Lens corrections equivalent to audio corrections when possible

[Tks]

is that what Nikon said when they introduced the 43-86 zoom?

Lmao

 

[Frank Dernie]

Zooms have no inherent detriments quality-wise compared to primes (in reality, zooms are far more R&D heavy on development costs by quite a bit).

Very, very, few zooms have boke competitive with primes. They usually have worse field flatness, depending on the set focal length.
It doesn't matter with wide angles simply because of depth of field but on longer lenses the differences are enough to, usually, rule out a zoom for my use.

Their resolving power even at 50MP+ resolutions aren't impacted.

The whole point of my original post was that the photographic community and press only seem to consider sharpness when there are other, equally important IME, lens characteristics which are hardly mentioned outside pro photographers who need them.
I did some tests with my printer and a loupe best part of 20 years ago, and found that 4 megapixels was enough, resolution wise, to produce an A4 (or 10x8") print indistinguishable from a higher resolution one.
Pro rata was also true, 16 megapixels for A3 - the biggest I print - and so on.
That means that the benefit of a camera with 50 megapixels and a lens with resolution to use them all ( and a solid enough support to achieve the potential) is purely for specmanship and for photographers who only look at tiny sections of their pictures on a screen at 100%.
It is pretty useful if you can't afford a long lens and need to crop heavily though .

OTOH if you have a lens with harsh boke it doesn't matter what resolution your cameras is, or how sharp it is, it will produce ugly photographs if there are any specular highlights in the background. Always, and on any camera. I have bought "sharp" recommended lenses which were completely useless for photography, for my use, simply because of this.

My original post was a lament for the one parameter obsession of sharpness in this internet age, when there are other, usually more important, parameters which either are never measured or never mentioned.

I observe that achieving sharpness from a lens is not a big problem if there is no limit to size and weight, nowadays there are a plethora of huge sharp lenses on the market. Mirrorless designs are easier because the "mirror area" can be used for elements.
Distortion is easily corrected, as coma and chromatic aberration, to an extent, in software.
Neither boke nor flatness of field can be, and the former is far more important than resolution in the vast majority of my, and IMO most other, photographer's pictures. As I bemoaned in my original post, we don't even have a reliable measurement metric to evaluate lenses in the lab for this crucial characteristic, other than the old Canon observation (in their lens works books) that lenses with sagittal and meridional mtf curves which closely overlay each other (they almost never do on zoom lenses) have good boke - no further information or explanation was given.
Field curvature is very subject dependant and rarely a problem for me.

Yes internal focus by small focal length changes are now common on autofocus lenses so many recent "primes" are indeed zoom lenses!

I still think photography on the internet is a "one trick pony" and the trick is not the only important one. Resolution is already way beyond what most photographers need, nice boke is still rare.

 

[Wes]

I think the reason "that the photographic community and press only seem to consider sharpness" in lenses is the wow effect from digital.
If you scan in your slides it is immediately apparent that that gear wasn't all that sharp.

Once this wears off, people will be searching for the next great thing.

BTW, Olympus has 3 different lenses specialized for great bokeh in their m43 line.

 

[Frank Dernie]

BTW, Olympus has 3 different lenses specialized for great bokeh in their m43 line.

The Olympus primes I have are all good for boke. The 150mm f2 is spectacular, 75 f1.8 and 300 f4 very good too.
I won't get the f1.2 models because they are big and heavy and, with digital, I don't need the speed or extra dof, not using Kodachrome 64 and 25 any more
I was tempted by the 45mm f1.2 but I already have super portrait lenses for FF cameras.

 

[paulraphael]

I still think photography on the internet is a "one trick pony" and the trick is not the only important one. Resolution is already way beyond what most photographers need, nice boke is still rare.

I can't say much about "photography on the internet" ... other than there's sure a lot of it. I will say, as a photographer and as someone with photographer friends (all of us doing it for art, not commerce) that you shouldn't make assumptions about what anyone's looking for from their tools. People have their esthetic vision, and they figure out what they need to make it happen.

