s2 and d3

[AlanG]

Advertisement (gone after registration)

I was surprised to see you as the author of this statement, Alan. Correcting a corner cast (colour values) is pretty light-duty, compared to Hasselblad's 28mm corrections (moving pixels around), for example, or the noise reduction going on in Canons at higher ISOs, leading to a very processed, smeary, low-noise look. Did you mean it in some sort of limited way?

 

I don't see how noise reduction has anything to do with the subject of optics. The best I know, Hasselblad applies correction to one lens that was designed with this correction in mind. The M8 applies vignetting and corner color correction to a number of lenses. And often requires one to shoot through a filter. Does it make a difference if the software correction was planned from the beginning or an afterthought?

 

I bet there is no lens that could not benefit from some type of software correction as I am not aware of any lens that has zero distortion, zero aberrations, and zero vignetting. What happens if the 24mm or the 30mm TS for the S2 will work better with some software correction?

 

Let's say you are a purist and are only going to work with very well corrected prime Leica lenses. So if you find yourself in a situation where shooting quickly over a wide range of focal lengths would be a benefit, would you be willing to use a zoom lens that has software correction if that would give you the best chance at getting the images you need?

 

The overall point is that to me the idea of some kind of "pure" image that is not technological is kind of an irrelevant fantasy. When shooting with w/a lenses on my view camera, I use a center filter to help reduce vignetting. When I dodged and burned a print, I was either making a creative decision or correcting for lens vignetting. When I tilted the printing easel, I either corrected or distorted the image.

 

In photography, technology and creativity are inextricably inter-related. I think that most photographers will use whatever technology they are comfortable with that gives the desired results. What advancements in technology does is keeps expanding your options, whether you avail yourself of those options or not. I don't see too many photographers who limit themselves to shooting Daguerreotypes. So when it comes to photographic technology, why would one arbitrarily draw a line somewhere?

 

I had a professor who called 35mm "miniature format." His idea of a reasonable camera was a 5x7 Linhof and he would never have accepted an assignment shot with a Leica. He drew the line at 4x5.

[wildlightphoto]

I bet there is no lens that could not benefit from some type of software correction as I am not aware of any lens that has zero distortion, zero aberrations, and zero vignetting.

 

Alan, when is the last time you used a Leica APO lens?

[AlanG]

Alan, when is the last time you used a Leica APO lens?

 

Oh I forgot. Every Leica lens is perfect. And there would be no point in Leica making a lens that is not perfect (let's say a wide angle zoom) that can be made nearly "perfect" by software. Because that just wouldn't be right. Even if that was the best way to get the image you want.

 

So lets say your APO telephoto could be made a bit sharper, or faster, or smaller, or lighter, but the trade off would be a tiny bit more distortion that could be corrected automatically by the camera's firmware? You wouldn't want that would you?

 

And what if the camera is aware of the focus shift that occurs when stopping down a lens. Should the camera be allowed to adjust focus for this? What if focus correction technology was part of the lens system itself? Sort of like a floating rear element but a shift that is connected to the aperture. Is that OK? Does it have to be mechanical, or can it be controlled via software?

 

And surely every lens for the S2 will also be perfect. No distortion, no aberrations, no vignetting, no focus shift when stopping down.. And what else? No MTF curve? Just a straight horizontal line, for the very first time ever. On all the lenses. At every aperture.

[sdai]

My thanks to Leica for posting these MTF charts, it proves the APO 180/2.8 R still can't be beaten - after 10 years!

 

Bring on the R10!

Welcome, dear visitor! As registered member you'd see an image here…

Simply register for free here – We are always happy to welcome new members!

[dfarkas]

My thanks to Leica for posting these MTF charts, it proves the APO 180/2.8 R still can't be beaten - after 10 years!

 

Bring on the R10!

 

You may have to adjust your scale a bit. The 180 APO-Elmarit-R (which I have always loved, especially on the DMR) MTF chart represents performance on a 24x36mm frame. The 180 APO-Elmar-S MTF chart represents performance on a 30x45mm frame.

