically for the purpose of evaluation of optics on an even playing field. The same projector and screen were used for all lenses and the distances were maintained for the equivalent focal lengths (3' for wide lenses, 6' for teles). A loupe was used to determine resolution in line pairs per height and width for the Center, Field (middle) and edge of the frame based on a super 35mm motion picture frame. All of the findings below should be interpreted as applying only to the S35mm frame in which the lenses are designed to cover. Numbers are specific to each lens wide open. Lastly, although we did measure the lenses in line pairs this is an observational procedure and rather than deal with the political ramifications of publishing these absolute numbers we've decided to present them as they apply to the S35 sensor in the Epic and Scarlet. The data presented below is such that even measurements with a "Fair" grade will, for the most part, resolve the full resolution of the sensor.

Notes:

Where you see notes in the table below represent specific observations, if a field is blank that means that the lens performed average for that metric. If it was either better or worse or particularly apparent it was recorded.

Perspective:

Keep in mind these are purely technical numbers and do not speak to the cinematic look and quality the lenses exhibit in a real world scenario.

Duclos Evaluation PDF (Click to download)

 

[Matt Ryan]

Torture Test:

This shot was designed to push all the lenses to the edge. If you plan on using these lenses wide open you need to view them at their worst. This test will show the differences in Resolution, CA, Contrast, Flare Characteristics, Bokeh, Breathing, Distortion, and Color Rendition. The test was performed both wide open and at T 2.8. For each focal length increase, the camera was moved back to maintain a similar frame for every lens. The test was captured with the Red Epic (5KFF, 320ISO, 23.976fps, 3:1 compression, 3200K) due to the high resolution sensor and ability to change lens mounts in a minute. Shutter speed was used to control exposure. Focus was confirmed for each lens in a 3 step process. First using the built in focus assist function, then by eye off the siemens chart, and finally confirmed on a 1080p on set monitor. All this was performed after the distance was confirmed with a tape measure.

Here is a look at the test in action:

Beauty Test:

This test was designed for the telephoto lenses of each set. We wanted to see how each lens treats a human face both wide open and again at T 2.8. Focus was set to the left eye closest to the camera with the same three step focus confirmation. We have to give a big thank you to Stephanie from Duclos Lenses for being our test subject. Shutter speed was used to control exposure. Shot on the Red Epic (5KFF, 320ISO, 23.976fps, 3:1 compression, 5000K post adjusted).

In action:

Video:

The video comprises both the Torture Test and the Beauty Test. It's about 45 minutes long but make sure to watch the entire thing. Hope the music makes it a little more entertaining for you!

Keep an eye out for the areas highlighted below. This will give you a baseline indication for comparison. View downloads at 100%.

Select Original for 2K video:

Download 2K H.264 video here. (Right-Click, save as.) (3.2GB)

Download 4K H.264 video here. (Right-Click, Save As) (5.08GB)

 

[Matt Ryan]

Comparisons: (Wide Open)

(at 100%)

Downloads: (R3D Grabs and Images)(Click to download RedcineX Pro)

(Each of these downloads contains R3D grabs, FF5K JPEGS, and projector results for that given set.)
Zeiss Master Primes
Zeiss Super Speed MKII
Zeiss CP.2 SuperSpeed
Leica Summilux-C
Leica Noctilux
Canon K35
Canon CN-E
Canon EF
Optica Elite
Luma Tech Illumina
Rokinon Cine
Comparison Images

Conclusion:

It’s safe to say the torture test is rather harsh on all of the lenses while all perform well on the Beauty Test. Like I said earlier this shootout wasn’t intended to pick a winner, but really to compare and contrast between the sets. This test is definitely not definitive and each real world shoot has a specific need or look. A set that works for one project may not work for the next. Download the test material which is provided in the links below and have a look for yourself (both R3D grabs and stills are available for each set). Before you make any conclusions I would urge everyone to perform their own tests if possible. Those of us who put this shootout together would like to thank all of you for taking the time to view this and hopefully you get as much out of this as we have. Phil will be posting some interesting Dragon info and what lens sets covered in our tests.


*Disclosure:

None of us were paid or guaranteed services or gifts in exchange for performing this test. We have no vested interest in making one set look better than the other. Those with lens sets in the test were not operating the tests for obvious reasons. The RED Epic FF sensor size is slightly larger than the S35 Motion Picture standard so any lens vignetting does not appear or affect the traditional S35 frame size.

BTS:

Links:
thecamerahouse.com
camtec.tv
http://www.redstudio.com/home
http://www.lumatechinc.com/S35.html
houseofsmooth.com
athosinsurance.com
ducloslenses.com
bandpro.com
willkeir.com

 

[Matt Ryan]

Thoughts from Evin Grant:

Sorry for the late post guys but I have a newborn in the house and finding time to sleep is hard enough these days.
First I'd like to say that Will, Matt R., Matt D., Matt U., Phil, Jeff Ryan and Luke did a fantastic job putting this together and although I was there to observe and give my opinion this was truly their effort and a job well done.

