fredmiranda.com Login
This topic is in the 2025 archive. You can read it and like its posts here. It is read only.

Archive 2025Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!

  
 
wastedimages
Offline
Details
p.3 #1Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
mawz wrote:
wastedimages wrote
mawz wroteI think it's pretty clear at this point that Cosina recomputes the optical design to account for the expected sensor filter stack for each mount (they're not the only ones, 7Artisans does the same for their FE+ versions of their M mount lenses).

So the same core design, but at least the rear group or element are tweaked to optimize for the mount in question.
So does that mean getting the VM mount version and adapting is a bad idea?
Yes, and repeatedly proved in testing for the wides and some (but not all) of the normals. The M versions and even the E versions are out performed by the native versions (and the same for E mount shooters, native E outperforms the M versions for wides and some normals).

Tele's generally work well regardless, so adapting M makes a lot of sense for >50mm and native for 50mm and shorter.


Damn, that is a shame, I might still adapt though as I have a Sony, will get a type 240 at some point and am currently using (and loving) a Nikon Zf for my vintage glass. Buying VM and adapting between the 3 cameras does make sense for me even at a slight loss in image quality
Jan 24, 2025 at 01:06 PM
Juha Kannisto
Online
Details
p.3 #2Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
Today Cosina announced the official (Japan) release date for the Z-mount version of this lens. It's 20th of February:

https://www.cosina.co.jp/news/%e3%83%95%e3%82%a9%e3%82%af%e3%83%88%e3%83%ac%e3%83%b3%e3%83%80%e3%83%bcnokton-28mm-f1-5-aspherical-z-mount-%e7%99%ba%e5%a3%b2%e6%97%a5%e3%81%ae%e3%81%8a%e7%9f%a5%e3%82%89%e3%81%9b/

Also, the U.S. price of E-mount version turned out to be $799 so most likely the Z-mount version would be priced very similarly. In U.S. 28/1.5 E was priced $100 below the new 90/2 VM ($899) that came out at the same time, whereas in Japan market 28/1.5 E is priced higher than 90/2 and higher than 28/1.5 VM Type I (which is $999 in U.S.). Interesting differences in pricing between different markets...
Jan 31, 2025 at 12:17 AM
Ripolini
Offline
Details
p.3 #3Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
In Italy, the price of the 28/1.5 E/VM I is € 1049 (including VAT):
28/1.5 E
50 Euro more than the 90/2 VM:
90/2 VM
28/1.5 VM II costs 1149 €:
28/1.5 VM II
I bet the Z-mount will be similarly priced.
Jan 31, 2025 at 03:09 AM
Jepser
Offline
Details
p.3 #4Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
Now I've placed a pre-order. Damn, addiction....
Jan 31, 2025 at 07:01 AM
stavrosf
Offline
Details
p.3 #5Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
Ripolini wrote:
In Italy, the price of the 28/1.5 E/VM I is € 1049 (including VAT):
28/1.5 E
50 Euro more than the 90/2 VM:
90/2 VM
28/1.5 VM II costs 1149 €:
28/1.5 VM II
I bet the Z-mount will be similarly priced.


I have preordered the Z version from the Greek distributor and indeed the price is 1049€.
Jan 31, 2025 at 08:33 AM
Ripolini
Offline
Details
p.3 #6Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
I have pre-ordered it just now
Same price.
Feb 14, 2025 at 12:51 PM
philip_pj
Offline
Details
p.3 #7Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
'Fred's review of 35/2 E-mount and VM versions clearly shows how the E-mount version is way better off-axis on Sony cameras'

Cosina made two optically different lenses in the VM 35/2 APO and the E-mount version of the same lens (the one Nikon users got later on).
The lens cut diagrams are identical for these two lenses, but the MTF charts don't lie. They clearly show the very different curves and overall performance wide open, with the E-mount vsn being around 5% stronger on-axis at f2, a pretty significant increase well beyond copy variation (you would hope) in a K8 device or similar.

So it's not surprising the VM would not match it, both used on a Sony. I think Cosina was trying to tame the 40 lpmm (detail) micro contrast a little in the VM, figuring perhaps that the M lens was going to see more use for people photography, street work, zone focusing at mid apertures.

'pretty clear at this point that Cosina recomputes the optical design'

Any doco for this? Hearsay, maybe? I've heard it too. They have never said more than using the words 'optimisation' and 'exclusive'. I'm not telepathic in Japanese. I think it's naughty of them not to show any changes in the lens block diagrams, That is much easier than going to the trouble of actually changing the glass formulations and/or the elements sizing, shaping and placement etc. Zeiss used to publish some great white papers, now there is no involvement with the user base. Learning anything is like getting blood out of a stone.

Most complex lenses take a year to develop and one reason why it takes so long is that it's not like changing cosmetics or ring designs.
The big one is quality assurance which takes a lot of testing by someone like the late Hubert Nasse, the spokesman for optical theory at Zeiss.