 

So the question is, which would be better on the R10? The 180 R lens will mount on the R10 directly. The 180 S lens will most likely mount via an adapter. This will be an interesting match-up.

 

David

[sdai]

You may have to adjust your scale a bit. The 180 APO-Elmarit-R (which I have always loved, especially on the DMR) MTF chart represents performance on a 24x36mm frame. The 180 APO-Elmar-S MTF chart represents performance on a 30x45mm frame.

 

 

Nothing needs to be adjusted.

 

The horizontal axis is frame size, the R's MTF chart used a radius of 20mm while the S lens' chart uses 27mm.

 

If you contract the R's chart horizontally to fit its edge at about 0.75 of the S chart's x axis. The R curve is still much flatter than the S curve. In fact, the S curve is not flat at all beyond 0.2, which is 5.4mm from the image center.

[wildlightphoto]

Advertisement (gone after registration)

Oh I forgot. Every Leica lens is perfect.

 

Alan, instead of a knee-jerk sarcastic reaction, can you answer the question?

[carstenw]

I don't see how noise reduction has anything to do with the subject of optics.

 

You are right, I just posted it as an example of non-voluntary processing which yields, IMO, unacceptable results. Note that the corrections in the M8 are voluntary. If you don't like them, switch them off.

 

The best I know, Hasselblad applies correction to one lens that was designed with this correction in mind. The M8 applies vignetting and corner color correction to a number of lenses. And often requires one to shoot through a filter.

 

I think Hasselblad made a mistake with that lens (btw, isn't the coming zoom also meant to require software corrections?). It is expensive enough to put it out of reach for some H3D users, yet it doesn't cover 645 completely, and will thus be compromised for the announced FF-645 sensors. In my mind, correcting colour values is acceptable, if the benefit is tangible, but moving those values around as a base correction means a net loss in sharpness, and is thus not acceptable.

 

I suppose Hasselblad decided to make this lens simply because they wanted their users to have an option to not have to move to LF with a DB. I think that the lack of coverage for FF-645 sensors was due to the failure to anticipate them. In the end, the 28mm lens is for Hasselblad's mid-level users, those rich enough for it, but not for those who buy the largest, FF-645 sensors (unless they simply crop away the edges; I believe it will still be a 31mm lens, or something similar).

 

To recap, the need for corrections on the M8 has two reasons: the IR debacle (which Leica failed to properly anticipate), and the fact that digital sensors add their own vignetting. I think that the latter is quite acceptable. The former is not, but accepting that it exists, I am still happy to have an M8. The M9 with FF-35 sensor and no IR problem will be my dream camera, especially if Leica can approach the size and feel of the original M6.

 

I bet there is no lens that could not benefit from some type of software correction as I am not aware of any lens that has zero distortion, zero aberrations, and zero vignetting.

 

That is just a tautology, as you are aware. But the acid test is not which lenses would benefit, but which aren't good enough to be used without. In that area Leica scores very highly.

 

The primary difference between Hasselblad and Leica in our example is that Leica is correcting deficiencies in the camera, whereas Hasselblad is correcting deficiencies in the lens. Given that bodies move through technological improvements at a much higher rate, I consider Leica's approach acceptable, but Hasselblad's not.

[Overgaard]

One could also say that today's camera manufacturers has inherited the film makers problems and challenges in how to capture correct colors in less light, with less noise.

 

As Carsten rightly states, that is an in-camera correction. It's a matter of how the "film" works. And I concur that there's a difference between correcting the lens and the "film" using software.

 

Todays way to manage how the "film" captures and relays the image is no longer a matter of chemicals but of software.

[georg]

I expect the quality of 39MP-backs (+ noise from 4 years more sophisticated CCD and microlenses) + Schneider Digitar / Rodenstock HR - Lenses from the S2. Everything less would be a disappointment! But when they achieve this quality, they are visibly superior to Hasselblad/Fuji!

