Problem: lack of definition, how to sharpen?

[Mark Baigent]

Hi Philippe

 

I am sorry if you took offenec, that was not my intention

 

> B.t.w. Mark, why don't you start a petition to drop the Let's talk about pics forum (Purely to talk about

> and share images, portfolio critique, inspire and general positivity). My God, you must hate it!

I like talking images, but this drifted into complaining about QC standards and lack of help from.

Then we had suggestion of letting "8s" through.

I am being positive by promoting professionalism.

 

> How many underexposed, tilted horizons, ridiculously overpolarized skies and other dross do you like me to show you,

> here at Alamy? 1000 pictures? 100,000? A million amateurish snapshots? "professional standard", are you kidding??????

I completely agree, the faults you mention should have been part of QCs standards set. No reason for others to suggest lowing standards further though.

 

You said

> In such a way it's unmarketable? Alamy doesn't sell blown up tiny portions of pictures, it sells whole pictures.

This could be solved by letting the buyers see a 100% portion so they can decide. For now I assume that buyers expect images to be sharp.

 

 

> If Alamy would take the OP's picture as a standard of passing/not passing and looks at ALL its images at pixel level....... I truly wonder how much

> of the 44 million pictures would remain.

I agree with you again, all images should be checked for technical quality and also exposure, tilted horizons, etc.

 

 

Philippe

If you feel that I overstepped the line I am happy to quit the thread, I have no interest in further upsetting anyone.

 

Mark

 

 

 

[Mark Baigent]

I'm fine with QC benchmarks, but let's not pretend that there's any kind of 'zero tolerance' approach. There's only one way to ensure that every pic in the collection meets the technical criteria... and that's to examine every one. Alamy's collection is unedited for subject matter, and part-edited for technical competence...

 Sadly true, I would be very happy to see them check all images for technical competance.

[Mark Baigent]

Hiya

 

 

Sorry for my harsh reply. Take it with a grain of salt.

 

No worries here, I did rant and that can attract attention.

 

All the best
Mark

[Craig Yates]

I just love the friendly banter on here   it's good healthy debate.

 

Back to the image, personally I would have hit the delete button on the camera. With modern camera technology we have the option to zoom the image and check

the sharpness straight away and retake if necessary. Not wait until it,s on the monitor screen to identify lack of sharpness or definition.

 

 

Regards

Craig

[John Mitchell]

I just love the friendly banter on here   it's good healthy debate.

 

Back to the image, personally I would have hit the delete button on the camera. With modern camera technology we have the option to zoom the image and check

the sharpness straight away and retake if necessary. Not wait until it,s on the monitor screen to identify lack of sharpness or definition.

 

 

Regards

Craig

Have to say that you've hit on a brilliant idea. I always zoom in to check sharpness after snapping. Have to say, though, that I've been surprised on occasion when looking at the image later on my computer. Still, today's high-res LCD camera screens do give a pretty good idea of sharpness and DOF.

[John Crellin]

I think it's an unusual case. At 100% the clip does show even very small twigs, but I find it impossible to look at the image - my eyes hurt! There's some shake or something present which is affecting the way it looks so you try to concentrate on it to see the detail but can't focus on it. Sometimes you have to step back from trying to study whether objects 2 oe 3 pixels in size are visible, and just look at the way the 100% image hits your eyes. Because that's how it will hit QC.

 

Hits the nail on the head - brilliant.

 

I don't see it as a 100% certain fail either but it is recognisable to me as the sort of image that bothers me greatly at my final checking stage before upload. Can't quite say what the fault is but... It's far easier when you blow what you thought was a brilliant pic up to 100% and immediately see obvious camera shake - a great disappointment but only one course of action...

 

John Crellin

[John Morrison]

If I'm looking at a pic at 100%, and I'm wondering "is it sharp?"... then the answer's always "no".

 

After a few years of QC I think I can recognise the 'snap' of a sharp image across a range of subjects. However, there do seem to be some minor anomalies with certain subjects. Some foliage in a sharp image can look slightly soft, as can people...

[John Crellin]

 

 

I think it's an unusual case. At 100% the clip does show even very small twigs, but I find it impossible to look at the image - my eyes hurt! There's some shake or something present which is affecting the way it looks so you try to concentrate on it to see the detail but can't focus on it. Sometimes you have to step back from trying to study whether objects 2 oe 3 pixels in size are visible, and just look at the way the 100% image hits your eyes. Because that's how it will hit QC.

 

Hits the nail on the head - brilliant.

 

I don't see it as a 100% certain fail either but it is recognisable to me as the sort of image that bothers me greatly at my final checking stage before upload. Can't quite say what the fault is but... It's far easier when you blow what you thought was a brilliant pic up to 100% and immediately see obvious camera shake - a great disappointment but only one course of action...

