Distortion correction software vs. well corrected lens???

[rummenigge2]

Maybe this is more a philosophical question, but still I'm going to ask.

Searching for distortion free lenses for digital I was asking myself: Equals distortion correction via software a well corrected lens in quality terms?
What I mean is, with film correcting the lens was important, as later it couldn't be fixed easily. Today you can use software to ease the flaws of lens design (like Hasselblad is using it as a design approach).
So, is our search for the "best" lens (distortion free) only a reminiscence to the way we used to work? With digital pushing pixels is part of the proposition of the medium.

Thanks in advance for your thoughts

P.S.: I posted the same question on the Lula forum, but I'm sure there are different people visiting this forum, as I'm interested in a diversity of opinions.

 

[thomas]

Interpolation always reduces IQ a bit... this goes for software based lens correction as well.
Then again even the very best lenses regarding distortion (for instance Schneider Digitar symmetrical designs) require a little software correction if you aim at a perfectly corrected and straightened image.

Personally I prefer "distortion free" lenses for 2 reasons:
- the last bit of software correction is not always required (depending on the scene), so in this case you preserve best possible IQ
- less distortion of the lens requires less software interpolation. IMO this helps to get a more homogeneous distribution of "sharpness" all over the image plane (the relation of center and edges is more balanced - while the correction of heavy and complex moustache distortion often leads to softer edges). less software correction also preserves more pixels (less cropping required).

 

[JonPB]

As Thomas said, yes, you lose qualities when correcting for distortion. This can be from stretching the image to correct barrel distortion, narrowing the image to correct pincushion, or possibly a bit of both with a strong mustache. Generally, though, if you lose more than you gain by correcting the image, you have the choice of keeping the image as it was recorded.

The question is, do you get anything in return for a bit of distortion? I would happily trade some corner resolution for reduced astigmatism or focal plane variation, which cause issues that cannot be fixed in post. Likewise, though probably not optically related, I'll take vignetting over axial chromatic aberrations, and I'll take lateral chromatic aberrations over focus shift. Modern lenses are generally designed this way, so you may be seeing more distortion in current lenses but there are probably lower amounts of other aberrations as a result.

My end goal is an image for output, so I don't mind aberrations that are correctible in software nearly as much as aberrations that will make it into the print. Theoretically, the best lenses from the film era might not be the best lenses for use on digital, even among film-era lenses, due to the improvements in how we can manipulate our images today. I haven't seen a case where that's true, though.

I would strongly recommend that you pick a lens based on what the final images look like after your typical workflow, not based upon any one optical characteristic. You'll probably want to see if your photo editor has a profile for correcting the lens/camera combination that you're considering; I don't know how common these are for medium format rigs. We all have our own opinions about what defects can ruin an image, and also of what characteristics improve an image, so the best test to see if a lens that works for you is to process some raw files that were shot with it. I suspect you'd find some people are generous if you ask for sample files from a particular lens/camera combination. They may even tell you what to look for as you manipulate the file.

Cheers,
Jon

 

[Shashin]

I have yet to see any proof that distortion correction impacts the quality of an image. Personally, It is not something I worry about.

 

[ondebanks]

I'll copy what I said on LuLa as well:

Personally, I don't like the Hasselblad "let the optics distort and laterally fringe, just fix it in software" design philosophy. From what I've seen, it works marvellously - as long as you use the 1 or 2 software packages that can make sense of the metadata and perform the correction. But if you want or need to use something else (and I do regularly use non-photo-mainstream products), too bad.

+1 to thomas' points as well.

Ray

 

[MarkoRepse]

I personally would rather pay more for a better lens than one which requires corrections. To me it feels like cheating, though to be fair its not really not much of an issue, it works well for the most part and the difference isn't a big deal. On the other hand an exceptionally corrected lens, for example any true apochromat, has a presence that lesser lenses don't, software corrections or not.

 

[rummenigge2]

I would strongly recommend that you pick a lens based on what the final images look like after your typical workflow, not based upon any one optical characteristic.

Cheers,
Jon

Point taken and understood. I'm just not used so much to distortion, shooting large format and the Mamiya 7 system. On the Pentax 67 the distortion of the 45mm drove me nuts.

 

[rummenigge2]

thank you all for the interesting observations and opinions.

I didn't mean to point fingers at Hasselblad, actually I think their approach makes sense for the system they designed.

In general I'm yet not convinced that distortion correction as example for interpolation is so different from correcting CA. If you start to push the pixel around, it makes no difference – you mess with the original captured picture. What I'm thinking is more in the direction of the obvious in the world of digital capture. Does it make a difference if there is interpolation, because the capture is 1 and 0 anyway and no concrete and physical evidence of the capture does exist ( unlike with film).