Suffice it to say I agree with pretty much all the technical conclusions that the rest of the group have arrived at, mainly...
1. All the lenses performed about equal stopped down to T2.8
2. If you want lenses that truly perform wide open you have to pay for the best (Leica C or Zeiss MP).
3. Rehoused still lenses do not a real cinema optic make.

My impressions:
Unlike the test it's self I do not hesitate to tell you that I'm biased. I have never hidden the fact that I like certain brands of lenses better than others (Leica, Nikon, Angenieux etc.) and that I generally prefer T2 lenses to T1.3/4 lenses. However I have on many occasions chosen fast lenses for the projects that dictate their use so I'm pretty familiar with MPs and Supers. The rest are mostly new to me but this test has revealed a few things I didn't know. Anyway here are my impressions from worst to best plus a little bonus addition...

7th... Canon K35s
To say that the K35s disappointed me is an understatement. For the last 2 years I've heard people raving about how much better they are than the Zeiss Supers but this set at least did not bear that out. I found pretty much nothing I could really love about these lenses, they showed far too much CA and the 85 was pure mush wide open. Even the bokeh had a bit of a double image I found unpleasant. Again, i've never been particularly fond of Canon glass but I expected better from the famous K35s.

6th... (Tie)... Canon CN-E/Zeiss Compact "Speed" primes
Both these sets are re-housed still glass from "L" or "ZF" lines and it shows, the barrels are nice and much more usable that the still equivalents but they do not really measure up to the purpose-designed cine lenses from the big boys. They still breathe too much and have too much CA, especially the Canon primes. On that note Canon's choice to only release these in EF mount tells me they are not really serious about this, the EF mount if fine for a certain type of project but it is not capable of standing up to the rigors of larger productions.

5th... Optica Elites
I like Russian glass, it has an organic character that really helps to soften the "digital" look of some other high end lenses. Technically there are many things not to like about these lenses but the images somehow still come out pleasing, unlike the Canon K35s which have most of the same issues but none of the visual character of this set.
The only thing to dislike is the barrels, Russian glass is fairly hard to maintain and work on and replacement parts can be hard to find.

4th... Illuminas S35s
My impression is these perform slightly better technically than the Elites but have a very similar "Russian" look as is to be expected. They also offer better barrels, mechanics and service since they are a currently manufactured product.

3rd. Zeiss Super Speed MKIIs
I've always preferred Zeiss T2.1 standard speeds to supers but to say that this set held it's own is a testament to how well they were designed in the first place.
They have their issues and are pretty much double what I think they should actually cost but the performance surprised me. Especially since they we're up against the best of the best. I would not have picked Supers for 3rd place (which honestly should go to the Cooke 5is if we had gotten them) but despite all that the ol' stand by Supers proved to be able to handle most of these tests with both a reasonably sharp image and more importantly a pleasing one. Not technically perfect but as good or better than any of the contenders save for their rightful heirs the MPs and Summilux Cs.

2nd. Zeiss Master Primes
There is very little to say here except maybe these lenses are too perfect! They are superbly consistent in both color and contrast, sharpness is impeccable and the look is pleasant if a little clinical, the only thing this set gives up to land in 2nd is CA, where as the MPs show almost none, the Leica's literally have none.

1st. Leica Summilux "C" primes
The new kid on the block truly is the king. They needed to truly do something special to beat the Master Primes from their rival german juggernaught Zeiss but boy did they make some scrumptious glass. I've been shooting Leica M for 15 years and they are hands down my favorite lenses and these live up to that heritage and then some. If I owned a super-car to hock in order to get a set it would have happened already, but you knew that of course.

Bonus!
There happened to be a set of Schneider Cine-Xenar IIIs at Duclos on our test day and it just wouldn't have been right to leave without at least throwing them up. My very quick impression is that these are the "Sleeper"set of 2013. They looked the closest to the "Optimo" look of any prime set I've yet tried. The lenses were larger than Ultras but not as big as RPPs or MPs. Breathing was non-existant, sharpness was stellar and the bokeh was luscious. I'm hopefully going to be able to borrow a set for a music video coming up so with any luck you'll have a chance to see them in action.

E.

 

[Matt Ryan]

Thoughts from Matt Uhry:

You're painted yourself in the corner , ISO already sky high, you opened the shutter to 1/32..… in the dark ass desert night exterior your lone 18k looks puny in the distance, casting "moonlight" over a few acres of rocks and cactuses. Suddenly the director wants to shoot 48fps, you wish your had talked the producer into more / bigger lights but fuck it, you're going WFO, wide f-ing open.

