AF100 Zeiss Ultra Vs Compact

So, how many situations have been in where you have said to yourself "if only I had that $12K lens so I get that shot..." Frankly, unless you have EVERYTHING else you need, and ALOT of business, is it worth it to own over renting? Frankly, you can have the best glass in the world, but unless you have proper lighting, support, etc, and the talent to bring it all together, it is not going to get you a great movie. I have seen awesome looking films shot on still glass with a $12K budget... Cine-glass is nice, but not the end-all be all of getting great shots.
 
BollingerPro said:
Guys,

Do the Zeiss ZE canon lenses allow for correct focus pull direction?

Is there a benefit to having a manual iris ring? I am thinking about getting three ZE lenses (28mm, 50mm 85mm) to go along with a Birger mount.
Click to expand...


The ZE's don't have an aperture ring but the ZF.2's do. And the focus pull direction is correct.
 
BollingerPro said:
Is there a benefit to having a manual iris ring?
Click to expand...
I'd say there's a huge benefit to it, and that would be -- what if you want to use these lenses on something other than a Canon? I bought the ZF lenses instead of the ZE's, specifically because they do have the focus ring, and I've been happily able to use them on the GH1, GH2, AF100, and my 16mm movie camera for about a year with no worries about iris rings, and will be able to use them on pretty much anything that ever comes out, because they're not dependent on any electronic support from anything else.

But, the tradeoff, is that I'm stuck with backwards focusing. Sigh.
 
One thing to remember is that cropping the lens's original image circle also means not using its full resolution. If we measure the resolution as a number of lines that the lens can resolve per picture height (or per inch, per mm, etc. depending on the method) then this is not an absolute number - absolute only for the given format. If you use a camera that has a smaller sensor size (and therefore smaller picture height) then the number of the lines resolved will be decreased. It is somewhat compensated by the fact that the center resolution of the lens is usually greater but that center resolution will also be proportionally decreased. That's why it is in general a bad idea to use i.e. medium format lenses on 35mm still photo camera, not to mention 35mm motion picture. The same lens, i.e. 80mm Carl Zeiss Planar optimized for 6x6 still photo will perform very poorly on a motion picture camera and perform beautifully on its "native" 6x6 body. In other words, the smaller the sensor size the greater the resolution the lens has to have. Some DP's would never use 35mm motion picture lenses on a S16 camera, even though they obviously cover the S16 frame - their resolution is just not good enough as compared with lenses specifically designed for S16. Ultra Primes are designed to perform excellently in a smaller image circle, benefit of that is especially visible in wide angle lenses. Their resolution is "packed" into a smaller area. CP's are still photo lenses really, just re-barreled to be convenient in cinema world (don't get me wrong, I want one set! ). Now, in reality, how much of an optical resolution do you really need for the camera that is not even 2K?
 
Last edited:
There is certainly truth in the notion that the smaller the format you're shooting to, the sharper the lens has to be in order to perform optimally.

However, it is also true that lenses typically (if not invariably) perform better in the center, than they do on the edges. This is the so-called "sweet spot" of the lens, and using a lens from a larger format on a smaller sensor results in the camera only "seeing" through the "sweet spot". That alone may be enough to overcome any worries about whether the lens is sharp enough, in general, for the smaller format. This effect can be seen in reverse when you put a smaller-format lens (such as a 16mm cinema c-mount lens) on a larger-sensor camera such as an AF100. The center looks great, but the edges get really soft and nasty.

In this AF100 forum, for use on the AF100, I'd say it's not something to worry about. The ZF's easily outresolve anything the AF100 sensor can see; any of the ZF's, even at wide open, deliver as much sharpness as the AF100 can resolve. Part of this is due to the AF100 being a less-than-2K system, and part is due to it having a much larger sensor than 16mm film, its sensor is over 3x as large as 16mm and it is almost as large as 35mm cinema film. The Compact Primes glass was originally designed as the ZF series to cover the stills "full frame" but Zeiss offers them as cinema lenses for the 35mm and S35 frame, which tells me they're comfortable having their name on these lenses as used for movie work.

It is an interesting question though, and since I have a PL-mount Super16 camera and some dedicated S16 lenses as well as the CP's, I'm planning on running some sharpness tests between them. The ZF/CP lenses are astonishingly sharp when stopped down to f/5.6 or f/8, so I have confidence that they are way more than enough for any large sensor (35mm "full frame" or 35mm/S35 cine or 4/3's) but it will be interesting to see if they still translate enough sharpness when rendering all the way down to S16...

