p.22 #1 · ZEISS Otus ML line officially announced!
zeitlos wrote:
This is interesting. I am always blown away by the 50mm GM‘s bokeh. And I do have a lot of good glas. But I don‘t know the Nikon 50mm 1.2. However, here they say the exact opposite of what you wrote => The Bokeh can be busier with the Nikon 50mm 1.2.
I cannot see an example of that busy bokeh with foliage in the video, however. The comparison should have also included contrast, distortion, and focus breathing in the score sheet.
p.22 #2 · ZEISS Otus ML line officially announced!
So tell me what manufacturers list specific camera models for which a lens is optimized and then imply that performance will be suboptimal on other cameras by the same manufacturers.
I don't see manufacturers providing that information, if you find an example of this, I suspect it's an exception...not the rule.
AcuteShadows wrote:
I think this is correct. In many ways, lenses, sensors and automatic electronic processing interact. Sensor stack thickness, sensor microlens design, sensor resolution all have an impact on the result of a given lens/sensor combination. Thus, a lens that is designed to be perfect on one particular sensor may not be perfect on a different sensor, and you actually should buy lenses for the sensors that you actually use, not for potential future sensor designs.
p.22 #3 · ZEISS Otus ML line officially announced!
tsdevine wrote:
So tell me what manufacturers list specific camera models for which a lens is optimized and then imply that performance will be suboptimal on other cameras by the same manufacturers.
I don't see manufacturers providing that information, if you find an example of this, I suspect it's an exception...not the rule.
No, they generally don't provide that information. You have to find out yourself. In my experience, Nikon lenses have never produced subpixel color moiré on the Nikon sensors I use. But I don't know if this is by design or by chance.
One example of a manufacturer that actually does comment on this is Leica, which states that lenses made for analog cameras are not optimal for digital cameras. There is also quite a number of comparisons of Leica lenses on different Leica cameras available.
p.22 #4 · ZEISS Otus ML line officially announced!
AcuteShadows wrote: [...]
If sharpness/acuity is not tailored to the sensor resolution, then you run the risk of having color subpixel moiré which is hard to correct and there are, to my knowledge, not automatic corrections for this. As an example, the Nikon 50mm f/1.2 S does no do this on the Z 7 sensor, but the Zeiss Milvus 25mm f/1.4 and the Leica Summicron 28mm can produce that kind of color moiré.
Jonas B wrote: I don't understand how that would work. I guess the ""sharpness" is tailored a little every time i move the focusing ring and when the distance to anything in the front of the camera changes....
Trying to search the net for this, to me, new phenomena gave nothing related to common photography. If this is something real, as opposed to marketing material or fanboy wishes, I would like to learn about it.
AcuteShadows wrote: Lenses are not actively tailored to a sensor (as far as I know). What I mean is that if sharpness exceeds the resolution of a sensor, then you may encounter adverse effects. You would need to choose the combination of sensor and lens, though.
Thank you for getting back on this.
Unfortunately I still don't see how this works. The claim "[lens] sharpness exceeds the [sensor] resolution" ► "adverse effects" contains several undefined factors:
(1) In a way nearly all decent lenses are capable of "sharpness" exceeding the sensor resolution somewhere in the image space. The opposite is also true, somewhere or overall (depending on focusing and aperture opening. It's not like a lens is "sharp" at the exact right level over the whole image. (That's like DxO old ratings...)
(2) Adverse effects? Is that the subpixel moiré you mentioned?
Is there a more technical explanation explaining about subpixel moiré somewhere you can direct me to?
p.22 #5 · ZEISS Otus ML line officially announced!
My understanding is moire / false color shows up when the texture of your subject is sharper than your sensor resolution. Which is why lower MP cameras, which are more likely to show moire have a low pass filter that slightly blurs the image to prevent false color.
If your lens is not sharp enough to fully resolve the details in the pattern, you will also not see moire. This is why one one of the ways to help fix moire is to steep your lens down into the diffraction zone, such as f11, to soften the image.
I got some (it was rare) images with my Nikon D800e and 14-24mm f2.8 G that had moire artifacts. Meanwhile sharper lenses like my Sigma 50mm f1.4 Art I don't remember ever getting moire with. I feel it has more to do with your subject than your lens.
p.22 #6 · ZEISS Otus ML line officially announced!
Jonas B wrote: AcuteShadows wrote: Jonas B wrote: AcuteShadows wrote:
Thank you for getting back on this.
Unfortunately I still don't see how this works. The claim "[lens] sharpness exceeds the [sensor] resolution" ► "adverse effects" contains several undefined factors:
(1) In a way nearly all decent lenses are capable of "sharpness" exceeding the sensor resolution somewhere in the image space. The opposite is also true, somewhere or overall (depending on focusing and aperture opening. It's not like a lens is "sharp" at the exact right level over the whole image. (That's like DxO old ratings...)
(2) Adverse effects? Is that the subpixel moiré you mentioned?