 

John Crellin

 

 

Camera shake wouldn't show the small branches and certainly not the little twigs as sharp. Everything would be soft and that's not the case. I believe there was a small breeze, slightly moving the branches laden with snow (bare branches don't move in a slight breeze).

 

Cheers,

Philippe

 

I wasn't saying the picture in question was camera shake just for the record. I think you may well be right and I have encountered this.

[John Crellin]

If I'm looking at a pic at 100%, and I'm wondering "is it sharp?"... then the answer's always "no".

 

After a few years of QC I think I can recognise the 'snap' of a sharp image across a range of subjects. However, there do seem to be some minor anomalies with certain subjects. Some foliage in a sharp image can look slightly soft, as can people...

 

Agreed - and one of the "certain subjects" is the case of flowers etc where the detail is in shades of yellow for instance. This is not surprising when you consider how the cameras actually work and the inevitable smearing from the algorithms that puts pixels together. Plus the human eye's acuity varies with colour...

 

But yes, sadly as you say if you are wondering...

 

John

[John Crellin]

Forgot to mention, camera shake can be elusive and twigs etc apparently sharp if they happen to be aligned with the direction of shake. That is often the explanation in my experience for apparently OK pictures where "something is wrong".

 

John

[John Mitchell]

Personally, I think that this image would have passed QC a few years ago. The times they have a changed, which isn't surprising, I guess, given the size of Alamy's collection now.

[Bill Brooks]

There are some subjects that have a naturally unsharp look, and I think the example is one of them. This does not mean that Alamy should accept the image. Does the snow on the trees have a surface coating of ice as well?

 

Anything with a coating of ice, such as tree branches, will look less sharp, because you are looking through the translucent ice covering, which is diffusing the edge of the branch underneath.

 

Same thing with marine mammals, or fish, or people when they are dripping wet and leaping out of the water.

 

Slight movement of branches due to rising hot air currents in a valley will cause some branches to be blurred at 100% and some not, unless your shutter speed is over 1/640.

 

Hot air rising over water or land will also cause unsharp distortions to distant objects.

 

A very cold -40C day may change the characteristics of your lens.

 

These distortions usually do not show up at 100% from a 12 megapixel camera but do on a 21 megapixel camera. I am wondering how Alamy sets the viewing for sharpness standards? Do they view all cameras at 100% regardless of file size?

 

A better comparison would be to view the 6 megapixel minimum allowed file size at 100% and view larger files at a lesser magnification determined by the file size. Give reviewers a 30 inch monitor, and allow the software to automatically choose the percentage that will display the entire image at once on the large monitor. About 33% magnification on my 30 inch for a 21 megapixel file.

 

Photographers should still track quality at 100% however.

[John Mitchell]

 

There are some subjects that have a naturally unsharp look, and I think the example is one of them. This does not mean that Alamy should accept the image. Does the snow on the trees have a surface coating of ice as well?

 

Anything with a coating of ice, such as tree branches, will look less sharp, because you are looking through the translucent ice covering, which is diffusing the edge of the branch underneath.

 

Same thing with marine mammals, or fish, or people when they are dripping wet and leaping out of the water.

 

Slight movement of branches due to rising hot air currents in a valley will cause some branches to be blurred at 100% and some not, unless your shutter speed is over 1/640.

 

Hot air rising over water or land will also cause unsharp distortions to distant objects.

 

A very cold -40C day may change the characteristics of your lens.

 

These distortions usually do not show up at 100% from a 12 megapixel camera but do on a 21 megapixel camera. I am wondering how Alamy sets the viewing for sharpness standards? Do they view all cameras at 100% regardless of file size?

 

A better comparison would be to view the 6 megapixel minimum allowed file size at 100% and view larger files at a lesser magnification determined by the file size. Give reviewers a 30 inch monitor, and allow the software to automatically choose the percentage that will display the entire image at once on the large monitor. About 33% magnification on my 30 inch for a 21 megapixel file.

 

Photographers should still track quality at 100% however.

 

Yes, the world is sometimes fuzzier-looking than we realize. You can't always blame it on the photographer. As mentioned earlier, downsizing to 24 MB might make a big difference with this shot in my humble estimation (but then I'm a somewhat fuzzy person by nature). My guess is that Alamy views everything at 100% regardless of file size. Perhaps with camera manufacturers cramming more and more pixels onto sensors in order to keep up with the Joneses, this policy is something that should be reconsidered. I don't know how many other stock agencies judge IQ at 100% (?).

 

In a similar vein, back in 2007-2009, I uploaded at least 1500 scans (35mm slides) to Alamy without a single failure. None of them were downsized (54 MB files). When I look at some of those scans now, I realize that I probably wouldn't dare submit them today for fear of failing QC. Fortunately, though, they keep on selling on Alamy with no complaints (touch wood) from buyers.