 

The MTFs of the S-Lenses are superior to the MTFs from Rodenstock/Schneider too, but we should be really careful comparing them from different manufacturers at all!

 

Hasselblad is dead to me, letting the better build-quality of the 200/500 series die, moving production to Fuji, bringing new lenses/bodies which are incompatible to bigger sensors or other bodies (what was the impvrovement from H1 to H2 to H3 within just 6 years?) and reducing the swedish staff from over 400 to 60 (the R&D-team of the S2 is bigger)!!! That isn't Hasselblad anymore...

[sdai]

bringing new lenses/bodies which are incompatible to bigger sensors or other bodies (what was the impvrovement from H1 to H2 to H3 within just 6 years?) and reducing the swedish staff from over 400 to 60 (the R&D-team of the S2 is bigger)!!! That isn't Hasselblad anymore...

 

You simply keep on repeating these non sense, Georg.

 

Keep in the mind the smallest Hasselblad sensor is still bigger than the biggest Leica sensor, and why would Hasselblad make new lenses/bodies compatible to other bodies?

 

Don't forget Leica isn't supporting their old but superior R system too.

 

What Leica is doing with the S system is purely wasting time IMO. Get a decent R working and they can run away with existing R customer bases and attract many more from Canon, Nikon camps.

[AlanG]

Alan, instead of a knee-jerk sarcastic reaction, can you answer the question?

 

Me??? If anything, your response was a knee-jerk reaction as it wasn't relevant to anything I posted. I presume you are implying that you don't think your lenses could benefit from any type of software correction. Good for you. Do you also want to deny Leica any opportunity to have software correction in its bag of tricks when designing new lenses?

[AlanG]

That is just a tautology, as you are aware. But the acid test is not which lenses would benefit, but which aren't good enough to be used without. In that area Leica scores very highly.

 

 

You just said that if you turn off the M8's correction you'll end up with vignetting and corner color shift on w/a. So in what way is it scoring highly? And this is using a cropped sensor which already is a major compromise in regard to the original design specs of lenses. My original post is that Leica is a manufacturer that applies more optical correction than others. I think that is true. I only posted in response that Leica is not so "pure." Not that I care if they are or aren't.

------------------------

Originally Posted by overgaardcom

 

Leica is so pure and natural. No nano-stuff or artificial software-fixing of lens capture. Photography is about light, not about technology.

__________________________________

 

A lot of people seem pleased with the results they get frame a variety of cameras without using software correction. That doesn't take away the fact that software correction can improve the results of almost any camera/lens combo and thus offers new opportunities to lens designers.

 

So the bottom line is, if you have a Leica lens that is better than most other lenses but still not "perfect," is there something wrong with using software to try to make it closer to "perfect?"

[georg]

"and why would Hasselblad make new lenses/bodies compatible to other bodies?"

 

They introduce crop-lenses (36x48mm) for a system designed for 41x56mm (e.g. film and the new H3D-60), they make bodies only a few years old non workable with certain lenses and backs! I don't expect them to open their system to other manufacturers, but I expect a new system (H1 was introduced just ~6 years ago!) to be consistent.

 

The S is a different system and cannot be fully compatible with the M/R by nature, and I think their skills (mechanical & optical quality) are more appreciated in the professional studio/fashion/commercial-market. They can make a R10 based on the S2 pretty easy, but they propably only had this one chance to step into the MF-market.

 

"is there something wrong with using software to try to make it closer to "perfect.""

 

Nothing is wrong with that, new technologies, new opportunities, I think we all agree with that! But this technology can also be used to cut production costs by simply don't designing the lenses to a professional standard (the 35-90 is sold for 6000€!!!) and this is the wrong direction, it always starts pretty harmless and then becomes worse over time.

 

Isn't this direction obvious when comparing a 200 series Hasselblad with the H-System?