I will say that I used the ZF's on Ultra16 on "A Verse Before Dying" and they were plenty sharp, intercutting easily with the AF100 and its Compact Primes.
 
All this is of course relative, it depends on what the delivery format is and where the final product will be screened. I understand that ZF's where plenty sharp on "Verse..." but was it blown up to 35mm in 4K and screened in a movie theater? That would be the test and I doubt they can handle it (again it is relative what we call acceptable results) - they would in my opinion not even come close to S16 lenses.
 
It has been screened on large screens and will be screened next month on a movie theater in NY, as far as I understand. Blowing up to 35mm and 4K doesn't do anything to add sharpness that wasn't already captured in the first place.

S16 lenses are a red herring -- they're useless on an AF100 so it doesn't matter if they're sharper or not, they don't cover the sensor so ... why discuss them?

But, again -- I have some S16 glass and some CP's and an S16 camera and will be able to compare them directly.

My point was -- seeing as this is the AF100 forum, the CP's are more than a match for anything the AF100 can deliver.
 
broughtonfilm said:
All this is of course relative, it depends on what the delivery format is and where the final product will be screened. I understand that ZF's where plenty sharp on "Verse..." but was it blown up to 35mm in 4K and screened in a movie theater? That would be the test and I doubt they can handle it (again it is relative what we call acceptable results) - they would in my opinion not even come close to S16 lenses.
Click to expand...

I don't see your point. If a lens looks sharp at 1080p, any softening of the image at 35mm would reflect the transfer, not the performance of the lens. If another lens did not perform as sharp at 1080p, it would not look as sharp at 1080p or blown up to 35mm...
 
Barry_Green said:
It has been screened on large screens and will be screened next month on a movie theater in NY, as far as I understand. Blowing up to 35mm and 4K doesn't do anything to add sharpness that wasn't already captured in the first place.
Click to expand...

4K doesn't of course add sharpness but reveals its (possible) lack more than 2K, that was my point. [/QUOTE]



Barry_Green said:
My point was -- seeing as this is the AF100 forum, the CP's are more than a match for anything the AF100 can deliver.
Click to expand...
Agree.
 
blackcat said:
I don't see your point. If a lens looks sharp at 1080p, any softening of the image at 35mm would reflect the transfer, not the performance of the lens. If another lens did not perform as sharp at 1080p, it would not look as sharp at 1080p or blown up to 35mm...
Click to expand...

How you perceive sharpness depends on the size of the image you are looking at and your distance to it. It is not absolute. In a movie theater image screen is usually huge and some people sit very close to it, in a distance smaller than diagonal of the screen.
 
broughtonfilm said:
4K doesn't of course add sharpness but reveals its (possible) lack more than 2K, that was my point.
Click to expand...
Not sure I follow though. When you project on a big screen, you're going to see all the detail that was recorded in the original signal. If you plug the camera directly into a 1080p projector and project it -- that's all there is. Uprezzing to 4K might make it look a little better because the uprez process might smooth out edges better than the 1080p projector does, I guess. Transferring to film won't improve anything, it'll instead be a degradation caused by the introduction of the film's grain and any transfer will degrade things... a digital projection from a 4K file from an uprezzed 1080p image is going to look sharper and clearer than a film transfer would (assuming, of course, the projector is up to the task).

Putting the AF100 image through an uprezzing process to a 4K projector is going to make it look better than if you just projected it directly. And we projected it directly, and I'm sure the NY screening is going to be a direct 1080p projection too.
 
Barry_Green said:
Not sure I follow though. When you project on a big screen, you're going to see all the detail that was recorded in the original signal. If you plug the camera directly into a 1080p projector and project it -- that's all there is. Uprezzing to 4K might make it look a little better because the uprez process might smooth out edges better than the 1080p projector does, I guess. Transferring to film won't improve anything, it'll instead be a degradation caused by the introduction of the film's grain and any transfer will degrade things... a digital projection from a 4K file from an uprezzed 1080p image is going to look sharper and clearer than a film transfer would (assuming, of course, the projector is up to the task).

Putting the AF100 image through an uprezzing process to a 4K projector is going to make it look better than if you just projected it directly. And we projected it directly, and I'm sure the NY screening is going to be a direct 1080p projection too.
Click to expand...