Is there a more technical explanation explaining about subpixel moiré somewhere you can direct me to?
Indeed, most decent lenses will outresolve the sensor in the sense that the image (at least in some area, usually the center) could be improved noticeably by using a sensor with higher resolution. Some lenses, will, for example, induce stepping artefacts with nearly vertical or nearly horizontal high contrast edges. But not all decent lenses induce what I call "color subpixel moiré".
Sensor pixels, as recorded in an image, are calcuated from color (sub-)pixels, which are present on the actual sensor. If a lens is sufficiently sharp, a bright spot (or high contrast edge) may touch one subpixel, but not the adjacent ones. Then, the resulting pixel is calculated with uneven color information, and will thus display a color that is not present in the object space. I consider this an adverse effect. My term for this is "color subpixel moiré", but maybe there is a more exact expression out there.
Some lenses are actually quite evenly sharp across most of the frame. But if not, then a loss of sharness in some area will not result in color subpixel moiré (but will result in some softness of the image, of course).
p.22 #7 · ZEISS Otus ML line officially announced!
Jonas B wrote:
[...]Unfortunately I still don't see how this works. The claim "[lens] sharpness exceeds the [sensor] resolution" ► "adverse effects" contains several undefined factors:
(1) In a way nearly all decent lenses are capable of "sharpness" exceeding the sensor resolution somewhere in the image space. The opposite is also true, somewhere or overall (depending on focusing and aperture opening. It's not like a lens is "sharp" at the exact right level over the whole image. (That's like DxO old ratings...)
(2) Adverse effects? Is that the subpixel moiré you mentioned?
Is there a more technical explanation explaining about subpixel moiré somewhere you can direct me to? ...Show more →
AcuteShadows wrote:
Indeed, most decent lenses will outresolve the sensor in the sense that the image (at least in some area, usually the center) could be improved noticeably by using a sensor with higher resolution. Some lenses, will, for example, induce stepping artefacts with nearly vertical or nearly horizontal high contrast edges. But not all decent lenses induce what I call "color subpixel moiré".
Sensor pixels, as recorded in an image, are calcuated from color (sub-)pixels, which are present on the actual sensor. If a lens is sufficiently sharp, a bright spot (or high contrast edge) may touch one subpixel, but not the adjacent ones. Then, the resulting pixel is calculated with uneven color information, and will thus display a color that is not present in the object space. I consider this an adverse effect. My term for this is "color subpixel moiré", but maybe there is a more exact expression out there.
Some lenses are actually quite evenly sharp across most of the frame. But if not, then a loss of sharness in some area will not result in color subpixel moiré (but will result in some softness of the image, of course).
Hi again,
So, now we agree that most decent lenses "outresolve" the [any] camera sensor in some area if focused correctly and we aren't using a small aperture opening.
We can also agree that we from time to time see color artefacts when enlarging the raw image sufficiently.
Now I assume that we from now on talk about sensors with no low pass filter in the sensor stack.
What the heck is a lens tailored to a sensor [resolution] now? I'll simply claim there is no such thing.
I checked the image in your link. There we have color artefacts all right. I think I would simply call them moiré. Do we really need many fine lines side by side to call the artefact for moiré? When we get the color artefact, as in your sample, from only one line of high contrast it's pretty pure - when we have a repeating pattern we get more colors but the background and idea is the same. This is what I think.
Hence we don't really need a new name. All moiré is a result from subpixel calculations.
Please note: This is in no way my area of expertice. I may be completely wrong and I'm happy to be corrected. I have now made two claims
(1) no lens is tailord to any sensor
(2) "color subpixel moiré" is really just (the beginning of) moiré. EDIT: Thinking about it, maybe just call it aliasing.
I think with current bayer sensor technology we basically have to choose between having no detail smaller than 4 pixels or dealing with some false color. Maybe we will get to the point where sensors out resolve lenses, but then I think we will have people complaining about diffraction.I keep hopping Sigma will succeed at making a FF foveon sensor. In theory, a foveon sensor should have no false color, help diffraction issues, and have better color accuracy.
p.22 #10 · ZEISS Otus ML line officially announced!
I wrote this to point out that, in my opinion, given a certain sensor, an addition increase in lens sharpness becomes less and less useful, and may even have adverse effects.
I don't want to propose any new nomenclature, so you may call it moiré, or aliasing, or some other term.
While to my knowledge, no lens is tailor made for a single sensor model, I assume that some lenses are made for a particular class of sensors. For example, Leica lenses can be made smaller because they a supposed to work with Leica sensor with very low sensor stack thickness. On the other hand, Fuji GFX lenses are, to my knowledge, optimized for the thicker sensor stack of sensors used in Fuji GFX cameras (with the benefit of reducing the effect of dust on the sensor, especially at smaller apertures).
p.22 #11 · ZEISS Otus ML line officially announced!
AcuteShadows wrote:
I wrote this to point out that, in my opinion, given a certain sensor, an addition increase in lens sharpness becomes less and less useful, and may even have adverse effects.
