[Gooshin]

I, for one, will be sticking with APS for a long while I think. As a hobbyist I believe the smaller / lighter / cheaper lenses deliver much better value for money at the moment.

given the size of the Canon 5D i do not see it a problem for them to make FF cameras small in physical size

[heatherslightbox]

If I had $8000 to put towards a new camera, I think I'd save up a couple more grand and pick up a Mamiya digital MF--never mind 24x36. About the only way that will happen is if the husband produces a winning Lotto ticket. In that case, I'll fly to NYC myself and go on a shopping spree at Adorama and B&H and load up on both Mamiya and Pentax gear.

A girl can dream can't she?

Very interesting article--it's a good justification for picking up the 55/1.4 when it comes out.

Heather

[rparmar]

When entry-level products like the EOS 450D are out resolving the lens they're supplied with

Well, there's the logical error. The supplied kit lens may be crap but other lenses are not. The camera is nowhere near out-resolving good glass (as the article already mentioned concludes). So the rest of the argument is based on a false premise.

[ryan s]

Don't rule out the fact that when you shoot with a cropped sensor (rather than FF) you are shooting through the center (sweet spot) of older full frame lenses. This gives you better performance from a less expensive older lens (so some extent).

Indeed...corner performance plays less of a part, although there are 2 sides to that debate...

I don't even want to think about how much money I would have had to lay out to get the AF/modern/digital equivalents to what I own. Plus old lenses just feel better.

I, for one, will be sticking with APS for a long while I think. As a hobbyist I believe the smaller / lighter / cheaper lenses deliver much better value for money at the moment.

Looking at just the economic side of it...I got my K10 for $540. It was the top of the line Pentax at the beginning of the year. How much was, say, the 5D at that time?

I'm not even going to spend the money to go to the K20. I'll hold off until the K40D...or if they come out with the 645D it's mine. K10D + 645D = perfect P kit

[cpopham]

given the size of the Canon 5D i do not see it a problem for them to make FF cameras small in physical size

I'm not talking about the camera body, I'm talking about the lenses.

I was less than impressed about this canon's DR in tests. Nothing even remotely close to a MF it supposedly can replace in studio.

Radu

[Wheatfield]

Well, there's the logical error. The supplied kit lens may be crap but other lenses are not. The camera is nowhere near out-resolving good glass (as the article already mentioned concludes). So the rest of the argument is based on a false premise.

The article quoted talked more about edge sharpness and chromatic aberration than raw resolution.
If you check resolution numbers, you will find that very few lenses are hitting more than 80 lpmm at 1.6:1 TOC.

[creampuff]

Love him or loath him, here is what Ken Rockwell has to say about FF and the lenses you can use. Quite different from what dpreview asserts.

The Full-Frame Advantage

[Gooshin]

I'm not talking about the camera body, I'm talking about the lenses.

because we all know how big and heavy our supertaks are

[rparmar]

The article quoted talked more about edge sharpness and chromatic aberration than raw resolution.
If you check resolution numbers, you will find that very few lenses are hitting more than 80 lpmm at 1.6:1 TOC.

I have written in detail about this before, based on the Osuna and García article, specifically the results in Table 3. In short, while I cannot fault the authors' methodology, I disagree with the assumptions on which they base their conclusions.

They use an "average" lens optimised to f/11, which requires 7 MP to resolve fully on an APS-C sensor. But let's look at something better than average, the FA77. Using their terminology, this is a "highly corrected" lens, having its peak performance at f/4 to f/5.6. The chart shows 52MP at f/4. A further wrinkle: these values are for green-yellow light of wavelength .55 microns. If you want to capture blue light (sea, sky) double the MP values.

Thus, seascapes taken with an FA77 on the APS-C "crop" sensor require 101MP to resolve fully. If the work in this article is correct, I don't think we're going to "out-resolve" our best lenses any time soon.

[Gooshin]

caveat!

even if we were to create a 100 MP sensor and stuff it in our cameras to fully take advantage of our top glass

would we not be required to print billboard sized posters to justify this sort of thing?

how much megapixels is needed to produce a standard letter or A4 format print for a magazine cover before the difference in resolution will be beyond our printers or even eyes to differentiate?

Love him or loath him, here is what Ken Rockwell has to say about FF and the lenses you can use. Quite different from what dpreview asserts.

The Full-Frame Advantage

Creampuff,

With all due respect the whole discution gravitates around HIGH RES FF sensors, not the 12-16 Mp tops of D3's, 5D's or 1Ds MarkII's. The examples from the review show what happens with a 21 Mp sensor and the next gen seems to be in the mid 20's range so even more demanding.

Radu

I have written in detail about this before, based on the Osuna and García article, specifically the results in Table 3. In short, while I cannot fault the authors' methodology, I disagree with the assumptions on which they base their conclusions.

They use an "average" lens optimised to f/11, which requires 7 MP to resolve fully on an APS-C sensor. But let's look at something better than average, the FA77. Using their terminology, this is a "highly corrected" lens, having its peak performance at f/4 to f/5.6. The chart shows 52MP at f/4. A further wrinkle: these values are for green-yellow light of wavelength .55 microns. If you want to capture blue light (sea, sky) double the MP values.

Thus, seascapes taken with an FA77 on the APS-C "crop" sensor require 101MP to resolve fully. If the work in this article is correct, I don't think we're going to "out-resolve" our best lenses any time soon.

Rparmar,

I think those values are 100% theoretical it doesn't mean that any existant lens could resolve a 100 Mp APS-C sensor. Maybe a "million dollars lens for CIA's satellites" could be way better than anything we've seen before but I doubt that even such a lens could go that far.

Radu

[rparmar]

I think those values are 100% theoretical it doesn't mean that any existant lens could resolve a 100 Mp APS-C sensor. Maybe a "million dollars lens for CIA's satellites" could be way better than anything we've seen before but I doubt that even such a lens could go that far.

Well sure it's theoretical. So is the Theory of Flight but that keeps planes in the air.

Each single stop improvement in a lens doubles the MP it can handle yet again. Something like the FA77 is two stops more optimised than a generic lens. Going to full-frame doubles the requirements again.

If that seems counter-intuitive it's only because manufacturers have been falsely telling us that 8MP is "better" than 6MP and so on. In fact one needs to jump to 12MP to get significantly better than 6MP. From there it's only three incremental jumps to 100MP.

Our lenses jump faster than our sensor technology. Which is why you can slap a (carefully selected) old lens on a K20D and take fantastic shots.

And that's not theory... it's practice.

[Pål Jensen]

Creampuff,

With all due respect the whole discution gravitates around HIGH RES FF sensors, not the 12-16 Mp tops of D3's, 5D's or 1Ds MarkII's. The examples from the review show what happens with a 21 Mp sensor and the next gen seems to be in the mid 20's range so even more demanding.

Radu

1) The 21mp sensor is less of a high res sensor than the 14,6mp in the K20D; it is just larger. From a resolution point of view, the K20D puts more demands on the lenses.
2) APS sensor, like the K20D, sets even more demands on the quality of the lenses because the format is smaller. Compared to FF, all optical defects - including purple fringing - are 50% more for the same lens for a certain output size.
3) What is a professional opinion anyway? Someone having opinions as a profession? Thousands of pros use older lenses on FF Canon bodies with good results. Thousands of pros use even older, even worse lenses on medium format camera with digital backs. The results are spectacular.
4) The argument in the original article was that you need to use the best lenses to get the best out of the sensor. Wow!! What a surprise! This is true for any sensor film or whatever. It is even more important the smaller the format is.
5) This post was also made at the Dpreview list where it was debunked....