Think of this as a torture test, not a friendly test. We chose test subject to try to make the lenses stumble, trip and maybe fall on their faces. The aluminum foil will show center CA and the latches on the violin case will show edge CA and sometimes coma. You can get a sense of contrast and veiling flare by sampling the RGB values in the center dark area and comparing it with when the MR-16 par light is on or off. Cower in fear as Phil uses his bright ass LED to aggravate lens flares. You can get a sense of breathing and foreground and background bokeh when we rack from the targets to foreground grid then to minimum and then infinity.

All of the lenses we tested preformed better when stopped down a bit, and they look much more equal to each other at T2.8. If best optical performance is important to you, don't shoot WFO.

My opinions begin here:

In real life overall color balance is pretty easy to correct, as is adding a bit more contrast. Edge resolution is important for tableaux shots, especially on wider lenses. I find that the edge / corners of longer lenses are usually defocused on typical subjects, unless i'm doing straight on copy stand style work which is pretty rare. Distortion and Chromatic Aberration aka. CA while possible to address in post are important to consider as are lens flares, desired or not they are baked into the image and would not be practical to remove. Flare ( different than lens flares ) is the hazy halo that appears around a bright object, a common and usually undesirable low contrast / foggy looking effect that you'll see in the older lenses. You can have good resolution and a large amount of flare and the image will look "soft" but have detail. Have a look at the nice old school portrait effects of the T1.3 K35 85mm at the end of the clip.

Master Primes

MP's along with S4's has been my go to set when budget allows. I was surprised to find a bit more CA and a bit more cyan haze when WFO than I had ever seen in the real world, which is to say they are a great set of lenses. I think there are 2 downsides, the cost and their imposing physical size. There's not many flaws or flares so not much "personality" either. Bokeh can be a little donut like for my taste.

Leica Summilux C's

The newest designs that we tested and I think you can see it in the low to nil breathing and great, even center to edge performance. The most controlled for CA and had the least wide open cyan haze color shift of any of the sets.

Zeiss Super Speeds

Classic set, not great wide open, great at 2.8 - not for perfectionists with the edges CA and flaring wide open. I don't think the cost of used sets is justified when the Elites and Optars offer better value.

Luma Tech Illumina

This was the factories demo set and is presumably optically the pick of the litter. There was a bit more CA and flare on some of the focal lengths and others looked decent. Mechanicals are much better now than on the early sets where there was a big tendency to bind when using clamp on matte boxes. Physical size is quite nice. Bokeh on this set is pretty. It's probably one of the easiest high speed set to buy, pay the money, they ship, shoot movie.

Optica Elite

My personal set…. so I may not be as objective as the others. There's a bit of CA and flare on some of the lenses when WFO and Not tested, but part of a typical set was the 100mm which is pretty sweet. I like the low breathing and nice cam driven mechanicals ( shared with the Master Primes and Leica Summilux C's ) which allow a non linear focus scale: more rotation at the distance you need it. Since the lens outer body is one piece you can use as heavy a clamp on matte box as you desire. I also like the physical size of the set. The lenses come in a larger 110mm diameter, those models focus a bit closer than my "compact" set but are optically the same. Optica Elite is phenomenally bad about marketing their lenses. The factory has been responsive for any parts / repair needs that I've had.

Zeiss CP speeds

It's annoying and confusing that Zeiss is calling the set "Super Speeds" Mechanically they are nice, to me the 114mm barrels are bit over the top, but that's becoming a standard so maybe it's OK. The 3 fast lenses are not optical stars having a bit more CA and flare than fresher modern designs.

Canon CN-E

All the Canon's had quite a bit of CA, we had some image shift on this set, but were not able to attribute it to the build of the lens or something in the EF mount. There was a noticeable improvement in the coatings of the CN-E's than on the L's. I wish a company like Canon would commit real resources and intelligence into the lenses themselves rather than the marketing

Canon L

Similar to the CN-E but with loose AF style mechanics and slightly more prone to lens flares. I cannot recommend them for motion use unless you've got a good electronic way to control the focus. Refer to SALT 2 for more Canon L action.

Canon K35

I did not find much "classic" in this set they just seemed old. Heavy CA, heavy breathing, a bit of image shift, mechanics worn and maybe a bit wonky in the first place. Design not clip on matte box friendly due to helical binding, different front diameter on 18mm.. Good coverage of course from 24mm on up.

 

[Phil Holland]

Phil's Notes (1 of 4)

First off I'd like to thank everybody involved with putting this test together. Especially the sponsors for allowing us to shoot at their spaces and use their lenses. I especially want to thank Will and Matt for inviting me to tag along. Good to see and work with some familiar faces and always great to meet new ones. Big thanks to Jeff for handling the DIT duties (and book transport!). Testing can be grueling work sometimes, but this was a whole lot of fun and extremely informative for me.

Matt's done a great job of writing up and providing all of these images.

Why was I there? I'm a bit of a lens head and do a lot of technical/VFX related imaging and tests. My job was to help out during the tests, take pretty pictures, and make observations. My perspective in general for this will be geared bit more towards the full sets of glass. I have a lot of thoughts, but I'll try to keep it concise.