[sdai]

They introduce crop-lenses (36x48mm) for a system designed for 41x56mm (e.g. film and the new H3D-60), they make bodies only a few years old non workable with certain lenses and backs! I don't expect them to open their system to other manufacturers, but I expect a new system (H1 was introduced just ~6 years ago!) to be consistent.

 

What Hasselblad is doing is not different from many others are doing, not at all.

 

If cropped sensor is good enough for you, you can buy the 2 HCD lenses and use them on a cropped sensor camera such as the H3D-31. If you want 645 full frame, Hasselblad makes 9 full frame lenses and I doubt many photographers can afford them all!

 

but they propably only had this one chance to step into the MF-market.

 

IMO they have no chance at all. It's all about money/affordability, 30x45 can never beat 36x48 or full frame 645 in terms of image quality.

 

There's only one reason why many folks haven't bought into digital medium format and that's about money.

 

Those who have waited and can not afford either a Hasselblad or Mamiya digital medium format are very unlikely going to be able to afford a Leica S. Simply as that, end of story.

[carstenw]

You just said that if you turn off the M8's correction you'll end up with vignetting and corner color shift on w/a. So in what way is it scoring highly?

 

You won't get any argument from me about the cyan shift corrections, that was a mistake, but it was not typical Leica, but a beginner mistake. The vignetting doesn't concern me, and actually I like a little. My point is not that Leica lenses are perfect, because they aren't. My point is that the (modern) lenses are so well designed that correcting the rest is strictly optional and a point of taste, not necessity. One might claim that the corrections on the Hasselblad are also optional, but they sure aren't marketed that way. Hasselblad seems to market it as an actual advantage, which I find annoying. I don't think that there is anything wrong with programs like DxO but I think that the lens company's work isn't done until the lens can stand on its own, and any further improvements are minor and optional.

 

Simon, the point here is that is no Hasselblad 28mm non-D lens. You are stuck if that is what you want for your new H3D-50. This is a mistake from Hasselblad. The same for any other D lenses they make in the future. My gut feeling says that there won't be another. The D-lens concept was Hasselblad's strategy du jour for yesterday. The 50/60MP sensors killed it.

 

By the way, I am another Hasselblad V fan. I wish they had updated the 203FE/205FCC with whatever the H has, and gone in that direction. That would have been my dream camera.

[AlanG]

 

"is there something wrong with using software to try to make it closer to "perfect.""

 

Nothing is wrong with that, new technologies, new opportunities, I think we all agree with that! But this technology can also be used to cut production costs by simply don't designing the lenses to a professional standard (the 35-90 is sold for 6000€!!!) and this is the wrong direction, it always starts pretty harmless and then becomes worse over time.

 

 

Maybe these lenses should not sell for so much. I think once photographers showed they were willing to pay $30,000 for a digital back, the lens companies decided to get in on the action too.

 

It is very hard to decide on the actual value of camera gear. (Vs. what it costs to make and market.) Phase One is re-selling the Hartblei 45mm TS lens at a much higher price than Hartblei sold the lens. What changed? Now this $700 - $900 lens is a $4,000 lens. I guess it is a lot better now. Other companies do the same thing when they re-brand. It makes me wonder why other lenses cost so much too.

 

But let's look at software correction of lenses again. I had a job not too long ago where a small part of it was to photograph some old framed newspaper ads that were at my client's office. In the old days, I would have used a well corrected macro lens and possibly a copy stand. Now I used a zoom lens (which had major barrel distortion and vignetting) and DXO which automatically made the image distortion free and even in the raw conversion process. And DXO has a feature where you mark the corners of the object and you'll get a perfectly squared up image.

 

So I just shot them roughly squared up and got them perfect later. They probably ended up being tiny images on some web site or corporate piece. It wasn't as if I was trying to do museum quality repro. But they were surprisingly good non-the less. What is wrong with using convenient zoom lenses and/or lower priced optics if they can be corrected properly for the job at hand? I am no longer a purist as I was when I shot everything on 4x5 or larger.

