p.37 #1 · Zeiss "Batis" AF lens line to be released for FF E-mount
TT, I wonder whether there is not some confusion here. The 85mm Batis, which may or may not be a Tamron or Zeiss/Tamron effort, has nothing to do with Sony. Similarly, the lenses badged "Sony-Zeiss" have much more to do with Sony than with Zeiss. Now Tamron may be cooerating with both design teams, each one on a 85mm, but I have not heard of a FE 85 so far, except the Batis...
p.37 #2 · Zeiss "Batis" AF lens line to be released for FF E-mount
philber wrote:
TT, I wonder whether there is not some confusion here. The 85mm Batis, which may or may not be a Tamron or Zeiss/Tamron effort, has nothing to do with Sony. Similarly, the lenses badged "Sony-Zeiss" have much more to do with Sony than with Zeiss. Now Tamron may be cooerating with both design teams, each one on a 85mm, but I have not heard of a FE 85 so far, except the Batis...
i agree to certain extend, but my view is that i really dont think there are that many expert/engineers available in this industry. especially tamron/sony relationship are rather close, so i believe there are some strong sharing relationship between the two companies i.e. i wouldn't be surprised if all three worked together on this lens then slap a zeiss badge on it.
p.37 #3 · Zeiss "Batis" AF lens line to be released for FF E-mount
Didn't SAR recently post/link to an interview in which a Sony designer explained that they had toyed around with the idea of a 85mm f1.2, but along the way realized that they just couldn't come up with a design that was sufficiently small. It would be interesting to know what happened afterwards. Did they decide to to do a f1.4 design instead or leave out the 85mm focal length and go for 135mm, because of the Batis? Then again, we have three 35mm lenses, so two 85mm may well be realistic.
p.37 #4 · Zeiss "Batis" AF lens line to be released for FF E-mount
Jman13 wrote:
3% pincushion is really high. If the lens is sharp enough to still be razor sharp after distortion correction, it could be a minor issue, but it's troubling that Sony and Zeiss are relying so much on digital correction. The FE 28/2 has horrible barrel distortion too, though it is sharp enough to make it generally a minor point once corrected, but I'd really like to see more optical correction done. Somehow Fuji manages to do a very good job with optical distortion correction on their lenses, even with wide-angle primes like the 14mm and 16mm, both of which are practically perfectly corrected optically for distortion....Show more →
p.37 #5 · Zeiss "Batis" AF lens line to be released for FF E-mount
Yes, some do, though it doesn't appear to be widespread. However, most of the ones that do are zooms. The primes are all very well corrected. Almost bizarrely so.
One thing to note is that Fuji bakes the distortion correction into the RAW file itself, so if you use LR or C1, you'll never see the uncorrected distortion, which is nice from a workflow standpoint, though annoying if you want to change it. However, Photo Ninja doesn't see that correction data, so I open the files in PN when evaluating distortion on Fuji lenses (and m4/3 lenses).
It's just a trend I've noticed...some E-mount lenses are plenty corrected optically, but a lot have pretty severe distortions. The Sony 18-105/4 is by far the worst offender, and the FE 28/2 is a great lens with massive barrel distortion, which ultimately affects corrected corner resolution due to the stretching. Now the Batis 85 has pretty strong pincushion. It's not a dealbreaker, but it is something that isn't just distortion on its own, as it ultimately affects sharpness of the final image as well. (I'm still interested in the Batis 85, BTW).
It's just a trend that's happening with these lenses, and it's something I wish they would focus on a bit more due to how it affects other imaging characteristics.
p.37 #6 · Zeiss "Batis" AF lens line to be released for FF E-mount
Jman13 wrote:
Yes, some do, though it doesn't appear to be widespread. However, most of the ones that do are zooms, and even there it's generally not correcting massive amounts of distortion. For instance, the Fuji 16-55/2.8 has software corrected barrel distortion at the wide end, but even uncorrected, it's better than the FE 28, despite being a zoom lens. That distortion is really only present between 16 and 18mm too. The primes are all very well corrected. Almost bizarrely so.
One thing to note is that Fuji bakes the distortion correction into the RAW file itself, so if you use LR or C1, you'll never see the uncorrected distortion, which is nice from a workflow standpoint, though annoying if you want to change it. However, Photo Ninja doesn't see that correction data, so I open the files in PN when evaluating distortion on Fuji lenses (and m4/3 lenses).
