[v5planet]

I never said it would lower the resolution of the lens. As MPs increase smaller apertures will become less useful and detail loss will be more significant. That has nothing to do with the resolution of the lens. It is a result of aperture size combined with pixel size and the size of the airy disk.

It is not brick wall. If the resolution gained by stopping down is greater than the detail loss from diffraction then you will still have a net gain.

Smaller apertures do not become less useful as pixel density increases, because they are still required for wider depths of field. Even when pixel size diminishes so completely that you have a net loss of detail on a PER PIXEL basis due to diffraction, you can just reduce the final image resolution to sizes we happily use now or have in the past. An image taken at f/11 on a 24MP sensor downsized to 14MP is not going to look worse than an f/11 image taken on a 14MP sensor.

---

Last edited by v5planet; 06-14-2011 at 03:36 PM.

[uccemebug]

What does everyone think?

That it's probably time to retire my film rangefinder. Hmmm, with a roll of Porta now $20 including processing, I think that lens and a compatible body might pay for themselves in about a year.

[v5planet]

The limitations on the finest details that can be captured imposed by diffraction are independent of the size of the pixels that the light hits. If you read that link you posted carefully, it discusses this in quite a fair amount of detail. The point is that as pixel density increases, you do not suddenly LOSE the ability to capture a detail that was seen by a sensor with a lower pixel density. In fact, because diffraction is NOT a brick wall, you will continue to see modest gains in the ability to resolve finer and finder detail, except that each iterative step of minutia will become more and more clouded by blur and loss of contrast due to the diffraction. That doesn't mean you are seeing less detail now - you are seeing MORE. It just means that pixel peeping becomes less satisfying because a 24MP image viewed at 100% will seem fuzzier than a 14MP image at 100% because the amount of detail PER PIXEL has reduced. At some point there is absolutely no more detail to be squeezed out of the equation due to diffraction, and if you continue to increase the pixel density on the sensor the only thing you will accomplish is spreading this same maximum amount of detail over more pixels. It is not destructive, just redundant.

You referenced an image on that site that showed the progressive loss of fine detail as you decrease the aperture of a lens, but decreasing the aperture of a lens is not the same thing as increasing pixel density for an image projected at a given aperture. So the argument that a narrow aperture lens becomes "less useful" with a higher MP sensor doesn't make sense.

When the K-5 first came out there was no end to the threads that complained about the new sensor "outresolving" all but the best lenses available, which seems like a related argument to me, and one I found irritating, because it caused people to think that their photography was now somehow worse than it was before or that their lenses were now deficient in some way they weren't with older cameras. That newer and fancier sensors reveal hitherto unseen deficiencies in a lens' optical characteristics, or that they are unable to overcome hard limits on the finest details that can be seen at narrow apertures due to diffraction does not somehow iteratively degrade the quality of photographs with each successive generation of camera. At WORST you reach a limit beyond which it doesn't make sense to focus on megapixels. This opens up an altogether different argument, one which is rational: i.e. should we bother with this or are our energies better spent elsewhere?

[RBellavance]

An image taken at f/11 on a 24MP sensor downsized to 14MP is not going to look worse than an f/11 image taken on a 14MP sensor.

Thank you for stating clearly what I was trying to get at in my earlier post on this subject !

[Digitalis]

When the K-5 first came out there was no end to the threads that complained about the new sensor "outresolving" all but the best lenses available, which seems like a related argument to me, and one I found irritating, because it caused people to think that their photography was now somehow worse than it was before or that their lenses were now deficient in some way they weren't with older cameras.

Yes V5, I was very entertained by these statements. However, there are some lenses that are frankly, dogs. And using them on a higher resolution sensor just makes this more obvious when you are viewing at 100% (though I have an eye for optical flaws and I can spot bad things even in thumbnails) the Pentax 645 45mm f/2.8 on the Pentax 645D being a case in point, the higher resolution of the 645D simply amplifies the fact that this lens isn't that great when compared to the Pentax 645 45-85mm f/4.5 zoom (and when a slow zoom lens can beat a prime you know it's bad).

Diffraction is also much more obvious on the 645D - but it is dependant on your output dimensions* - on most DSLR cameras AA filter already is softening things. On cameras built sans AA filter, e.g: the Leica M9 -stopping down to f/16 is as far as i'm willing to go with the M9 (thankfully most M lenses don't go any further than f/16 as a rule) Because at f/16 there is a distinct softening of finer details that would otherwise be clearly rendered at f/4 - the higher resolution just makes diffraction blur more obvious - even on a 3Mp D30 diffraction was there, the camera didn't have the resolution to make it as apparent as current higher resolution cameras do.


* for instance taking a Macro photograph at f/32 on a 1Ds MKIII and printing it at 13X19 will be less detailed than the same image printed at the same size but taken with the lens at f/8 with DOF stacking.

[v5planet]

Because at f/16 there is a distinct softening of finer details that would otherwise be clearly rendered at f/4 - the higher resolution just makes diffraction blur more obvious - even on a 3Mp D30 diffraction was there, the camera didn't have the resolution to make it as apparent as current higher resolution cameras do.


* for instance taking a Macro photograph at f/32 on a 1Ds MKIII and printing it at 13X19 will be less detailed than the same image printed at the same size but taken with the lens at f/8 with DOF stacking.