For this test the focus was on looking at these high speed lens's optical performance by testing them at their widest aperture and closed down to T2.8. As Matt stated we were looking for information regarding breathing, image shift, coverage, build quality, general usage notes, and unique lens characteristics. With that said my needs and desires from a set of lenses may be different than others and may even vary from shoot to shoot. Especially when time, budget, and portability come into play. I'll try to express my opinion where I can, but I'll also be playing devil's advocate and try to take myself out of the equation where possible. We all have opinions about what we're after. None is more valid than the other. It's more about how you want to create your frame and how you can use the tools to accomplish that.

You'll be able to look at these images and make your own conclusions as to how they perform on an optical basis. For me, I'm looking at these as individual sets and how they perform as a whole. These were mostly "working lenses" and in various states of service. Also, some of these lenses are several decades old. So there will be variances.

Let's look at the test setups.

---

The Stress Test was conducted at Red Studios Hollywood. Big thanks to Jarred and Carrol for allowing us use of Stage 3 for the day. Will brought the books and Matt Uhry brought his highly effective "Torture Board". Most of the weird props are from me, but they were chosen for a reason. Matt Hayslett and Ryan O'Hara operated the camera for the whole day. Everyone kept open eyes to make sure nothing was being missed out on. Matt and Ryan will have some great info after operating the whole day.

The critical observations I'm primarily looking out for this setup are various types of optical aberrations, illumination falloff, how/where detail resolves on the focal plane, depth of field, focus plane object separation, perspective draw, chroma/tonal consistency, and how nice the out of focus image (bokeh) appears. Additionally, during the focus pull, you'll be able to make notes on breathing and image shift.

What's extremely important to keep in mind is that this was a stress test. Meaning we were trying to make the lens do funky things. That said, most of these lenses cleaned up a lot of their wide open performance aberrations by T2 or T2.8. Especially the newer glass.

---

Stephanie was shot at Duclos Lenses using a simple three light setup using daylight balanced CFL bulbs diffused under Matt Hayslett and Matt Uhry watchful eyes. In frame you have a color chart and gray card to observe skin tones. Lenses used were the 75mm or 85mm from each manufacturer and were shot wide open and stopped down to T2.8. A notable thing to look out for is how the out of focus image behaves behind our model. Also, you can see the general difference in contrast between sets and the subtle increase in contrast on certain lenses as they are stopped down.

---

Also conducted at Duclos Lenses, Matt Duclos loaded up each lens on the projector to make critical observations regarding resolving power, optical aberrations, lens distortion, ghosting, light fall off, breathing, flatness of field, and contrast. I took the notes during this test, but Matt will have the most to say regarding the hard number results of this test because it's his "everyday" business over at Duclos Lenses and he has a great deal of expertise in pin pointing lens characteristics.

Something to note regarding projection. We tested at specific distances and locking very accurate focus. You can indeed notice varying performance in certain lens characteristics by shooting in the field at different focus distances across the scales.

---

Depending on the type of work you shoot color (chroma) and tonal (luminance) consistency across a set might be important to you. I extracted samples from the gray card off of Uhry's Torture board to do some quick comparisons. I should say that this is not normally how I conduct this test, but due to the setups and lighting being consistent I feel it's worth looking at. Also, normally I test the full aperture range of a set. Here we'll be looking at the wide open and the T2.8 samples we used on the stress test.

Nearly all of the lens sets were within a 1/4 stop of each other in regards to tonal variance, except for the Elite's, which exhibited a bit over a 1/2 a stop difference. All within correctable levels in reality, although 1/2 a stop isn't lovely and would be considered a pretty noticeable difference.

Color is a different story and is a bit more noticeable during lens changes on similar setups. What I'm looking for here is "harmony" between the lengths in terms of chroma. Meaning color chips on the same horizontal aperture row that appear to be close to the same color is ideal. Some lenses render warm, some cool. Not a bad thing and certainly something that can be corrected if needed.

Based on that, the sets that I feel did a good job overall here are Zeiss Master Primes, Leica Summilux-C, and Zeiss Super Speed MKII sets. The tonal variance on the Elites seems to go away pretty quick as you stop down. I like the chromatic feel of the Illuminas, but they do produce a fairly subtle tonal variance overall. The Elite's and this particular set of K-35s exhibit the most color variance.

Again, this is pretty picky stuff that can all be corrected relatively easily. Some DP's have very strong feelings about this subject though and it's a relevant concern/topic.

---

Observation Tips

There are a lot a great areas to look at across each focal length. Besides resolving details here are some nice areas too make some observations.


*note the neutral chromatic nature of the Leica's chromatic aberration. I personally enjoy that look.



*observe these lights during the focus pull.



*also check for ghosting around the knife on the cutting board and the effect this has on the henchmen frame right.