It's just a trend I've noticed...some E-mount lenses are plenty corrected optically, but a lot have pretty severe distortions. The Sony 18-105/4 is by far the worst offender, and the FE 28/2 is a great lens with massive barrel distortion, which ultimately affects corrected corner resolution due to the stretching. Now the Batis 85 has pretty strong pincushion. It's not a dealbreaker, but it is something that isn't just distortion on its own, as it ultimately affects sharpness of the final image as well. (I'm still interested in the Batis 85, BTW).
It's just a trend that's happening with these lenses, and it's something I wish they would focus on a bit more due to how it affects other imaging characteristics. ...Show more →
Yeah, this is a very disappointing trend and kudos to Fuji for mostly staying with optical correction. My slowly growing enthusiasm for Sony has diminished severely after hearing about how poorly the lenses are on their own.
p.37 #7 · Zeiss "Batis" AF lens line to be released for FF E-mount
rattymouse wrote:
Yeah, this is a very disappointing trend and kudos to Fuji for mostly staying with optical correction. My slowly growing enthusiasm for Sony has diminished severely after hearing about how poorly the lenses are on their own.
I actually had a lot of fun with the X-pro1, X-E1 and the X100 cameras.. A few quirks I wasn't to happy with, but overall a joy to use. Some amazing JPG files from all three cameras, surprisingly.
As far as corrections are concerned. I'm not to sure. But that 35/1.4 was a pretty killer lens.
If Fuji ever makes a X100 with a larger sensor (FF or medium format), I'd be itching to play with it.
p.37 #8 · Zeiss "Batis" AF lens line to be released for FF E-mount
TrainTicket wrote:
i agree to certain extend, but my view is that i really dont think there are that many expert/engineers available in this industry. especially tamron/sony relationship are rather close, so i believe there are some strong sharing relationship between the two companies i.e. i wouldn't be surprised if all three worked together on this lens then slap a zeiss badge on it.
In the past though there has been a distinct separation between the Zeiss branded lenses directly sold and marketed by Zeiss and those sold and marketed by Sony. The Batis line falling into the former camp (sold directly by Zeiss), it would seems odd to me that Sony have any direct design/ manufacturing connection but who knows...I believe the Zeiss AF lenses sold directly prior (Touit) were said to perhaps be Fuji sourced. Maybe these lines are becoming more blurred and, in the end, the brand/ badge is just marketing.
p.37 #10 · Zeiss "Batis" AF lens line to be released for FF E-mount
Tariq Gibran wrote:
In the past though there has been a distinct separation between the Zeiss branded lenses directly sold and marketed by Zeiss and those sold and marketed by Sony. The Batis line falling into the former camp (sold directly by Zeiss), it would seems odd to me that Sony have any direct design/ manufacturing connection but who knows...I believe the Zeiss AF lenses sold directly prior (Touit) were said to perhaps be Fuji sourced. Maybe these lines are becoming more blurred and, in the end, the brand/ badge is just marketing.
What would you make of the ZE/ZF lineup then?
I'm a huge fan of quite a few older Zeiss lenses. But I think the last lenses I really loved from Zeiss were the 50MPZE and 100MPZE. Who designed those, let's bring those guys back!
BTW, you know how small the 50MPZE is? If you take away the macro feature, it's optics including spacing from sensor make it one TINY TINY LENS. The 100MPZE being bigger, isn't really all that big either.
Why not remove the macro feature and put those in batis cases!
Optical issues aside, I love the rendering of those fellows.
p.37 #11 · Zeiss "Batis" AF lens line to be released for FF E-mount
lightskyland wrote:
The bottom line for me is the file after corrections.
That's what you see in your file and that's what you compose with in the viewfinder.
That's what I want Sony and the lens designers to focus on, not on things they had to do in the film era.
What exactly are you paying for then when you spend money on a lens that requires software correction to be fully useable? Suppose that the lens costs $1000 or more (perhaps much more like a Zeiss Otus lens). Is it really worth that large amount of money if the glass cannot produce a useable image but needs software correction?
p.37 #12 · Zeiss "Batis" AF lens line to be released for FF E-mount
What exactly are you paying for then when you spend money on a lens that requires software correction to be fully useable?
