R10 "Easy Predictions" from Erwin Putt

[AlanG]

Advertisement (gone after registration)

... Instaed of complaining about lens design I would rather complain about heavy recalculation of images inside the camera.

 

Well that's what digital photography is about - recalculating an image. Are you implying that somehow the depiction of the subject is being altered? Because it has been my experience that by "correcting" lens faults, the depiction of the subject becomes more accurate.

[AlanG]

As I previously stated, I see a net gain in quality when correcting for the lens faults.

 

Here is a 100% crop from the edge (not corner) of an image made with the Canon 24 TSE on a 5D. The lens has been shifted up a bit. One was processed in C-1 the other was in DxO. The principal difference here is the c/a correction, although for some reason, DxO also brings out a little more detail in the fine branches against the sky. (It does this even if I turn off DxO's c/a correction) Maybe C-1 could regain this detail if I made the sky darker. But the c/a is pretty bad and needs correction in my opinion. I used the DxO default setting here and it left a tiny bit of c/a. I've since gone back and manually adjusted it and DxO can remove the rest. The Image in DxO also had some geometric correction to make the building perpendicular - I shot handheld and was off a little. And I don't see any negative impact from this either.

 

All I'm saying is that I would not have anything against this being handled by the camera's firmware as I correct for c/a and other lens faults on nearly every photo.

 

This being an unsupported lens, there is no correction for vignetting or distortion. I'm not sure why c/a correction works on unsupported lenses but I guess it isn't so lens specific that a general c/a correction for a 24mm focal length will work and then one can always fine tune it.

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!

[EH21]

Getting back to the R10 or the hope for one, I'd rather see leica spend their resources on developing features such as a Full Frame sensor, a bigger brighter viewfinder, some kind of flash metering built in, and AF than put energy into software to correct final images. I do not need a smaller camera or want to replace a bunch of my R glass.

[EH21]

Yeah Alan,

What we need from Leica is also some tilt shift lenses so I can dump that sh*tty Ts-e 24 of mine. Honestly the reason I am here in this forum is that Leica lenses are better - they don't normally exhibit the kinds of problems that you are needing to correct with your canon body and lens - and I don't think its needed for a Leica camera like it is for some of the others. Show us an example of a shot taken with the DMR and a leica lens that needs such correction and then maybe I'd be interested to continue the discussion about software correction. If its not a problem with the DMR, I don't see it being a problem with a successor.

[AlanG]

. Show us an example of a shot taken with the DMR and a leica lens that needs such correction and then maybe I'd be interested to continue the discussion about software correction. If its not a problem with the DMR, I don't see it being a problem with a successor.

 

This discussion was not about current lenses but about new designs of cameras and lenses. The original post was about predictions for both the M and R series. Leica currently is using digital image correction in the M8. I and others gave examples to show how far along that technology is today. And some of the issues that need digital correction are not specific to the lens but may be a lens/sensor issue. Especially noticeable on full frame sensors.

 

So let's start with the assumption that all current Leica lenses are perfect and wouldn't benefit from any correction for distortion, vignetting, c/a or any other lens defect. So when it comes to the primes and current zooms, Leica may not need to do any more R&D. (But of course there is an issue with w/a vignetting on the M8 - corrected in firmware.)

 

But what about future designs? Isn't that what this thread is about? I think it is easier to make a well corrected 21-35 f3.5 lens than a 16-35 f2.8 or a 17-35 f2.8. Or a 35-70 f4 rather than a 24-70 f2.8. Let alone a 12-24 that covers full frame. (Sigma is alone here.)

 

Yes that 24TSE could be better, although it is pretty good after DxO. But nobody else is making a similar lens. Maybe it isn't very easy to do so.

 

Some photographers will not be interested in compromising their images for convenience by using fast wide range zooms instead of primes. But if Leica offers a choice, they will have more potential customers.

 

Isn't there a possibility that if you combine Leica's optical design expertise, with the best digital correction, Leica may come up with better solutions for the types of lenses I just listed?

[robsteve]

Isn't there a possibility that if you combine Leica's optical design expertise, with the best digital correction, Leica may come up with better solutions for the types of lenses I just listed?

 

 

Alan:

 

I posted earlier in this thread than the current DMR does lens corrections for ROM lenses. You may have missed this.

 

I think it may be the same corrections as advertised by Hasselblad since their image capture device is designed by Imacon, like the DMR.

 

These corrections are easily seen by taking a picture, and then taking the same picture with the ROM contacts of the lens taped over. The picture with the ROM corrections enabled will have at least vignette correction. I have not done any more testing to see if there is CA correction too.

 

Robert

[AlanG]

Advertisement (gone after registration)

Alan:

 

I posted earlier in this thread than the current DMR does lens corrections for ROM lenses. You may have missed this.

 

I think it may be the same corrections as advertised by Hasselblad since their image capture device is designed by Imacon, like the DMR.

 

These corrections are easily seen by taking a picture, and then taking the same picture with the ROM contacts of the lens taped over. The picture with the ROM corrections enabled will have at least vignette correction. I have not done any more testing to see if there is CA correction too.

 

Robert

 

This is good to know. I think all manufacturers will have no choice but to embrace this type of technology if they want to make cameras and lenses that are capable of making the best quality digital images.

[fotografr]

The first digital camera I bought, the Olympus C2000Z, about 8 years ago. It had a barrel distortion at the wide end but came with software that corrected it. I thought that was pretty neat for an early consumer digital camera.

 

The fact that a camera purchased just 8 years ago is now considered one of the "early" ones says a lot about this digital age in general. It is amazing how far the technology has come in such a short period.

[jrc]

The discussion of software and lenses was begun when somebody (too lazy to look back) suggested that Hasselblad was deliberately *not* building the best possible lens for a number of reasons, but making up the difference with software. That, I think, would be a mistake -- Lenses still count in the final image, and the only reason at all for shooting MF, and paying the price (perhaps $50,000 for one top end system, or $80,000 with two backs?) is image quality. If final image quality with a Canon, Nikon or Leica SLR matches Hasselblad, software correction or not, then Hasselblad is toast. Hasselblad needs to be the most and best of everything, because the SLR stuff is now so good, and the SLR software is so good, that there's not a lot of space left for MF. Hassy best make the most of what space is left...

 

JC

[AlanG]

The discussion of software and lenses was begun when somebody (too lazy to look back) suggested that Hasselblad was deliberately *not* building the best possible lens for a number of reasons, but making up the difference with software. That, I think, would be a mistake -- Lenses still count in the final image...

 

So tell me, just how bad are the pictures from the 28 on the H3 and 39 megapixel back? Do you know anyone else who makes a 28mm retrofocus lens (allowing for a mirror) that works on the 36x48 format?

 

I think Mamiya was going to make a 28 (or 24) for the ZD but never did.

 

 

Here is something that Hasselblad says about their digital correction:

 

For example, in the case of our new HC series 28mm lens, the only way we can get the lens to produce images of Hasselblad quality is by having an integrated system and utilizing our knowledge of our optics, our cameras, and our digital magazines, having them communicate with each other, and applying software within the system to compensate for issues that arise (such as lens distortion) and optimize the image.

 

What is “Ultra Focus” and why is it important?

 

A: Put very simply, light bends as it transits transparent materials (such as filters and covers used in front of CCD sensors). In addition, variations in aperture slightly change the focus point of the image on the sensor. Since we are very familiar with the filters and sensors we use in our H3D cameras, and since it is a DSLR, we have found a way to correct for this digitally, and we call this “Ultra Focus”.