To stir the pot....

[Djarum]

So, as we know, panny is getting ready to release the new 14mm lens. This thread at dpreview poses a few questions:

Poor quality of new 14mm f2.5 lens [Page 1]: Micro Four Thirds Talk Forum: Digital Photography Review


Basically, it appears that the lens has quite a bit of distortion without software correction.

So the question is, do you as a consumer care?

Is the quality of the lens determined by the format that the lens is attached to only?

My own thoughts are is that if I'm not interested in using the lens outside of the format and the results are good, thats fine by me. If I expect a lens purchase to hold value as a tool in the future for other formats, I want the lens as good as it can be without software correction.

Secondly, the lens in and of itself is only as good as the actual optic, which means that judging a lens regardless of the system is based on how the lens performs without any correction.

If we are judging the lens quality attached to the format with software correction, we aren't judging the lens, but we are judging that particular setup.

 

[Conrad]

It's the final result that counts.

Software distortion corrections will impact sharpness negatively. But if the lens is sharp, the sensor is good, it may very well be the right choice. The designer always has to balance a large number of tech specs (distortion correction, all types of aberrations, performance wide open, weight, size, etc.) with the cost of manufacturing (or better: the value) of the lens. In the electronic age you would be stupid not to use the extra degree of freedom offered by software correction.

The same holds for a lot of other technical choices that manufacturers make for getting optimal image quality: auto white balance algorithms, AF technology, image stabilization, design of the analog part of the sensor electronics, etc.

So I agree: judge the system, and don't pick just one tech spec of one element in the whole chain.

 

[Brianetta]

Basically, it appears that the lens has quite a bit of distortion without software correction.

As a consumer, I understand that part of the Micro Four Thirds system is optical, and part of it is software. That the short register is going to cause distortion, and that the system is designed for this to be corrected by the image processor. Discovering that a particularly wide lens is basically acting the way that the system meant it to doesn't bother me a jot.

(-:

It's a bit like the Eurofighter Typhoon. Without its avionics computers, it can't be flown. Without its computer, the Micro Four Thirds system can't make photographs.

 

[Streetshooter]

I think it's great as a shooter, that the designers, engineers, software developers do such a great job. It makes it easy for me to get great results.
With film, these distortions meant a lot as there was not too much one could do to get past it. So, do I care about this stuff, sure but I care more about what I do.

 

[oldracer]

... the short register is going to cause distortion ...

I assume you mean the distance from the lens flange to the focal plane.

Just avoid confusion on anyone's part, I'll point out that the shorter distance really has nothing to do with causing distortion. A lens designer who wants to take advantage of the short distance to design a very compact lens may well count on software to correct distortion intrinsic in his design, but that is not a fault of the M43 flange distance. It's simply an option that the whole system approach makes available.

Re software correction, I agree it's the final product that is important. Does anyone judge a painting by the number of brushes used or by the number of paint colors used? I don't think so.

It's a bit like the Eurofighter Typhoon. Without its avionics computers, it can't be flown.

Yes. That's been true of fighters for some time. Dynamic instability means no human could possibly fly them but it increases their performance significantly. Not a bad metaphor for M43 distortion correction, actually.

 

[mary poppins]

for me, the final product is paramount. i'm not particularly concerned with what the corrective software is doing so long as i have a cursory understanding of that which it does, the corrective process doesn't 'get in my way', and truly does help me make technically better images. a closer analogy than a plane that you can't fly without computer assistance might possibly be traction control and steering/suspension/fueling adjustments on a sports car - you can still drive the thing to a good degree without these computer assists, but you can drive a more flawless lap with them turned on. i.e., you can still make images - even fine ones - without the distortion/CA corrections, but you can take technically better pictures with them.

cheers

 

[brnmatsumoto]

I am probably going be the odd man out on this, but in the case of the micro four thirds system, a software correction of the distortion is troublesome in that this is supposed to be an open architecture encouraging other manufacturer's to make bodies that fit the micro four thirds standard. My question is whether the open standard of the micro four thirds includes this type of software correction. If you mount this lens on an Olympus micro four thirds body, will its distortion be corrected in this camera body as well as the Panasonic?

Another aspect about this software correction--it is working on the JPEG file. The RAW file being minimally processed will display the geometric distortion. That is how the poster on DPreview spotted this problem. So if you are working in RAW does Panasonic provide the correction factors for the distortion so that you can "dial" out the distortion when working with these files? I have no problem with software correcting problems--but we also need to know to what extent it is doing this correction. It strikes me that if I am working with the RAW file, in say Aperture or Lightroom, I need to be able to correct for it in these third party programs. Maybe I am wrong, but it strikes me that porting out the JPEG from a third-party RAW converter program will display the distortion.