---

Just a note on coverage. All of these lenses cover 5K Full on the Red Mysterium X sensor, which is larger than the S35 4-perf image circle. They may exhibit light falloff and likely won't be the sharpest in the corners, but if you're wide open that's not likely what you're after. A quick look to the future. The Red Dragon sensor isn't here just yet, but what I can say is that many of these lenses will cover Dragon 6K based on what's currently known as of 02/01/2013. Where some might get in trouble is on the the wider focal lengths, like 18mm and beyond. It is just too early to make that call though (anything can change in reality). That said, with Dragon's 5K being nearly the exact size of a S35 3-Perf take comfort in your current and future lens investments being extremely valid for common production work. 6K HD isn't that much larger than a S35 4-Perf Full Aperture frame in reality. An interesting thing to note based on observing this graphic is that there might be a reason to have a 5.5K/5.6K format on Dragon which would lay within the image circle of S35 4-perf Full Aperture. The final formats haven't been announced yet though.

 

[Phil Holland]

Phil's Notes (2 of 4)

---

Zeiss Master Primes have earned their reputation in the industry. I'm fairly familiar with these lenses and I'll be using them as the measuring stick when looking at the other lenses from this test. Optically these lenses perform extremely well, virtually no breathing, well controlled aberrations, and impressive wide open performance. From 14mm-135mm you are looking at a consistent front filter size of 114mm. From 16mm-100mm you are looking at a consistent lens length of just over 6 inches. This is desirable on set and helps speed things up and minimize rig adjustment during lens changes. Focus pulls are smooth as can be really. This is a well maintained and serviced set as well. Some things that people don't particularly love about Master Primes is that they are fairly hefty and bulky per lens. This may or may not be of concern to you, but I've heard more than one horror story involving people dropping them due to small slippery hands. They do have a look to them overall. It could be described as cool and clinical. Very contrasty too. Overall, these are top tier lenses and what many hold as the highest standard in cinema primes.

---

The Leica Summilux-C lenses are the relatively new guys on the block in the high end cinema glass realm. There's a lot of good things to be found within their design. They are around 3.5-4lbs per lens. Throughout the range they have a consistant length of 5.6 inches and a front diameter of 95mm. Built really well and also virtually no breathing. From a Red shooter perspective, I love the size of these things on Epic and Scarlet.

I've had experience with the 40mm and 75mm before this test and really liked what I saw. After looking at our test images I'm beginning to like them even more. Whatever Leica is doing with it's modern elements and coatings is fairly tremendous. There's nearly no chromatic aberrations exhibited across the range (it almost produces what I would describe as a "holy glow"). Better performance on that front than Master Primes in my mind. Spherical, coma, and astigmatic aberrations are all really well controlled as well. Additionally I like the flatness of field and separation that you see on the Summilux-Cs. They have a bit of flavor to them. You will notice a bit of a sparkle on your out of focus hot spots for certain lengths. This is clearly part of the design and you will notice as you close your aperture down you will see a unique pattern form from the iris blades. They are pretty contrasty and I would say have a very slightly warmer feel to them compared to Master Primes. These are cinema lenses that seemed to be designed to have a subtle yet distinct look to them.

---

I have to say this is a pretty nice set of Zeiss Super Speed MKII lenses. Especially the 50mm. These are great little workhorses overall. They are all nice and compact and feature an 80mm front diameter across the range. Built to last, which is one of the many reasons you still see them in use. The aged and dated coatings show a bit. Fairly contrasty overall, especially the 50mm in this set. There's certainly chromatic aberrations exhibited across the focal range. Also, they do ghost and "glow" a bit. I would say the Elites exhibit that the most agressive ghosting, the Leicas, K-35s, and MKIIs having a bit less of it at 50mm. It's there though. That all mostly clears up by T2/T2.8. The do exhibit a bit of breathing when racking focus, but it's not too bad. I would say these have a similar Zeiss coolness about them. I've always liked the look of the MKIIs. No real shockers on this front. Hard to beat for such a portable set.

---

This is actually my first time working with the Canon K-35 lenses. One of the sets I was most excited to use mostly for their reputation for having a unique look and pretty high image quality. All have 80mm fronts except for the 18mm, which is 110mm. Pretty light weight and impressive close focus. Up to 35mm gives you 12 inches, which is nice. Short and compact, similar to the Zeiss Super Speed MKII lenses, although the K-35 glass feels a hair lighter in my hands. These lenses are a little older than I am and just like other vintage glass of it's class it exhibits similar chromatic aberrations to the MKII's. I would say, when directly compared to the those Zeiss Super Speed MKII lenses, the K-35 lenses are a bit softer in the field towards the edges of the imaging plane. Noticeable even at T2.8. We did put up the Canon CN-E and EF lenses of equivelant lengths and it's interesting to see what's changed in the last 36 years. Generally speaking the modern optics are better. The K-35 lenses do in fact breath a bit as well. They do have a unique look to them that I can see being appealing. To me the Zeiss Super Speeds MKII are better lenses in nearly every way when compared side by side like this.