Any of what I would define as 'change' (a measurable and clearly defined physical alteration in the above-mentioned determinants of lens performance, for example) absolutely requires extensive QA not just in the production of new input components and assembly line processes, but the identification of unpredictable shifts in performance itself (MTF), bokeh, and particularly in aberrations inadvertently introduced that interfere with ray angles inside the lens.

Veiling glare, ghosts and flares in such a lens are not considered aesthetic as they are in many character and cine lenses, they are something to be eradicated. Nasse spoke of a vast range of carefully controlled and structured tests that had to be exhaustively undertaken with feedback and interaction with the design team.

He said they found a lot of this kind of problem in his line of work at Carl Zeiss (their real name, even now). He was not some narrowly academic scientist, but the guy who conducted the very challenging hands on work looking for shot situations that would interfere with the resulting images once in the hands of the user base.

Once the lens passed muster in his hands doing this battery of tests and possibly making further changes to the physical lens (which could also include coatings and/or further arranging elements to deal with the aberrations that had appeared). This 'invasive' level of optimisation would be akin to the minor reworking of existing lens designs such as the growing host of Leica M lens remakes and updates made by (say) the Chinese makers.

And Leica themselves, as in the early 20s remake of the Summilux-M 50/1.4 (CF), probably their most famous lens. Some of the derivative lenses actually appear to not just be approaching the technical and artistic performance of the original models, but improving on them. Or at least providing a similar level and style of image, but with their own flavor being a part of it.

Anyway, I am quite suspicious about this inferred mount optimisation process, so if any readers have detailed information of how the maker would proceed, what the steps are, and how long it takes, etc., I would like to see it. Genuinely interested, not being provocative here. It's a big gap in our knowledge and I like precision; a lot of lens design is cloaked in mystery, particularly the ingredients put into the (what I call) the 'house style' of the maker: the general rendering look of the brand or range. You won't be seeing much on the lens configuration and glass types, or even element counts, for the Zeiss Supreme Primes, all $192,000 worth of them. Not unusual, either.
Feb 15, 2025 at 12:26 AM
taildraggin
Offline
Details
p.3 #8Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
Yes! Zeiss gave you some info on the design case. Excellent.

Since these are still hosted on the 'new Zeiss" site, Cosina might be aware of this past goodness they inherited(?).

Zeiss CF Planar 100mm f/3.5

Zeiss CF Sonnar 150mm/f/4

Zeiss CF Planar 120mm f/4 Makro-Planar
Feb 15, 2025 at 07:06 AM
Keith B.
Offline
Details
p.3 #9Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
My hunch: The minor design tweaking required to change the curvature of field performance of a given lens for the difference in filter stack thickness between M, E or Z mount sensors would not show on the simple line drawings of lens element configurations provided by CV for our amusement.
Feb 15, 2025 at 01:33 PM
philip_pj
Offline
Details
p.3 #10Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
The reason for my take on what exactly is done for each mount is I read a lot of cine material. They use adapters all over the place, which I put down to the footage not requiring such a high level of IQ to permit close inspection because of constant motion.

DZO have a great diagram to show why M-mount has a very bright future in low and mid-end productions here, near the page bottom:

https://thypoch.com/en/simera-c

Just watching a video on Zeiss's Nano Primes, they give you a new mount so you can simply unscrew four bolts and swap out the mount. These are a six lens range selling for $26,000, and they do not do any optimization or optical recalculation of the kind suggested to be done by Cosina. Nanfos are very high end, corner to corner sharp so quality matters greatly here. Below see how they 'optimize' for a new mount:
(spacer for new page design gap)
..





five minutes of optimization on $26k lenses

Feb 16, 2025 at 12:43 AM
stavrosf
Offline
Details
p.3 #11Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
I am not a technical expert, but I will try to explain what I understand with a (great) bit of simplification.
At the film years the film had the ability to receive the light coming from the lens even on an angle. On digital era, the sensors do not work the same way. They need the light to be received as vertically as possible. Of course if you understand what an optic is this is much easier for the center of the frame and much more difficult on the corners. In order to handle the corners correctly for different sensor stacks it may need some handling by the lens manufacturer.

Let's go to the Zeiss Nano Prime. These are optics that have a much greater diameter and this means that even the corners are receiving the light (I guess very close to) vertically. So, this is the reason that you just change their own adapter and you are changing mount.