Right, diffraction is very real and I wouldn't argue that it doesn't start sapping away the sharpness of an image past a certain threshold. The argument I'm making though is that if an image looked fine at say, f/11 on a 14MP sensor, it will not look worse at the same aperture on a 24MP sensor at the same print/output size and same viewing distance. The image may look fuzzier at 100% on a screen, but that is only a reflection of reduced per-pixel sharpness when you start looking at previously unseeable details of such a small size that they are susceptible to diffraction blurring. The image and all its myriad details projected by the lens occupy the exact same physical dimensions on the sensor, except now each pixel is seeing less and less of the entire image, eventually down to a granularity too small for the lens (and therefore the sensor) to resolve additional detail.

[Digitalis]

The argument I'm making though is that if an image looked fine at say, f/11 on a 14MP sensor, it will not look worse at the same aperture on a 24MP sensor at the same print/output size and same viewing distance.

quite right, if the sensor size remains constant, there won't be any discernible difference. However, a larger sensor with the lens stopped down to a similar aperture will render the scene with greater clarity if all things remain equal, because there is less magnification for the original image - that is one of the reasons why I still use 4X5 and 8X10 format.

[Yamahaman]

How do we even know it will be called the K-3?
We don't.
Therefore I'm going to call it Rupert. Because I can.

Yeah, if it had gone K11, K9, K7, K5 then I reckon we'd have a decent shot at K3, but in this series it's really just been K-7, K5. Therefore I reckon it will be called K-rati.

[Winder]

Yeah, if it had gone K11, K9, K7, K5 then I reckon we'd have a decent shot at K3, but in this series it's really just been K-7, K5. Therefore I reckon it will be called K-rati.

If there is no body change I kind of like the idea of a K-5 Super.

[Winder]

$100.00 price drop in the K-5 and a $300.00 rebate from Pentax just in time for the holidays. Looks like they are getting ready to roll out a new model this spring.

$1,200.00 K-5 is a pretty good deal.

[sjwaldron]

$100.00 price drop in the K-5 and a $300.00 rebate from Pentax just in time for the holidays. Looks like they are getting ready to roll out a new model this spring.
$1,200.00 K-5 is a pretty good deal.

If there is no body change I kind of like the idea of a K-5 Super.

I think Pentax would be wise to keep the K-5 around as well. A K-5 Super version makes sense.

Someone posted a screenshot in another thread from the French Pentax site with a note about a show in early October, showing a blacked out body that looked like a K-5. I just hope a K-5 Super isn't the only SLR release they announce. I think it would be pretty lackluster assuming Canon and Nikon will be offering new SLR releases around that time as well.

Here is the PDF: http://www.pentax.fr/fr/news/1549/media/db8665807ae75d146df2db715e0e9631/E_I...011_PENTAX.pdf

At least based on past releases, Pentax should have a new body design lined up for the next release. The practical person in me would think it will be a basic step up from the K-5 with that new Sony APS-C sensor, advanced video controls, faster fps, faster AF, and other improvements. The idealist in me would hope for a larger sensor camera in addition to the K-5 Super.

[Winder]

I think Pentax would be wise to keep the K-5 around as well. A K-5 Super version makes sense.

Someone posted a screenshot in another thread from the French Pentax site with a note about a show in early October, showing a blacked out body that looked like a K-5. I just hope a K-5 Super isn't the only SLR release they announce. I think it would be pretty lackluster assuming Canon and Nikon will be offering new SLR releases around that time as well.

Here is the PDF: http://www.pentax.fr/fr/news/1549/media/db8665807ae75d146df2db715e0e9631/E_I...011_PENTAX.pdf

At least based on past releases, Pentax should have a new body design lined up for the next release. The practical person in me would think it will be a basic step up from the K-5 with that new Sony APS-C sensor, advanced video controls, faster fps, faster AF, and other improvements. The idealist in me would hope for a larger sensor camera in addition to the K-5 Super.

Considering Hoya was looking to sell Pentax I doubt they put a lot into R&D of product that would not be released while they owned it. It would make a lot of sense for them to design a K-5 Super using the same body and controls. Make improvements that don't cost any real R&D money like updating the Prime image processor, improve the AF & SR algorithms, & drop in a new sensor. Assuming the new sensor is an improvement over the current 16MP sensor.

Maybe we will finally see the AA filter removed..... I know wishful thinking.

[Zygonyx]

Can you imagine "dropping" a new sensor without R&D ?
Me not : the entire image processing chain is to be modified...

[sjwaldron]

Considering Hoya was looking to sell Pentax I doubt they put a lot into R&D of product that would not be released while they owned it. It would make a lot of sense for them to design a K-5 Super using the same body and controls. Make improvements that don't cost any real R&D money like updating the Prime image processor, improve the AF & SR algorithms, & drop in a new sensor. Assuming the new sensor is an improvement over the current 16MP sensor. Maybe we will finally see the AA filter removed..... I know wishful thinking.

Wasn't it said that Hoya was the one who pushed to get the 645D to market? It's true they trimmed a lot out of Pentax, but I think they would have kept R&D working full force to keep their unintentional investment worth something so they could sell it for a profit. Now that Ricoh is on the scene, we see that Hoya seems to have given them an amazing deal and kept their foot in the door to mutually make the Pentax brand stronger. Also, the Pentax Q survived (or started?) with Hoya as it must have been in the most important stages of development when Hoya was in charge.

I guess we will find out in October how well Hoya treated Pentax with SLR development.

[Ubuntu_user]

I guess we will find out in October how well Hoya treated Pentax with SLR development.

Mhm. And something inside of me says that Hoya was a whole lot better than we've thought...