[Ed Rooney]

Talking about music, Duke Ellington once said, "If it sounds good, it is good." I wonder what he would have said about sharpness? 

[John Morrison]

I wonder what he would have said about sharpness? 

 

"See sharp"...

[David Kilpatrick]

I have seen Alamy QC working a few years ago. They had 24" LG Screens, probably have something bigger or better now. First, the screen fills up with large thumbnails. Reviewer has to see many pages of these - maybe thousands a day. Even at this size they can spot doubtful situations/image look. They can then punch any one thumb up to full screen fit. Some will get rejected at that stage. If they want to do they can do 100% view at any point on the screen. Reject as needed. Then - back to the next full screen of big thumbs.

 

All done in very egonomic conditions (position and distance of monitor) in neutral grey environment and controlled light in shifts or rota which prevent fatigue. QC is well planned.

[John Mitchell]

I have seen Alamy QC working a few years ago. They had 24" LG Screens, probably have something bigger or better now. First, the screen fills up with large thumbnails. Reviewer has to see many pages of these - maybe thousands a day. Even at this size they can spot doubtful situations/image look. They can then punch any one thumb up to full screen fit. Some will get rejected at that stage. If they want to do they can do 100% view at any point on the screen. Reject as needed. Then - back to the next full screen of big thumbs.

 

All done in very egonomic conditions (position and distance of monitor) in neutral grey environment and controlled light in shifts or rota which prevent fatigue. QC is well planned.

Always interesting to hear about what goes on behind the scenes. It certainly does sound like a well-planned setup.

[David Kilpatrick]

Philippe - I am working with 36 megapixels at the moment (having done so briefly with the Nikon D800/E in the past and also with MF). I have invested in tilt adaptors and will need to use them - this is like working with 5 x 4 film. Even f/22 is not enough. Landscape yesterday, final image size was 125MB but decided to knock it down to just under 100MB, and was composed from two shots - a foreground focus shot, blended with a distant focus shot, joined in the dark area below the horizon:

 

 

My assumption is that the line of trees needs to be sharp, as well as the central foreground (not worrried about the very bottom of the shot). Don't have a tilt adaptor for this lens yet, crude focus stacking only works for some shots. Some of my panoramas or joiners on this camera are well over Alamy's maximum file or JPEG size limit so they have to be reduced down.

 

You suggest that it may be unfair on 24 megapixel+ users that their images are scoped at 100% (if necessary) and the same goes for 9 or 12 megapixel users. I'd say the unfairness was the other way round. If a 2400 x 3600 image is unsharp I'll bet it actually looks unsharp at full screen QC view, but a 5000 x 7500 pixel image could be unsharp and look OK, so escape getting 100% view. My approach is to look at the best sharpness I get routinely (normally prime lens in the studio with flash) and if any large pixel count file falls short of this, reduce its size. Objective, for all images regardless of size to be sharp at 100% view. That includes 24 and 36 megapixel images, which plenty of photographers are now using. If these large images are not just as sharp as smaller sizes, they don't have much point, you might as well use a smaller file size camera and reduce your processing overhead.

 

It's worth noting that with the right lens and shooting conditions, the image on a Sony A7R (no AA filter) is so detailed that it can be exported from raw to 60-70 megapixels. When Alamy first set a minimum file size, cameras were 6 megapixels and the required file size was 17 megapixels, involving an enlargement from approx 3000 pixels long dimension to approx 5100, or 1.7X linear. Despite the heavy AA filters and often quite inaccurate focus of these early DSLRs, these scale-up image were accepted - subject to QC - and many remain in Alamy's files. I have a few thousand myself. When I return to look at these, they are softer to a degree I can hardly believe, compared to my submissions today. With the camera I'm using now, I can reduce files by 0.5X linear, one quarter of the data size, for Alamy's current file size lower limit of 24MB. With the choice of cameras we now have available, every image you submit to Alamy should be as sharp as they require it to be, at 100% view.

[David Kilpatrick]

Can't agree Philippe. You have to view high pixel count cameras as a different format. Heather Angel managed to keep butterfly wings perfectly sharp on 6 x 6cm, I might well fail to do so APS-C but that's my fault and my images won't sell for A3 calendar pages as a result. The whole point about Alamy is that one standard applies, independent of file size, and therefore you should be able to licence and download any of the listed file sizes for a given image, and know it will reproduce well when used at a typical print setting (300dpi for print, 240dpi for inkjet, 150dpi for murals/banners). How it looks at 72dpi is irrelevant unless you want to use a tiny clip on screen, as the file will always be purchased/downloaded at a web size, and we should all hope that every single image on Alamy will be more than OK for that.