Keep in mind that the price to rent the Zeiss for a week is similar to the price to buy this Voigtlander.
Feb 16, 2025 at 01:26 AM
Ripolini
Offline
Details
p.3 #12Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
In this forum we have users who have reported observing different performance of (mostly wide-angle) E-mount lenses fitted on Nikon Z bodies with adapters from different manufacturers; the difference between the adapters was in their thickness. Adapters of inaccurate thickness induced worse corner performance of E-mount lenses on Nikon Z cameras.
Cosina states generically that ‘optics optimised for the image sensors of mirrorless cameras with the Nikon Z-mount are implemented’. This may simply mean that the distances between certain elements/groups within the optical scheme, and/or the distance of the rear element from the sensor, are slightly modified to compensate for the difference in sensor stack thickness. Under this assumption (it's just an hypothesis), lenses/groups employed in the optical scheme are the same independently of the mount (E, Z, VM); only their mechanical assembly could be modified/tweaked.
Feb 16, 2025 at 09:45 AM
stavrosf
Offline
Details
p.3 #13Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
The english translate of the official Cosina video of the lens on youtube (the one at the first post) is that "Cosina has optimized the performance by improving the design to suit each sensor". What is the possibility Cosina to change optical elements, but keeping the same formula in order to have better corner sharpness..
Feb 16, 2025 at 11:32 AM
glassartist
Offline
Details
p.3 #14Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
stavrosf wrote:
I am not a technical expert, but I will try to explain what I understand with a (great) bit of simplification.
At the film years the film had the ability to receive the light coming from the lens even on an angle. On digital era, the sensors do not work the same way. They need the light to be received as vertically as possible. Of course if you understand what an optic is this is much easier for the center of the frame and much more difficult on the corners. In order to handle the corners correctly for different sensor stacks
...Show more →

Why would film would have " the ability to receive the light coming from the lens even on an angle" when the goal was to have it lie flat against the 'platen'?
Feb 16, 2025 at 02:21 PM
mapgraphs
Offline
Details
p.3 #15Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
Ripolini wrote:
...
This may simply mean that the distances between certain elements/groups within the optical scheme, and/or the distance of the rear element from the sensor, are slightly modified...


This is actually something that one can see by trying a lens on adapters with different registration values. Some of it has to do with the effect of the field curvature spread as the lens is closer or further away from the correct registration distance. From playing around with this one can see that there's also the possibility that tweaking distances between a group or groups would also have an effect how the light rays align on the sensor plane.

I've never seen anything from Voigtländer about them changing the lens recipe for a new mount. I have found that poor edge performance with an M mount lens on an M body can be corrected on a mirrorless camera by using a short adapter, i.e. simply changing the registration distance until the edge definition improves. A lot of it does seem to depend on field curvature (aperture and point of focus being variables). Given the extremes one can encounter with lens asymmetry, it obviously doesn't take much to throw peripheral rendering off, so perhaps not much to alter it once the recipe is set...
Feb 16, 2025 at 06:17 PM
Ripolini
Offline
Details
p.3 #16Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
mapgraphs wrote:
I've never seen anything from Voigtländer about them changing the lens recipe for a new mount.


According to data published in Cosina website, the 50/2 Apo-Lanthar VM has a different design from that of 50/2 AL for E- and Z- mount. VM version has only two abnormal partial dispersion elements.


glassartist wrote:
Why would film would have " the ability to receive the light coming from the lens even on an angle" when the goal was to have it lie flat against the 'platen'?


Because film doesn't have microlenses on top of the emulsion. Both film and sensor must be flat, being (theoretically) flat the focus plane of SLR/DSLR/mirrorless cameras.

Feb 17, 2025 at 12:55 AM
mapgraphs
Offline
Details
p.3 #17Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
On the other hand, high performance lenses like the 75 Nokton f/1.5 or 40 Nokton f/1.2 are exactly the same for E, Z & M...
Feb 17, 2025 at 04:36 AM
Ripolini
Offline
Details
p.3 #18Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
mapgraphs wrote:
On the other hand, high performance lenses like the 75 Nokton f/1.5 or 40 Nokton f/1.2 are exactly the same for E, Z & M...


Sure. Same holds for the 35/2 Apo-Lanthar & 28/1.5 Nokton. However, the performance of the 28/1.5 VM adapted on Sony is visibly different from that of the E-mount version, as Fred's test has shown:
https://www.fredmiranda.com/forum/topic/1889861/

Feb 17, 2025 at 05:46 AM
glassartist
Offline
Details
p.3 #19Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
Ripolini wrote:
mapgraphs wroteI've never seen anything from Voigtländer about them changing the lens recipe for a new mount.
According to data published in Cosina website, the 50/2 Apo-Lanthar VM has a different design from that of 50/2 AL for E- and Z- mount. VM version has only two abnormal partial dispersion elements.

glassartist wroteWhy would film would have " the ability to receive the light coming from the lens even on an angle" when the goal was to have it lie flat against the 'platen'?
Because film doesn't have microlenses on top of the emulsion. Both film and sensor must be flat, being (theoretically) flat the focus plane of SLR/DSLR/mirrorless cameras.



Maybe I'm thinking about this wrong (which wouldn't surprise me), but shouldn't micro lenses help direct oblique rays down to the sensor surface? Or are you suggesting that unlike film where the film emulsion is on the surface, the micro lenses add depth and further offset the sensor?
Feb 17, 2025 at 01:49 PM
stavrosf
Offline
Details
p.3 #20Cosina Voigtländer Nokton 28/1.5 lens for Z-mount got announced!
BTW it seems that people in Germany started receiving their lenses (facebook posts and idealo.de availability site). I still waiting for mine (in Greece).
Feb 17, 2025 at 02:47 PM
FM ForumsNikon ForumJoin Upload & Sell
  
 

You are not logged inLog inRegister