---

The Elite S35 Digital are a pretty interesting set of lenses. They have a consistent front diameter of 85mm through the set. I'm somewhat baffled by how long the 24mm design is though. I would say it's going to be a showdown between the Elites and the Illuminas as to which set has the most "personality". I mean this both in positive and negative ways. The Elites exhibit a healthy amount of chromatic aberrations and ghosting. Much clears up stopped down to T2.8. The lenses do have a unique quality to them, especially the out of focus image. They aren't as contrasty as the premium cinema lenses. I think the 18mm is pretty nice overall. The 24mm isn't exactly a 24mm and is noticeably wider than the other 24mm lenses tested. The 50mm ghosts and blooms wide open in a an aggressive way. Overall this set left me wanting more. They provide a unique look and familiar bokeh across the set (wide open it's very exciting on not so controlled). But these have some quirkiness about their build and optical performance. Chromatic aberrations still pop up at T2.8 at most lengths. I would say they are more contrasty than the Canon K-35s and pretty close to the Illuminas. The Elites are prone to flare, which is both a good and bad thing depending on your creative intentions.

---

**edit 14mm T1.8 coming soon. Not 16mm.

The Illumina S35 lenses were another set I was happy to finally get my hands on. I've heard and seen extremely varied and confusing things about these lenses. There are simply not many sets out here in Southern California. First handling them you get the feeling that they have crammed a whole lot of optics into the lens tubes. They have a solid feel to them. Mostly around 3lbs, they all feature consistent 95mm front diameters. Their size is actually very appealing to me for use on Epic and Scarlet cameras. In a similar way to how the Leica Summilux-C lenses feel "ideal". One design issue is that pressing down on the front of the lens induces friction which causes an effect that I would describe as a "break pad" on your follow focus ring. That may be an issue for some folks who use clamp on matte boxes and operate by grabbing the sides of the matte box itself. I never use clamp on matte boxes, but it's note worthy for sure.

At some point my interest in these lenses became somewhat of a fascination. The guys were even ribbing me about my weird infatuation with these Illuminas. I started noticing some interesting things about the design of these lenses. They overall have a distinct and consistent "look" across the set. Wide open you'll see some "exciting" aberrations (more exciting than the Elites) in the out of focus highlights, however, upon testing them further that goes away at T2/T2.8. Which is oddly similar to the Elites' performance, but just a different flavor. As with the Elites, the Illuminas are mighty prone to flare. Chromatic aberrations do exist, but most clear up at T2.8. I'm left wanting to look at these lenses closer and shooting a larger variety of subjects with them. They have character and a feel I rather enjoy.

 

[Phil Holland]

Phil's Notes (3 of 4)

Zeiss is now making new CP.2 Super Speeds that all have 114mm fronts. These are based on still lens designs. If you've used other CP.2 Zeiss glass you know what to expect. The 50mm CP.2 T1.5 for me was one of the worst wide open performers out of this test, which was rather shocking. These aren't perfect cinema lenses and do breath a bit. They are relatively affordable and a good entry set to cinema glass. Also the CP.2 lenses come in PL, EF, and F mounts which are user interchangeable. Based on the logic of the CP.2 releases it's unclear if we'll ever see other super speed focal lengths in the immediate future because that would require new lens design. The 85mm performed pretty well overall, but they all breathed. If you are considering these lenses you may want to explore finding a good set of Zeiss still lenses and get them Cine-Modded. It is the same glass. I do like the build, size, and weight on these.

---

Canon has followed Zeiss's lead in the design of their CN-E cinema primes. They even have the same front diameter of 114mm. I am familiar with these lenses and can confirm that it's also the same still EF glass inside these lenses. Just different coatings and mechanics. You can see similar chromatic aberrations comparing them against their EF relatives. There will likely be a 35mm on the way soon, which is good news. However, like the Zeiss Super Speeds, there is probably not much chance in the immediate future for a much larger set of CN-E fast primes. They have announced slower 14mm and 135mm lenses though. These are currently only available in EF mount for some reason. If Canon wants deeper market penetration they should release a PL version. I'm confused about this as they are offering PL cinema zooms, yet not primes. Perhaps they will have some surprises for us at NAB.

---

Canon EF/Still Glass/Cine Modding


We put up the Canon EF glass I had on hand against the CN-E's and also the much higher quality cinema glass. We all know these aren't designed for motion and exhibit image shift and various levels of breathing. What I will say though is that a Cine-Modded set of still lenses from Canon, Nikon, Zeiss, Leica, or "other" represent perhaps the best bang for your buck when it comes to fleshing out a lens set. The value to price ratio is pretty high in that regards. One of the great advantages of shooting on Red is that you can you these lenses and have electronic control over supported focal lengths. Also, most still lenses are compact and light weight. Nice if you need to hike 16 miles to get the shot I've used still lenses on many productions over the years. Large budget productions. These will never be production workhorses like the top tier cinema lenses out there. However, for relatively little money you can get a full shooting set and get out there and get quality work done.