For most of my own projects over the last couple of decades, sharpness has been important, and bokeh has been irrelevant. I wanted the kind of tactile sense of detail in my prints that traditionally came from 8x10 contact prints. And I was doing work that mostly didn't divide the world into "subject" and "background," so I didn't make use of selective focus.

I would not say it's easy to achieve this, especially if your prints get at all big. Most modern lenses are impressive on-axis, but in the corners suffer from field curvature (which sometimes actually helps, other times not), astigmatism, and who knows what else. Sometimes corners soften just a bit in an ok way; other times the detail smears in ways that can look quite ragged and distracting. I've got an architectural lens that I love, but I've found it has to be used at one of a couple of apertures, and at a certain magnification range, and within a small range of its shift capability, or everything goes to pieces. I typically have to sell lenses in order to buy the one I need for a project. Sometimes I'll rent one first to make sure it's good enough.

I would have bought a technical camera with 100+ megapixel PhaseOne back for one project—if could have remotely afforded it. Instead I use a pro-sumer dslr similar to what people use for work and hobbies totally different from my own. It's then a matter of figuring out what kind of glass will get it to do what I need. These cameras are remarkable jack-of-all trades tools. Which makes all the more important to remember that people are using them for many different things.

 

[charleski]

Yes internal focus by small focal length changes are now common on autofocus lenses so many recent "primes" are indeed zoom lenses!

Lenses have been doing this ever since they were first developed. It’s an inherent consequence of the thin lens equation. Keeping the focal length constant with changes in focus requires a far more complex lens, and this is one of the reasons that cine lenses are more expensive.

 

[JoachimStrobel]

I am not sure what is meant here. There is one company that measures lenses and provides corrections for their software to apply it: DXO with their Photolab. And that works and is great. But that is it. No lens has EPROMs which holds corrections to be applied during raw to jpeg conversion outside the cameras. And the internal camera jpegs that might use some lens correction stuff are what Spotify mp3s are for music compared to raw processing outside the camera. And there is no room, hence lens corrections do not have to care about a room as speakers have. And there is only one lens, not 2 to 7.

 

[paulraphael]

I am not sure what is meant here. There is one company that measures lenses and provides corrections for their software to apply it: DXO with their Photolab. And that works and is great. But that is it. No lens has EPROMs which holds corrections to be applied during raw to jpeg conversion outside the cameras. And the internal camera jpegs that might use some lens correction stuff are what Spotify mp3s are for music compared to raw processing outside the camera. And there is no room, hence lens corrections do not have to care about a room as speakers have. And there is only one lens, not 2 to 7.

I'm not sure if I understand what you're saying here. Adobe camera raw and lightroom software both support lens profiles from a large database. They can automatically correct distortions and lateral chromatic aberration, and can reduce the effects of longitudinal chromatic aberration. Nikon's own raw processing software includes these features (with profiles only for supported Nikon lenses). Hasselblad's newer generation lenses are designed from the ground up to be corrected with lens-specific profiles in their raw processing software. This means they can ignore distortion and certain chromatic aberrations almost entirely, and put their engineering efforts into aberrations that can't be fixed in processing. This seems to me very much like dsp correction in speakers—not for correcting room problems, but for correcting cabinet resonances and shortcomings of drivers, crossovers etc.

 

[Wes]

DXO has better corrections than Adobe, etc. according to what I've heard. Also, CaNikon don't release all the lens correction data to anybody outside their co.

OTOH, at some point you are picking nits for non-military or NASA lenses. F/8 and be there is still very true.

 

[paulraphael]

DXO has better corrections than Adobe, etc. according to what I've heard. Also, CaNikon don't release all the lens correction data to anybody outside their co.

Maybe, I don't know. Lightroom's corrections do what I need them to do. If straight lines look straight to me, and fringing is gone without noticeable artifacts, then I'm not going to be too hungry for improvements. There are other qualities in a raw processor that are more important to me, including the workflow and the quality of the tools.

 

[Frank Dernie]

I will say, as a photographer and as someone with photographer friends (all of us doing it for art, not commerce) that you shouldn't make assumptions about what anyone's looking for from their tools.