Unless you are planning to use a lens on a film body, you will be using very extensive software transformations to achieve any images with a lens. RAW conversion, deconvolution etc. are all extremely sophisticated software operations, without which a digital camera and lens are nothing but heavy, expensive lumps of glass, metal and plastic.
Suppose that the lens costs $1000 or more (perhaps much more like a Zeiss Otus lens). Is it really worth that large amount of money if the glass cannot produce a useable image but needs software correction?
You are paying for resolution, bokeh, well controlled LoCA and other characteristics that can't be achieved via RAW file deconvolution and software transformation. As well as build, handling, AF etc.
I certainly don't want engineers wasting trade-offs in the hardware that are better achieved in software, like trading image sharpness for lower distortion!
If someone is going to use a lens on a film camera, I can see caring about what the projected image looks like at the focal plane. If not, I think it's just an irrational prejudice against modern photographic design.
p.37 #13 · Zeiss "Batis" AF lens line to be released for FF E-mount
I'm not sure how blind people are if they don't mind software distortion correction. I find it extreme annoying that on fine textures the contrast reduces toward corners artificially - it does that already on almost all lenses naturally (even Otus level lenses), and when you enhance it by extra interpolations due to
a) CA-removal (=scaling color channels independently)
b) distortion correction (=stretching and compressing areas of picture)
it just gets more and more annoying. Fine textures take hit even when just rotating image to straighten horizon, but reduction is same overall on whole image in that case, looks less artificial. Maybe when we have 100-200mpix sensors these extra pixel interpolations become irrelevant, but with current sensors I find it extreme annoying.
Luckily for me only photography, where I care about image quality, is nature photography and neither CA or distortion cause rarely issues and I don't need digital microcontrast raping done to them. Hopefully I don't get into cityscapes or architecture...
From my perspective Zeiss made a good compromise for getting flat field on Batis 25 at cost of wave type distortion (max 1.5%). The most annoying thing in ZE/ZF series 2/25 is the field curvature (very strong curve in corners of image, which @ infinity won't be fixed even by using f/11). It seems that Batis is free from field curvature and can be used for landscapes already at f/4 if DOF is sufficient. I just hope we get them before summer ends...
p.37 #14 · Zeiss "Batis" AF lens line to be released for FF E-mount
I love it when the statement is made "what are we paying for" You are paying for a matched system lens and camera working together for the best possible IQ that the engineers can put together for the price. Sony lenses aren't designed for brand x or y but for the Sony E or A mount.
p.37 #15 · Zeiss "Batis" AF lens line to be released for FF E-mount
adamdewilde wrote:
What would you make of the ZE/ZF lineup then?
I'm a huge fan of quite a few older Zeiss lenses. But I think the last lenses I really loved from Zeiss were the 50MPZE and 100MPZE. Who designed those, let's bring those guys back!
BTW, you know how small the 50MPZE is? If you take away the macro feature, it's optics including spacing from sensor make it one TINY TINY LENS. The 100MPZE being bigger, isn't really all that big either.
Why not remove the macro feature and put those in batis cases!
Optical issues aside, I love the rendering of those fellows....Show more →
I'm pretty sure those are pure Zeiss designed and Cosina manufactured. Great lenses and solidly made (as are the Zeiss ZM's and even many of the Voigtlanders). Cosina is underrated as a lens manufacturer imo.
p.37 #16 · Zeiss "Batis" AF lens line to be released for FF E-mount
Tariq Gibran wrote:
I'm pretty sure those are pure Zeiss designed and Cosina manufactured. Great lenses and solidly made (as are the Zeiss ZM's and even many of the Voigtlanders). Cosina is underrated as a lens manufacturer imo.
That's what I thought... But now I'm skeptical
Anyway, I don't mind Cosina at all. They're fully capable of making a good lens. I've never had problems with the ZE lineup. I've had a few problems with ZM lenses (loose barrels etc). The only thing I hate is the silver ring on the end of the lens.. What a dumb idea!
p.37 #17 · Zeiss "Batis" AF lens line to be released for FF E-mount
Samuli Vahonen wrote:
I'm not sure how blind people are if they don't mind software distortion correction. I find it extreme annoying that on fine textures the contrast reduces toward corners artificially - it does that already on almost all lenses naturally (even Otus level lenses), and when you enhance it by extra interpolations due to
a) CA-removal (=scaling color channels independently)
b) distortion correction (=stretching and compressing areas of picture)
it just gets more and more annoying. Fine textures take hit even when just rotating image to straighten horizon, but reduction is same overall on whole image in that case, looks less artificial. Maybe when we have 100-200mpix sensors these extra pixel interpolations become irrelevant, but with current sensors I find it extreme annoying.