A kind of diminishing return yes, also when not counting the dollars.
AcuteShadows wrote:
I don't want to propose any new nomenclature, so you may call it moiré, or aliasing, or some other term.
OK. But it was the term "color subpixel moiré" that had me confused and wondering at first.
AcuteShadows wrote:
While to my knowledge, no lens is tailor made for a single sensor model, I assume that some lenses are made for a particular class of sensors. For example, Leica lenses can be made smaller because they a supposed to work with Leica sensor with very low sensor stack thickness. On the other hand, Fuji GFX lenses are, to my knowledge, optimized for the thicker sensor stack of sensors used in Fuji GFX cameras (with the benefit of reducing the effect of dust on the sensor, especially at smaller apertures).
Yes, so it is. As a Sony user I'm well aware of the problems using wide angle M-mount lenses on my cameras The sensor stack is too thick causing smearing at the corner with these lenses. But that is basically a totally other thing than discussing artifacts due to a lens being too "sharp".
Anyway, it's good for me to have it sorted out. Thank you for your patience explaining!
p.22 #12 · ZEISS Otus ML line officially announced!
AcuteShadows wrote:
While to my knowledge, no lens is tailor made for a single sensor model, I assume that some lenses are made for a particular class of sensors.
For this discussion it might be interesting to note that there are few cameras that might actually have tailored lenses for their sensors: Sony RX1 and Zeiss ZX1.
The RX1 series have used off-the-shelf sensors which Sony also uses for their A7 cameras. But Zeiss ZX1 was advertised to have specific 'Zeiss designed' sensor, whatever it means in practice, but nevertheless it's a sensor (37mp) we haven't seen in other products. Both of these cameras have their lenses tailored to their sensor. With the RX1 not perhaps more than having optimal size and image quality, but with the ZX1 seems that each single sensor unit was calibrated individually and the Distagon 2/35 lenses were hand built (not mass produced) where they could provide much tighter tolerances (which translate to better performance). The Distagon 2/35 was, considering its small size (especially entry pupil), something we have not seen before. It provides an Otus level performance (practically APO), and I believe it's because the lens was tailored to specific sensor (and they didn't have to account somewhat loose tolerances which come with the interchangeable systems).
p.22 #14 · ZEISS Otus ML line officially announced!
NatDeroxL7 wrote: ??
Taking a guess, B&H discontinued the wrong Otus lenses. I got the email notice indicating the discontinuance of the new ones. Still see them on the Zeiss Otus site so sent a message to B&H to confirm it is not a mistake.
p.22 #15 · ZEISS Otus ML line officially announced!
burningheart wrote:
Taking a guess, B&H discontinued the wrong Otus lenses. I got the email notice indicating the discontinuance of the new ones. Still see them on the Zeiss Otus site so sent a message to B&H to confirm it is not a mistake.
Ah, I'm pretty sure it's a kind of mistake coming from the fact that the lens isn't available yet. At some point the sign will be exchanged into a "Coming soon!" message.
p.22 #16 · ZEISS Otus ML line officially announced!
Hopefully this is just a mistake. Adorama still lists them as available. Otherwise, this may become the most disappointing lens release of all time.
Mar 21, 2025 at 02:37 PM
Steve Spencer Online • • • • • • • Upload & Sell: On
p.22 #17 · ZEISS Otus ML line officially announced!
DWOfPaul wrote:
Hopefully this is just a mistake. Adorama still lists them as available. Otherwise, this may become the most disappointing lens release of all time.
It isn't schedule to be released until late in the year, so what is surprising is that they have a listing for it at all. I wouldn't make anything of the discontinued listing. It certainly isn't coming soon, so it is hard to know what to call. "Coming in six months or so," or "coming at some point," or "eventually available" would be a lot more accurate, but not really any more helpful.
p.22 #18 · ZEISS Otus ML line officially announced!
Steve Spencer wrote:
It isn't schedule to be released until late in the year, so what is surprising is that they have a listing for it at all. I wouldn't make anything of the discontinued listing. It certainly isn't coming soon, so it is hard to know what to call. "Coming in six months or so," or "coming at some point," or "eventually available" would be a lot more accurate, but not really any more helpful.
Acording to Adorama the 50mm is suposed to ship in May 5th and the 85mm is suposed to ship in October 23rd. So atleast for the 50mm we are only talking about 1.5 months away.
Mar 21, 2025 at 03:12 PM
Steve Spencer Online • • • • • • • Upload & Sell: On
p.22 #19 · ZEISS Otus ML line officially announced!
DWOfPaul wrote:
Acording to Adorama the 50mm is suposed to ship in May 5th and the 85mm is suposed to ship in October 23rd. So atleast for the 50mm we are only talking about 1.5 months away.
Yes, that is right. The picture with the discontinued label was of the 85, so I was talking about the 85. Sorry I wasn't more clear.
??