Sorry, I was referring to my needs, I realise that a landscape photographer, for example, is more sensitive to full field sharpness and, to an extent flatness of field, which are of little importance to me.
It is just that sharpness is like SINAD, important but by no means the whole story.
I do take a few landscapes and here I do want sharp all over and no vignetting.

 

[Frank Dernie]

I can't say much about "photography on the internet"

I worded this badly too.
I was not referring to photography but equipment discussion!
A very different thing
Like in "hifi" where there is a long wide spectrum of enthusisats with people being music lovers at one end and equipment perfectionists at the other, "Photography" ranges from people who just want to get the picture they want to people who spend all their time on test shots and measurement graphs.
It was the real value of the latter I took issue with.

 

[paulraphael]

...ranges from people who just want to get the picture they want to people who spend all their time on test shots and measurement graphs.

I suspect a lot of people are just looking for an excuse to collect more cool stuff. You see it any hobby that involves nice shiny things...

And some people just love to geek out. A friend of mine wanted to start a satirical photography forum that was all about art critiques of people's histograms.

 

[JoachimStrobel]

I'm not sure if I understand what you're saying here. Adobe camera raw and lightroom software both support lens profiles from a large database. They can automatically correct distortions and lateral chromatic aberration, and can reduce the effects of longitudinal chromatic aberration. Nikon's own raw processing software includes these features (with profiles only for supported Nikon lenses). Hasselblad's newer generation lenses are designed from the ground up to be corrected with lens-specific profiles in their raw processing software. This means they can ignore distortion and certain chromatic aberrations almost entirely, and put their engineering efforts into aberrations that can't be fixed in processing. This seems to me very much like dsp correction in speakers—not for correcting room problems, but for correcting cabinet resonances and shortcomings of drivers, crossovers etc.

Some people commented, that these lenses are only useable with corrections. Meaning, for speaker we would end with a scenario where each driver is connected to an outboard DSP that does the crossover. Why not. But one would need a feedback loop with a mic, like the Echo studio. So each speaker pair comes with a remote wireless microphone, all cabled up with a DSP. Yes, that could work. Except, the discussion on the correct room curves fills hundreds of posts already. At least the lens industry knows what they are aiming at.

 

[paulraphael]

Some people commented, that these lenses are only useable with corrections. Meaning, for speaker we would end with a scenario where each driver is connected to an outboard DSP that does the crossover. Why not. But one would need a feedback loop with a mic, like the Echo studio. So each speaker pair comes with a remote wireless microphone, all cabled up with a DSP. Yes, that could work. Except, the discussion on the correct room curves fills hundreds of posts already. At least the lens industry knows what they are aiming at.

The lens corrections we're talking about don't require any kind of feedback loop. They're just based on knowing the transfer function of certain lens aberrations at different apertures and different angles, and reversing them.

This works for linear degradations, like rectilinear distortion and lateral color. It doesn't work for nonlinear ones like flare. It would be my assumption that diffraction is also nonlinear and irreversible, but that post about the Canon software claims otherwise. I don't know if that's for real or not.

 

[JoachimStrobel]

Let me try to organize my thoughts:
- Most Boomboxes have a Dsp to get the maximum out of their geometry and driver.
- Some (I only know the Amazon thing) have an additional mic that checks if the DSP does it good and changes things.
- I am not aware of widespread use of such DSP stuff in loudspeakers. Neumann seems to use them.
- Then there is the area of Room correction software with a mic and some target curve.

For lenses, there is a process where a DSP looks at the lenses picture and then tries to tweak results knowing the lens design and shortcomings by working on the raw data, probably including special features of the camera bodie’s Bayer sensor. But outside the military there is no adaptive optics where lenses are manipulated like drivers in a speaker. The said corrections work well, and they seem to work best with DXO and PL where one can actually look at a database of lens performance and can infer the correction.
The mic equivalent for photos would be the color analyser that is used to calibrate monitors, printers and/or camera responses. This then is very similar to room correction including the discussion on target curves which means colorspace and temperature in this context.