I suppose my post is asking more questions than it answers, but it strikes me that the lens has potentially less value if it forces an additional post processing step to be done in working with RAW files.

Brian

 

[grebeman]

In the electronic age you would be stupid not to use the extra degree of freedom offered by software correction.

Oh dear, call me stupid then

I agree with the points raised by brnmatsumoto, does that software correction translate from one manufacturer to another and across the various RAW post processing programs available or does it tie us to one manufacturer and the software that that manufacturer supplies and thus rules out exercising choice, that's not my definition of freedom.

Barrie

 

[hmpws]

To my knowledge, correction data are embedded in RAW outputs, and any compatible RAW converters will use them automatically.

Though, I think with Oly bodies, only distortions are corrected, whereas Panny bodies correct for both CA and distortions automatically.

 

[mary poppins]

...good questions.

 

[grebeman]

"I really love the 35mm effective focal length (efl) field of view (fov) of this lens... it really needs distortion/CA correction though, currently I'm using Olympus Studio, Capture One v5 PRO and Bibble 5 pro beta but none of them really nail the correction automatically."

The statement above is a quote from Brian Mosley in the Olympus 17mm f/2.8 Image Thread. This still leaves some doubts in my mind about leaving distortion correction to the software and how well each software is able to deal with the problems given it by any particular native lens.

Barrie

 

[Pelao]

If I understand correctly, this is all based on one persons sample images. Given this, and the rather subjective harsh views of other lenses that were expressed, I will await a full, professional review.

 

[Djarum]

Peleo,

Yes, you are correct, but thats not really what I am trying to discuss.

I think one of the arguments is that there are users who don't want the RAW conversion of the correction done automatically. There is some merit to this argument. I'm not much of a RAW user, so I really don't have an opinion on this.

As long as there is a standard method for correction that is open to the companies to keep the lenses interchangable between Oly and panny, I really have no problem with it.

Dj

 

[Brian Mosley]

Peleo,

Yes, you are correct, but thats not really what I am trying to discuss.

I think one of the arguments is that there are users who don't want the RAW conversion of the correction done automatically. There is some merit to this argument. I'm not much of a RAW user, so I really don't have an opinion on this.

As long as there is a standard method for correction that is open to the companies to keep the lenses interchangable between Oly and panny, I really have no problem with it.

Dj

Hi Dj,

the only problem I see, is that not all raw developers have the same access to 'adequate' distortion/CA correction algorithms... over time, hopefully raw converters will develop better approaches to getting the best quality image out of relatively poor lenses.

The purist in me still wishes for optically perfect performance... Surely the latest computing power and production techniques could deliver this?

Kind Regards

Brian

 

[Ray Sachs]

I thought Andy Westlake, from DP Review, did a very good job of explaining the tradeoffs of optical correction (ie, higher priced and larger lenses with, as I understand it, more corner softness) vs smaller and less complex lenses with sharper corners with the distortion corrected in the software. Its not that one is necessarily better than the other - they're philosophically different approaches to camera/lens design that are available in the digital age (and were NOT available in the film age). The final results, he claims, are pretty much the same - the path to get there is different. To keep size and weight lower, the m43 philosophy is to use software correction. There's nothing wrong with either approach but if size and weight matter to you, as is part of the driving philosophy behind m43, that priority results in a clear preference. If someone prefers more optically accurate lenses, there are plenty of directions to go for them, but m43 is probably not the ideal format for those people. My understanding of this stuff is cursory at best, but this makes sense to me.

I think it was a great thread over at DP Review. It clarified this issue for me quite a bit - I'd seen it mentioned but never really understood what people were talking about. Now I have at least a basic understanding of it. As an end user, if the end results work for me, I don't give much of a rat's ass how we got there. And the results from my ep2 and native lenses are more than good enough for me - its the best stuff I've ever shot with, including in the film days.

-Ray

 

[Djarum]

Ray,

I agree. Andy has done a good job of explaining the differences.

I think people were very spoiled by the 4/3 offer by olympus, and expected the same results. I also think that panny's 20mm also spoiled folks, and everyone expects similar performance.

Personally, for my 8x10's I print I think the quality is fine.

 

[BBW]

Ray do you have a link to that piece that Andy wrote?

 

[Djarum]

BBW,

Dpreview has a two page "blog" I guess on the issue here:

Distortion Article: 1. Page 1: Digital Photography Review

Now, Andy Westlake, in the thread I posted, makes several statements regarding distortion correction in Software.

 

[BBW]

Thanks!