You are right, sure, sure - no benefit here.I am in the film world and 4K is always better there... However, transferring to film and introduction of grain helps. It masks any digital artifacts, aliasing, noise, etc. It becomes less digital. I experienced it many times Du Art lab in NYC has presentations of 35mm blowups from different formats and magic happens. Footage also gains little bit of color reproduction of light sensitive film stock - of course nothing that hasn't been captured is not magically brought back but... looks better...Hard to explain it rationally.
 
Barry_Green said:
Not sure I follow though. When you project on a big screen, you're going to see all the detail that was recorded in the original signal. If you plug the camera directly into a 1080p projector and project it -- that's all there is. Uprezzing to 4K might make it look a little better because the uprez process might smooth out edges better than the 1080p projector does, I guess. Transferring to film won't improve anything, it'll instead be a degradation caused by the introduction of the film's grain and any transfer will degrade things... a digital projection from a 4K file from an uprezzed 1080p image is going to look sharper and clearer than a film transfer would (assuming, of course, the projector is up to the task).

Putting the AF100 image through an uprezzing process to a 4K projector is going to make it look better than if you just projected it directly. And we projected it directly, and I'm sure the NY screening is going to be a direct 1080p projection too.
Click to expand...

Actually I know where I was coming from, we were talking about your Ultra 16 footage, not digital footage, so there is no plugging of the camera to the projector , it needs to be scanned. And there 4K will definitely be a tougher test and will reveal lack of lens's sharpness more than 2K scan. I wasn't clear in few posts up. I was talking about testing how ZF's performed on your Ultra 16 camera, not on AF100. I have no doubts they perform well on AF100 but have big doubts they perform well on Ultra 16.
 
Last edited:
broughtonfilm said:
You are right, sure, sure - no benefit here.I am in the film world and 4K is always better there... However, transferring to film and introduction of grain helps. It masks any digital artifacts, aliasing, noise, etc. It becomes less digital. I experienced it many times Du Art lab in NYC has presentations of 35mm blowups from different formats and magic happens. Footage also gains little bit of color reproduction of light sensitive film stock - of course nothing that hasn't been captured is not magically brought back but... looks better...Hard to explain it rationally.
Click to expand...
Oh, I totally agree. The grain texture just makes it feel more organic, smoother, and just ... more pleasing. I've transferred video to film before, and seen the effect firsthand. I was only puzzled about the idea that the film transfer would reveal more defects, when I think it would instead mask any defects, but I think we're on the same page now, right?
 
broughtonfilm said:
Actually I know where I was coming from, we were talking about your Ultra 16 footage, not digital footage, so there is no plugging of the camera to the projector , it needs to be scanned. And there 4K will definitely be a tougher test and will reveal lack of lens's sharpness more than 2K scan. I wasn't clear in few posts up. I was talking about testing how ZF's performed on your Ultra 16 camera, not on AF100. I have no doubts they perform well on AF100 but have big doubts they perform well on Ultra 16.
Click to expand...
Roger that, that makes sense. They may or may not perform as well as the best S16 glass when used on S16 film, I don't know yet, they are obviously one of the most recent designs but Zeiss doesn't rate them as high in sharpness as they do Master Primes, obviously. For an AF100 or other large-sensor 1080p system I think they're certainly enough, as to whether they'll deliver fully on the smaller S16 frame, that will need to be tested. It's on my list of things to do... :thumbsup:
 
Ryan Patrick O'Hara said:
The Cooke's were designed and advertised to work optimally at wide open and stopped down. With that said, I have yet to notice any aberrations or noticeable sharpness differences from wide open to stopped down. They are advertised to be just as sharp wide open as the cooke s4's when at the equivalent T/2.8 stop. I am wanting to do some very rigorous testing soon in regards to comparing them. If I could get my hands on a set of S4's for a day, I'd like to test. I'll be putting the 18mm through some tests when I get it, but I suspect and expect it to function the same.

I don't know about the CP.2's. I've played around with them but have never put them through any official tests.
Click to expand...

The cooke pachros also have an easter egg. If you light for a T2.8 and open the lens up wide, there is a small round "matte" that lets you achieve spherical bokeh at a measured T2.8, but if yo ustop down to T2.8+1/3, you get the cooke "flower" bokeh. So you have the options of circular, and traditional... at a nice working stop, at the lens' peak performance.
 