I hope my thoughts make sense?!

 

[Shashin]

Does it make a difference if there is interpolation, because the capture is 1 and 0 anyway and no concrete and physical evidence of the capture does exist ( unlike with film).

I hope my thoughts make sense?!

The digital image is not ones and zeros--you would have a binary image, just two colors or tones--you are confusing computing with datasets. Actually, the digital image does exist physically--it has to be written to memory where it has a physical existence. There really is no real difference between a matrix of color density in a gelatin substrate and a matrix of color values in a file that is in anyway significant in making both photographs.

The question is whether spacial processing to an original image more destructive than capturing the image and applying no processing. As far as I can see, folks simply assume any spacial processing will degrade the image because a digital image has a finite unit, the pixel. However, when you actually do this kind of processing, there appears to be no effect--a perfectly rectilinear projection gets distorted anyway simply by viewing distance (the wide-angle effect is one example, but happens at all focal lengths). Seeing is believing in photography. To assume a perfect optical projection is some kind of ideal over any type of projection is actually hard to argue. The simplest is for architectural photographers, but then the arguments is from simply one aspect of photography, when there are other uses and criteria.

 

[chrismuc]

CA correction in PP (like directly in PS ACR) workes very well resulting in a very sharp image even if the lens is not free of CA.
Linear distortion can be corrected also very well if the lens behaviour is known precisely. Using either the correction in PS ACR (for lenses that are recognized) or the Alpa lens corrector plugin (for lenses that are not recognized or for shifted images) gives also very good results.
Generally one has to imagine that a correction of for example 2% geometrical distortion reduces the image resolution only by the square of 1.02 (= 2%) equals roughly 1.04 = 4% meaning f.e. a 60 MP image has about 57.5 MP after applying the correction. So that's a very slight resolution loss compared to the advantage of a distortion-free image for architectural photography.

 

[rummenigge2]

CA correction in PP (like directly in PS ACR) workes very well resulting in a very sharp image even if the lens is not free of CA.
Linear distortion can be corrected also very well if the lens behaviour is known precisely. Using either the correction in PS ACR (for lenses that are recognized) or the Alpa lens corrector plugin (for lenses that are not recognized or for shifted images) gives also very good results.
Generally one has to imagine that a correction of for example 2% geometrical distortion reduces the image resolution only by the square of 1.02 (= 2%) equals roughly 1.04 = 4% meaning f.e. a 60 MP image has about 57.5 MP after applying the correction. So that's a very slight resolution loss compared to the advantage of a distortion-free image for architectural photography.

super info, danke!!!

 

[Bob]

One person's "aberration" is another "character".
Mathematical aberration correction is well advanced and algorithms are well known.
Lens designers are always faced with a cost and complexity tradeoff in pursuit of the perfect lens.
At some point the price vs perfection relationship will shift in favor of mathematical correction as a particular market no longer can tolerate the cost.
Fringe area losses in linear resolution due to sample-stretching are usually less obnoxious then the aberration if one is so inclined.
OTOH
We do not have as many useful aberration or character enhancers until we evolve to better wave-front recording techniques.
-bob

 

[torger]

Usually "distortion" refers to the barrel or pincushion distortion of lenses (ie not chromatic aberrations or other). In that case distortion-free lenses require symmetric lens designs, something which was one of the typical properties of large format lenses.

The Schneider digitar series largely keeps this property also for digital. However as digital sensors have issues with color casts, symmetric designs seems to be abandoned on wide angles. Live view features also require very large aperture which also difficult to achieve with symmetric designs.

To get really good corner performance on wides you need to have a heavily corrected lens, and so far I've not seen a wide angle lens that combines this with a symmetric design (although I guess it's possible?). The retrofocus Rodenstock 32 has more correction and sharper corners than the symmetric Schneider 35 for example. While the Schneider is distortion free and thus do not require software correction I'd assume that you still get more corner resolution from a Rodenstock 32 after distortion correction in software (in the cases you care to correct for it, the 32 distortion is rarely noticable).

From a philosophical perspective I prefer the symmetric distortion free designs, I like when the camera itself projects a finished picture on the sensor, readily shifted, tilted and distortion free. I hope this genre of camera will continue to exist. It feels more "photographic" to me, and I hope the spirit of large format design continues to live for yet some time. I guess I'm a bit nostalgic.

In terms of raw performance we can probably get more resolving power with a combination of heavily corrected glass and software corrections though.