Personally I own a full set of still lenses and usually rent higher end PL glass when needed. That's worked for a good while for me. Now I'm moving on and looking for a set of quality PL glass of my own. I'll be holding onto my still glass for motion and still shooting though. Even if I had a set of Master Primes laying around.

We previously tested still lenses in the "SALT II" tests. Make sure to give it a look if you are considering still glass, Cine-Modding still glass, the CP.2 lenses, or the Canon CN-E.
http://www.reduser.net/forum/showthread.php?73380-SALT-II-results-(Part-1-2-3-4-Bokeh-Breathing-amp-Flare-and-Part-5-Impressions)

---

Rokinon

We briefly tested out the Rokinon Cine Lenses while at Duclos. Rokinon needs to be commended on releasing extremely affordable cinema style lenses. Nobody has really done this before. Once the 50mm comes out you'll be looking at around $2,000 USD for their entire range of fast glass. They are also working on a 10mm f/2.8 which currently nobody makes. Pretty exciting. That said, these lenses don't perform as well as the more expensive "mainstream" still lenses out there. They suffer from flare, low contrast, and chromatic aberrations when wide open. There's no free lunch with glass in reality. However, if you are just starting out, this is a good starter set with "cinema minded" mechanics to get you on your way.

---

Missed it by "that" much.

There was indeed a Cooke missing from our kitchen. We didn't have the Cooke 5/i set to test out. I also have zero experience shooting with these lenses. I have however worked on material shot off the 5/i lenses and there's a few things we can say about them. From 18mm-100mm they have a consistent front diameter of 110mm. They are similar in size and weight to Zeiss Master Primes. The 5/i are very sharp lenses as well. I would say they resolve a smooth yet crisp image. There is such a thing as the "Cooke Look" and you either love it or loathe it. I'm on the love side of that fence. I've always liked what I've seen off the S4/i lenses and the 5/i, while slightly different overall, do indeed maintain that heritage and feeling. The Cooke 5/i are right up there with the Zeiss Master Primes and Leica Summilux-C lenses in terms of performance as well as price.

Another set that didn't appear at this test are the Vantage One T1-11 lenses. They are brand new to the market and I don't currently know anybody who has a set. They have a nice range of focal lengths: 17.5, 21, 25, 32, 40, 50, 65, 90 and 120mm all hitting T1 at their widest aperture. Which is madness. Wide open they are supposed to provide a unique donut/bubble shaped bokeh. They also all have impressive close focus abilities. Hopefully we'll get a chance to shoot some tests with these in the near future.

---

At the end of our tests we had a bit of time to do some real quick shooting outside and inside of Duclos Lenses. We all grabbed material of each other and tinkered with individual lenses that peaked our interests. I played with the three sets that most interested me and over the last week I've been playing a bit with grades and seeing just how skin feels on the glass I like. One graphic I wanted to show is actually running one of these fast lenses through some deeper apertures and observing the depth of field increasing. Here's the Leica Summilux-C 75mm T1.4.



In the above graphic you can see some very light purple fringing. You won't be able to judge critical sharpness from this sample set as this tree was moving a good amount in the wind. Check out the leaves and the power lines as they materialize. The leaves in particular are pleasing to me at the widest aperture. In a previous test with this particular lens I noticed that diffraction was coming into play about T11 and more noticeable, yet still very usable at T16. The Summilux-Cs for me "sing" from T1.4-T8.


Another thing to point out is that we are shooting 5K 3:1 REDCODE Raw on the Mysterium X sensor. Depending on what level of "crispness" you want out of the image is largely up to you.



In the above example at T1.4 we can carve a nice sharp edge from our 5K material to reveal all of the captured detail or create a more natural feel (perhaps more "film-like") by not touching the sharpness at all. Personally I tend to enjoy the look of a bit more detailed finished when mastering for a 4K delivery. That may even change based on my subject matter. The point I'm making here is you have the power and control to "sculpt" your final level of detail because we are shooting RAW.

 

[Phil Holland]

Phil's Notes (4 of 4)

So what can be said after these tests? First, a bit of a reality check. All of these lenses are usable. They all get better stopped down 2 or 3 stops as well. Some have their quirks. All have their own unique characteristics. Breathing, image shift, chromatic aberrations, and distortion can all be "fixed in post". However, that's not a great answer. Especially if we are talking the business and general production side of things. Fixing in post means time and time always means money. Good mechanics and solid lens design pays off in a big way in use.

No real shocker on the top performers in this test for me. The Zeiss Master Primes and Leica Summilux-C are top tier glass. With great mechanics and stellar optical performance. However, that all comes at a premium cost.

So with all of this data and testing where does that leave my actual opinion? After two days of testing and several days pouring over footage I'm left impressed with 3 different sets that I was able to get my hands on. And I'm indeed in the market to make a lens set purchase in the next couple months, so this test was highly relevant to me.

