when does lense resolution no longer matter?

[squeegee]

I'm reading all these lens reviews that go on about resolution and MTF charts and numbers and I'm wondering... at what point does it not matter at all. Specifically... will my picture (as a regular joe consumer, not a professional) look any different with say a $400 lens v.s. a $2000 lens.

Here's why it's a complicated question (instead of the obvious "yes").

1) lets be honest about our audience, my audience and the audience of the regular "consumer" is almost exclusively civilians looking at photos on computer screens. Most posted images will be the size of computer screens or less, so right now that would be less than 1920x1080 on average, i.e. about 2mp max.

2) lets assume we're using the correct focal length lens for the picture we want, we frame what we want properly while taking the picture, only very minor cropping or trimming is required. This is a relatively safe assumption given how easy it is to get say a 14-150mm zoom lens.

Lets assume...
- A $2000 lens may get every single pixel right on a 12mp sensor.
- A $400 lens gets blurry pixels smeared across 2 or even up to 4 pixels...

once I scale a 12mp image down to 2mp, won't they look awful close?

Does it mean that for digital-viewing people, it's worthless to buy professional grade lenses?

Does anyone have a $2000+ lens and a $200-$400 lens they can take 2 photos of the same thing with and post scaled down versions for us to pick at and wonder if it shows any real difference scaled down?

(This was inspired by reading reviews of the olympus 150mm F/2 super high grade lens... it has desire factor, but I doubt it's practicality value.)

 

[Djarum]

Its a good question. It depends on how large you print or look at pictures at 100 percent. Some people just want the sharpest lens possible just so they can say they have it, even though they are only displaying web sized images.

 

[Optical Thrombosis]

Partly correct, partly false. In many circumstances you are absolutely correct, and I frequently see pictures in magazines that are beautifully shot, and turn out to have been made on a kit lens of no great quality.

However I will suggest that with digital sensors, lens quality (in terms of resolution at least) is critical. It is a truth that the smaller the imaging area (sensor/film) the more resolving power you need in your lens. Its ironic, but when you view a 5x4 sheet film image, it appears fabulously detailed and sharp, but the truth is these large format lens are frequently of lower resolution than the best digicam lenses today.

Ultimately, resolution is only part of the story. To make an image "sharp" you also need good contrast and the subject matter has to be accurately contained within the zone of sharpness determined by the depth of field, in turn defined by aperture. Most unsharpness is frequently camera shake, despite advancements in "shake reduction" systems (tripod is best!)

Frequently people comment on how "sharp" macro lenses look - actually most of the apparent sharpness is due to your brain defining the differences between in focus and out of focus areas. This is why portraiture can get away with 'soft' focus on a lady with less then perfect skin, but to our mind she is still perfectly in focus as she contrasts so well with the blurred background.

In broad answer to your question, there are very few lenses that are available today that give poor results, even on the most demanding sensors. Modern lens design is so good that even modest zooms produce fine pictures rich in detail. However, each successive generation of multi-megapixel sensors ups the ante on the lens makers. A lens can only resolve so much, and that is why it is pointless having the 24MP full frame sensor instead of the 12MP, if your lens can only resolve to 12MP. You cant put any more light information onto the sensor than the lens can give.

Another area that is frequently overlooked (and one reason why I don't trust lens charts) is that the overall 'look' of a shot has as much to do with the contrast, depth of field, diffraction, reflection at lens surfaces and the sensor design and image processing in camera. Of course the biggest and most non technical reason is - you need a photographer with that create 'eye' for the shot, correct composition, correct exposure etc. All this has massive impact on the final image.

You could have a bag full of Canon L spec lenses and take lousy shots, unfit for publication. An amateur with a cheap body and kit lens can produce art. Ultimately, dont worry about what kit you have, just what you can do with it.

Sorry, thats a rether hastily compiled ramble - hope something is useful in it!

 

[Fiddler]

I take photos for myself, and work according to the 'good enough' standard. Which means that if it's it's good enough for me, it's good enough.

 

[squeegee]

yes I understand that composition is very important and yes I understand that you can make do with any lens.

I'm just wondering, if there will be any difference at all in the picture given the SHG zuiko 150mm and the m.zuiko 40-150mm zoom @ 150mm once both images are scaled down to electronic viewing size like 1600x1200.

I realise you may need slightly more detail / resolution / mp than the final output, just so you can do post processing and retail details etc.

I suspect the answer is that there's no real visible difference once it's scaled down - but it's hard to make such a statement never having tried it, that's why I was wondering if anyone with one of the expensive lenses has actually tried this experiment.

 

[BillN]

Leica and Zeiss lens will still easily "out resolve" the sensor - (even on the M9) - so the "experts" keep indicating

 

[feppe]

Back on-topic!

Lets assume...
- A $2000 lens may get every single pixel right on a 12mp sensor.
- A $400 lens gets blurry pixels smeared across 2 or even up to 4 pixels...

once I scale a 12mp image down to 2mp, won't they look awful close?

It is highly unlikely you or anyone would be able to tell the difference. I did a pretty thorough survey of medium format digital backs vs DSLR images on Luminous Landscape forums, and people were unable to reliably tell whether web-sized images were created with an MFDB or DSLR - many of them pros who use both cameras.

There was another test done very recently there on 100% crops of DSLR vs MFDB, and even then only a few people were able to tell the difference. Again many who use both camera types daily.

For yet another perspective with similar results in even more demanding case.

So if seasoned pros can't tell the difference most of time between a 40+ megapixel digital back vs. a DSLR, I can't imagine anyone could reliably tell the difference between resolution of a kit lens and a premium prime at web resolution.

Please note there are other factors on lens quality than mere MTF, though. Contrast, geometric distortion, and chromatic aberration are important ones - these can be fixed to a certain extent in post. Also, with more expensive glass you'll have wider apertures, which are necessary for blurred backgrounds and low-light photography, and some sports photography.