 

But if I am going to invest in high end m/f gear, the only reason to do it would be to get a very high level of quality from the back and optics. (Along with the wide angle performance I need of course.) And if these optics can be made even a tiny bit better through software correction or micro focus compensation, that would be the way to go for sure. Otherwise, they might not consistently be enough better than 35mm with DXO to be justifiable on many projects. And surprisingly few of my jobs or clients are demanding enough to benefit from MF. (I've worked my way down from 4x5 to 6x9 to 24x36.)

[sdai]

Simon, the point here is that is no Hasselblad 28mm non-D lens. You are stuck if that is what you want for your new H3D-50. This is a mistake from Hasselblad. The same for any other D lenses they make in the future. My gut feeling says that there won't be another. The D-lens concept was Hasselblad's strategy du jour for yesterday. The 50/60MP sensors killed it.

 

By the way, I am another Hasselblad V fan. I wish they had updated the 203FE/205FCC with whatever the H has, and gone in that direction. That would have been my dream camera.

 

28mm on 645 full frame is equivalent to 17mm on a 35mm full frame, Carsten. Back in the good old CF days, the widest Zeiss for the V system was the Cfi 3.5/30 so my guess is it must be pretty hard to build a 4/28 to cover 645 or 6x6 full frame, especially when considering to deal with the sensor issues of a digital capture.

 

I agree with you, Hasselblad probably will never bother with another HCD design again now they have the 4/28 and 35-90 already, just like Nikon's DX and Canon EF-S assortment, enough to cater the cropped sensor flocks.

 

We'll probably never know why Hasselblad broke up with Zeiss ... but that's not important, if you must stick to the V system, Phase, Leaf still build backs fitting the 200/500 series, the CFV back is pretty good too.

[georg]

I think the way professional consumers will react to these systems with their different approaches (high-res 35mm, S2, closed MF-Systems, open MF-System...) will be crucial for the way the market develops.

 

Do we want full-frame sensors (56mm width) or entirely new formats (30x45mm)?

Do we want digital APO (the name...) or expensive, advanced lens design/manufacturing?

Do we want modular or closed systems?

How important are durability and mechanical quality of bodies & lenses?

...

Some of these aspects will lead to different systems for different demands, others will most likely lead to certain solutions with nearly no other choice left.

 

Do we buy overpriced, re-branded lenses? Then we get more of them!

Do we buy a body that has a just two years old predecessor but isn't capable (software) of using different backs/film but is the only chance to use new lenses with some kind of revolutionary software correction? So we propably get a new body every two years (well, 2008 we got no new H4D...)!

Do we accept more and more software correction? Then we get more!

Do we accept cheaper materials for body/lenses? Well, someday that's all what you can buy for all the money in the world...

 

I simply feel that this is a time where customers have to decide which route they want to take. There are many new, cool technologies - but do we also want that certain qualites disappear? Like AF 20 years ago: a new cool technology with lots of new possibilites, but customers ("in exchange") also accepted loose plastic focus mechanics and dark, small viewfinders...

 

"30x45 can never beat 36x48 or full frame 645 in terms of image quality"

In theory, with the same lenses, yes. But practical, a new system, new sensor and new lenses of this quality can compensate a smaller sensor (= higher magnification) to a certain level. From the S2-samples I have seen (not the small web-pix on their site) the S2 offers the same IQ as a 39MP-back + Schneider/Rodenstock - that's why my expectations are so high.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[sdai]

From the S2-samples I have seen (not the small web-pix on their site) the S2 offers the same IQ as a 39MP-back + Schneider/Rodenstock - that's why my expectations are so high.

 

Most people using a 39MP back + Schneider/Rodenstock combo are folks on serious landscape, architecture or food stuff utilizing a large format cameras tilt and shift ability, which the S2 doesn't have.

 

I have no idea why someone into large format photography would have the slightest interest in the S2 but that's just me.

 

The higher your expectations have, the bigger disappointment it will be.