TimurCivan said:
The cooke pachros also have an easter egg. If you light for a T2.8 and open the lens up wide, there is a small round "matte" that lets you achieve spherical bokeh at a measured T2.8, but if yo ustop down to T2.8+1/3, you get the cooke "flower" bokeh. So you have the options of circular, and traditional... at a nice working stop, at the lens' peak performance.
Click to expand...

Yeah, wide open bokeh without having to be wide open at unreasonable apertures (1.4) is nice! Also, the iris assembly's in the Panchro/i's are the same as the Cooke S4's... so unless you're shooting at T/2.8, the bokeh will match exactly between the sets, meaning a 'mother-daughter' camera setup can have A-camera on S4's and B- camera on Panchro/i's and if at the same T-stop over T/2.8, they will have the exact same out of focus shape.

Also important if you own Panchro's but occasionally have to rent S4 lenses that are not in the Panchro/i focal range yet must match color, contrast and look... such as a S4 12mm, 40mm, 135mm+, etc.


PS: In regards to my quoted response from earlier, I did do the lens tests with a set of S4/i's and there wasn't a difference at all. Color, contrast, sharpness all matched. On a nano-scale I could barely see that the S4's had a pinch more contrast and were slightly warmer. When I say this, it was pixel peeping and quick juxtaposition of 4k images to even notice in the first place. At a casual look, I even missed it. The degree of difference was probably close to the degree of difference between a lens set, not to mention we are talking between two different lines of Cookes. Cooke S4's have always been known to be one of the most color matched lenses, and the Panchro's are no different, being from the same stock.

For those interested the findings and 4k images of the test can be found and downloaded from a thread here in DVXuser.
 
Last edited:
broughtonfilm said:
Thanks. That will be a great test. I think I know the results but seeing is believing.
Click to expand...

Sometimes the answer is so obvious that it hurts that I didn't think of it before. I mean, I've gone and shot some S16 with both the Compact Prime and the 16mm/S16 lenses, but -- now it becomes painfully obvious that I don't have to go doing a 4k scan of s16 film to find the answer. I've got a camera sitting here that can shoot 5k digital frames on its own - the GH2! Even though there's some cropping involved in S16 (the GH2's sensor is ~1.5x the width of S16 film) if I frame a 4K chart to fit within that 12.4mm width, I should have an easily-accessible 3K+ digital scan of the full S16-sized frame.

Accordingly, herewith, are some comparison shots of an Angenieux 17.5-70 Super16 lens, a Cinema Products 25mm Ultra T 16mm lens, and a 21mm Zeiss Compact Prime. All stopped down to f/5.6 for maximum sharpness. The Angeniuex is brand-new old-stock, the Ultra T was a premium lens for CP cameras designed to compete with the SuperSpeeds of its day, and the CP is basically a full-frame design rebuilt as a modern cinema lens.

http://dvxuser.com/barry/GH2/CP21mm-4k.jpg
http://dvxuser.com/barry/GH2/Angenieux-4k.jpg
http://dvxuser.com/barry/GH2/UltraT-4k.jpg


Yeah, this was a slam dunk and easy to determine. After reviewing the results of this test, I have no doubt that the Compact Primes (and the ZFs, for that matter) are definitely resolving at least as much as the dedicated S16 and 16mm lenses. And I'm pretty happy with how the Angenieux performed, it looks largely on par with a modern Compact Prime! The Ultra T doesn't measure up to the others, especially in the corners, but in the center it's on par.

With this test in hand, I don't think it's fair to say that vintage S16 glass is necessarily sharper than modern stills glass, when used for an S16 frame size. And most definitely not for when using the full m43 frame size.

Of course, this isn't comparing against a modern Ultra 16 lens, granted...

(and don't read those resolution numbers as being absolutely accurate... the chart was vaguely framed for what I figured would be a S16 frame crop out of the entire GH2 full 16:9 frame, and I got actually quite close... the GH2's frame size is approximately 1.53x wider than S16 (18.9 vs. 12.4mm); that would make the ideal S16 extraction about 3265 pixels wide. I did a little math and vaguely got it close, my S16 extraction ended up being 1.49x, or 3334 pixels wide. Not too bad for eyeballing it!)
 
You must log in or register to reply here.
Back
Top