If I rolled out of bed tomorrow into an ocean of money I'd snatch up a Leica Summilux-C set. So many good things about their design, size, and performance impressed me. They produce a look I like. Business-wise, for those who rent gear, these lenses make sense. They have titanium mounts and lens barrels (titanium is about 3x stronger and 50% lighter than stainless steel, about 2x stronger than aluminum as well). That's nice for the sales pitch, but they feel pretty sturdy. They are set color matched from what I can tell (just like the Master Primes and Cooke 5/i lenses). Are they "better" than Master Primes? In some ways yes, in others no. It's that difference that makes me favor the Leica's a bit more. I'm looking forward to shooting with these more this year and in the future.

The other set that fascinated me and is a bit more within my reach are the Illuminas. I'm very curious about these lenses and I took three lengths to the side to play with before tests were done. These also have a look I rather like, but in a different way. I like the creative aspects of a lens that flares like this and I can clearly see how the image cleans up and transforms very nicely stopped down. I shoot a lot in the T2-T4 range and they fit most of my wants. I need a proper set of prime focal lengths, which is one of the reasons the CP.2 Super Speeds and CN-Es aren't very appealing to me, there's not enough glass in their sets to be considered a full consistent range. The Illuminas are a full set with more on the way. Their build and size feels good to me. I'll certainly need to shoot more on them to fully get the feel of what they can do. I don't fully understand my love for this set, but they have gotten me to take notice.

There was a third set of lenses that peaked my interests. It was outside of the scale and scope of this particular test, but when Matt Duclos walked out with the newest Schneider-Kreuznach Xenar III lenses and put one up on the Epic my eyes opened a bit wider. I think Evin had some very nice things to say about these as well. They aren't super speeds and they aren't super small. However, what they seem to be is properly designed cinema lenses (and likely extremely useful for commercial production). Their optical design is telecentric. They have very nice close focus performance. Also, they don't breath or shift. These are also within my reach. I won't say much about these now, but I'll be looking at the Xenar III lenses in the very near future. Schneider seemed to correct pretty much everything that was wrong with the previous versions of these lenses. There is something that feels very right about these now. Good on them.

Thanks again to all involved. It's been a blast.

One last word for those reading this. Take a look at the test images. Make your own observations. Push them around a bit. See what you can do with them. That's really what's important here. This part of the "chain" really does effect the look of what you are producing. There's always the technical marks you want to hit with glass, but you should always consider the feel of the image that you are capturing. These lenses all have character. Price really comes into play if you are looking purchase of set of glass and not just rent. The good news is if you are paying top dollar you are getting top of the line optics and mechanics. But I wouldn't be scared of the "old dogs" if you like their mojo. It's just what they do. If all glass looked the same a test like this would be rather pointless. There's really never been more options for motion picture lenses and more are on the way. These represent the fastest and latest. But there's a ton of other lenses out there in standard speeds that are truly amazing as well.

 

[PatrickFaith]

Thanks Much Guys!

I do a lot of geometry solves in Nuke, so i was wondering if you guys could comment on the "spherical aberrations" comments in the Notes column.

 

[Jiri Vrozina]

Looks like LEICA rules.....
Any word how SCHNEIDER Cine-Xenar lll compared? Matthew?

 

[Nick Morrison]

Very cool guys. Thanks.

 

[Matt Ryan]

Looks like LEICA rules.....
Any word how SCHNEIDER Cine-Xenar lll compared? Matthew?

They weren't directly compared. We just shot outside after the test was over. They were not a part of this test (only fast t1.3-t1.5 invited).

 

[Mike 'Fireman' Ross]

I'm curious, while you had the lenses on the bench, did you measure *actual* focal lengths and transmission values?

(it's not entirely unknown for some manufacturers to err on the side of optimism, shall we say? No names, no pack drill...)

Mike

 

[Phil Holland]

Kick ass job Matt (good stamina operating too). I have a feeling you and I were staring at these images a whole lot this last week.

Any word how SCHNEIDER Cine-Xenar lll compared? Matthew?

I'll be testing these soon. They are one of the sets I'm eyeballing at the moment.

I'm curious, while you had the lenses on the bench, did you measure *actual* focal lengths and transmission values?

Nope, but I made some calculated measurements from the Stress Test setup, which isn't "exactly" the right way to do that, but it shows some useful info.

 

[Brad Webb]

The Leicas just blew me away. Wow.

 

[Joseph Seuferling]

Great posts, Phil!

 

[Paul Ellington]

Great phil looked like fun wished I could have been there.

 

[Phil Holland]

We'll have to get you out here for one Paul.

Maybe once those Vantage One's are available. Those are going to be something worth experimenting with.

 

[Clayton Burkhart]

Thanks guys. Really one of the most useful and thoughtful tests that I have seen in a long time.
Shame about the Cooke's. They should have sent their lenses themselves, the test was that good.