Luckily for me only photography, where I care about image quality, is nature photography and neither CA or distortion cause rarely issues and I don't need digital microcontrast raping done to them. Hopefully I don't get into cityscapes or architecture...
From my perspective Zeiss made a good compromise for getting flat field on Batis 25 at cost of wave type distortion (max 1.5%). The most annoying thing in ZE/ZF series 2/25 is the field curvature (very strong curve in corners of image, which @ infinity won't be fixed even by using f/11). It seems that Batis is free from field curvature and can be used for landscapes already at f/4 if DOF is sufficient. I just hope we get them before summer ends...
That's why you want a high resolution camera... Most of the pixel-level detail you get from a Bayer-based-camera conversion is a bad estimation anyway. It has very little to do with the reality in front of the lens. *It might "look" good, but there's no target correlation. See older Leica cameras, large sensors with low resolution and no AA filters, often shot with stellar lenses. Looks sharp, has moire, has very little reality-correlated detail.
Linear transforms have some HF losses by design. That is because we're mixing up two totally different image representation methods when we look at an image "on screen".
The image is stored and is calculated on as if it was a purely point-sampled set of data. We must do this since working with square cutout samples of the image (which is what the sensor records and the monitor shows...) is very much harder computationally. There's a conflict between the sampling theorem and the reality of looking at images-on-screen.
To avoid making the problem worse and inducing artefacts in the image, most resampling schemes are on the "safe side" of HF detail. You see what happens if you don't do this if you look at incorrectly sampled thumbnails of large images, they're full of aliasing, moire and strange stuff that shouldn't be there.
Having a high resolution camera, and applying the transforms on the original resolution to downsample the size as one of the last steps in the processing chain is actually the most effective way we know of and can apply from a practical PoV right now if you want objective data. If you want the images to actually represent what was in front of the camera at the point of capture...
Distortion is correctable, lateral CA is somewhat correctable in post. Yes, you lose HF contrast, but that wasn't really any "real" HF contrast to start with.... What you lose is mostly digital artefacts, synthetic constructs. I prefer downsampling those away anyway (downsampling is effectively a low-pass filter... Scaling an image 1:2 leaves you with about the same image detail remaining as applying an r1 Gaussian blur). Then do the output sharpening as the last step of the PP.
Larger lenses, more expensive lenses and field curvature, well, you just have to live with that. There's not much you can do in software / PP to correct those problems... Distortion - well, as long as your camera has more resolution than you need, you're fine.
p.37 #18 · Zeiss "Batis" AF lens line to be released for FF E-mount
rattymouse wrote:
What exactly are you paying for then when you spend money on a lens that requires software correction to be fully useable? Suppose that the lens costs $1000 or more (perhaps much more like a Zeiss Otus lens). Is it really worth that large amount of money if the glass cannot produce a useable image but needs software correction?
If you have two $1000 25 f2 lenses for two different systems and they produced identical fully usable images only one used more software correction than the other but you had no way of knowing which one ----- which lens would you buy?
p.37 #19 · Zeiss "Batis" AF lens line to be released for FF E-mount
RobCD wrote:
If you have two $1000 25 f2 lenses for two different systems and they produced identical fully usable images only one used more software correction than the other but you had no way of knowing which one ----- which lens would you buy?
If you had 2 identical pairs of shoes, that cost you exactly the same, but one is genuine leather and the other is fake leather, which one would you wear?
p.37 #20 · Zeiss "Batis" AF lens line to be released for FF E-mount
edwardkaraa wrote:
If you had 2 identical pairs of shoes, that cost you exactly the same, but one is genuine leather and the other is fake leather, which one would you wear?
Would I be able to tell which one is genuine leather and would you be able to tell ? If yes then I would wear the genuine leather. If no, then I wouldn't care because I wouldn't know anyway.
Distortion - well, as long as your camera has more resolution than you need, you're fine.