I agree that Stereophile grade loudspeakers could use more DSP input than endless tweaking of driver position, crossover stuff and enclose optimization. My fear is, that it will lead to cheap design efforts, meaning: The DSP will fixit all. I guess the best would be a mix of good design and DSP. Or the Echo studio approach will sweep the market and Stereophiles are doomed.

 

[111db]

Some people commented, that these lenses are only useable with corrections. Meaning, for speaker we would end with a scenario where each driver is connected to an outboard DSP that does the crossover. Why not. But one would need a feedback loop with a mic, like the Echo studio. So each speaker pair comes with a remote wireless microphone, all cabled up with a DSP. Yes, that could work. Except, the discussion on the correct room curves fills hundreds of posts already. At least the lens industry knows what they are aiming at.

"God, grant me the serenity
to accept the things I cannot change,
the courage to change the things I can,
and the wisdom to know the difference."

An important concept for design engineers! Especially those with marketers breathing down their necks.

 

[curiouspeter]

I think the reason "that the photographic community and press only seem to consider sharpness" in lenses is the wow effect from digital.
If you scan in your slides it is immediately apparent that that gear wasn't all that sharp.

Once this wears off, people will be searching for the next great thing.

BTW, Olympus has 3 different lenses specialized for great bokeh in their m43 line.

Olympus was great when they had the only 5-axis IBIS in the world. What was achievable with a handheld OM-D was quite amazing when it came out. Now that practically everybody offers it, Olympus is no longer that special.

A Fuji 35mm f/1.4 has very nice bokeh. It is cheaper and lighter (half the weight!) than an MZuiko 25mm f/1.2 too.

That said, I prefer m43 to film up to 645.

 

[Wes]

I think the biggest similarity between lenses and audio is the "magical thinking" that goes on in the high end of both hobbies. The "Gestalt" of an unubtainium Leica lens .

Do you have much experience with the special rendering of certain Leica lenses?

In some circumstances the photos taken with them can easily be distinguished - in a blinded testing.

That is not the same thing as determining mechanistically what factors cause that, just as an epidemiological result is not the same thing as determining mechanistically what factors cause the disease or disfunction.

This may help:

Leicaness Leica Look

https://www.artphotoacademy.com/the-leica-look/

Leica Glow - a localized haze around highlights; residual spherical aberrations

- out of focus areas & colors that ‘pop’

high contrast

high levels of detail where it is needed

pleasant softness in areas that do not require much attention

ability to see the whole tonal range effortlessly.

Deep 3D look

Positive Bokeh - balls are brighter in the middle and darker closer to the outside of the shape

very high micro-contrast - register a nearly full variety of tonal variations between slightly darker and slightly brighter areas of very similar colors; gives rich colors and smooth tonal transitions to produce a three-dimensional “feel”

and no, it is not accurate - nor is a wah wah pedal

 

[curiouspeter]

Do you have much experience with the special rendering of certain Leica lenses?

In some circumstances the photos taken with them can easily be distinguished - in a blinded testing.

That is not the same thing as determining mechanistically what factors cause that, just as an epidemiological result is not the same thing as determining mechanistically what factors cause the disease or disfunction.

This may help:

Leicaness Leica Look

https://www.artphotoacademy.com/the-leica-look/

Leica Glow - a localized haze around highlights; residual spherical aberrations

- out of focus areas & colors that ‘pop’

high contrast

high levels of detail where it is needed

pleasant softness in areas that do not require much attention

ability to see the whole tonal range effortlessly.

Deep 3D look

Positive Bokeh - balls are brighter in the middle and darker closer to the outside of the shape

very high micro-contrast - register a nearly full variety of tonal variations between slightly darker and slightly brighter areas of very similar colors; gives rich colors and smooth tonal transitions to produce a three-dimensional “feel”

and no, it is not accurate - nor is a wah wah pedal

I thought the "Leica Look" is what Leica owners see in their minds.

After paying so much money, I can trick my mind into seeing anything.