 

If Alamy accepted soft images or inadequate depth of field just because the image was 16, 18, 20, 24, 36 or whatever megapixel size and not 9, 10, 12 or 14 (for example) the client would be misled. If I file a 36 megapixel image it is my intention that the client should able to print this to A2 size and still have it look like an exhibition print - just the same way that when I used to shoot 6 x 9cm slide film for libraries, I hoped to find buyers who needed that size (I never shot 5 x 4" except for commissions). I know that my libraries made 6 x 9cm enlarged dupes from 35mm originals in order to sell my work better - I've got dozens of them back from when the agencies scrapped real film. I've looked at them. Of course they were never any better than the 35mm original. Clients were slightly deceived. Alamy avoids that (except with the early uprezzed material).

 

I will go even further than this. I'll say that I would accept it if Alamy automatically limited the maximum offered size of all images originally filed before 2008 to 9 megapixels. I would even accept it if Alamy decided to purge all unsold images pre-2006 (Canon introduced a rather poor 8 megapixel quality in 2005, other makes went straight from 6 to 10 megapixels in late 2006). 

[Bill Brooks]

Good discussion, and David has changed my workflow.

 

In the days of film, when 35mm scenic shooters tried their hand at 4X5 their initial efforts would often be rejected by the library (not Alamy) for being slightly soft. The library's position would be that, yes the slightly soft 4X5 had more detail than a sharp 35mm, but when clients leased a 4X5 they expected much higher standards than 35MM. They expected mega detail and a slightly soft 4X5 containing more detail than a sharp 35MM would still disappoint the client.

 

Click on an Alamy thumbnail and you get a statement about maximum image size at 300 DPI based on the file size. The buyer is probably using that maximum image size as a guarantee of high quality up to that size. Therefore sharpness should be judged in the Alamy manner at 100%, regardless of the camera's native file size. I take back all my blather about judging images at other than 100%.

 

I can't believe I said this. Cold day yesterday. Brain frozen.

 

 

 

A better comparison would be to view the 6 megapixel minimum allowed file size at 100% and view larger files at a lesser magnification determined by the file size. Give reviewers a 30 inch monitor, and allow the software to automatically choose the percentage that will display the entire image at once on the large monitor. About 33% magnification on my 30 inch for a 21 megapixel file.

 

High resolution shooters beware. More is expected of you.

 

I use 3 cameras, a 5D11 - Alamy rating 47.5 CM long side at 300 DPI, a SonyRX100 rating 46.3 CM, and a Nikon1 AW1 rating 39 CM. This guarantee means that I am rating the Sony close to the 5D11 in sharpness, when in reality it is closer to the Nikon. From now on, I will be outputting Sony JPGs for Alamy at Nikon size.

 

For further clarity this Sony downsizing is not to sneak Sony images past an Alamy quality inspector. It is rather to meet the quality expectations of myself, Alamy, and the client.

[John Mitchell]

The fact that my fuzzy scans and upsized (to meet the old 48 MB requirement) images continue to sell on Alamy tells me that buyers are still more concerned about content than how sharp an image looks at 100%. Also, Alamy has older, lower megapixel DSLRs on its recommended cameras list, and I'm sure plenty of contributors are still using them. This emphasis on super-sharpness, almost above all else, is getting a bit out-of-hand IMO. Just my two pxels' worth...

[MDM]

Philippe - I am working with 36 megapixels at the moment (having done so briefly with the Nikon D800/E in the past and also with MF). I have invested in tilt adaptors and will need to use them - this is like working with 5 x 4 film. Even f/22 is not enough. Landscape yesterday, final image size was 125MB but decided to knock it down to just under 100MB, and was composed from two shots - a foreground focus shot, blended with a distant focus shot, joined in the dark area below the horizon:

 

 

My assumption is that the line of trees needs to be sharp, as well as the central foreground (not worrried about the very bottom of the shot). Don't have a tilt adaptor for this lens yet, crude focus stacking only works for some shots. Some of my panoramas or joiners on this camera are well over Alamy's maximum file or JPEG size limit so they have to be reduced down.

 

You suggest that it may be unfair on 24 megapixel+ users that their images are scoped at 100% (if necessary) and the same goes for 9 or 12 megapixel users. I'd say the unfairness was the other way round. If a 2400 x 3600 image is unsharp I'll bet it actually looks unsharp at full screen QC view, but a 5000 x 7500 pixel image could be unsharp and look OK, so escape getting 100% view. My approach is to look at the best sharpness I get routinely (normally prime lens in the studio with flash) and if any large pixel count file falls short of this, reduce its size. Objective, for all images regardless of size to be sharp at 100% view. That includes 24 and 36 megapixel images, which plenty of photographers are now using. If these large images are not just as sharp as smaller sizes, they don't have much point, you might as well use a smaller file size camera and reduce your processing overhead.

 

It's worth noting that with the right lens and shooting conditions, the image on a Sony A7R (no AA filter) is so detailed that it can be exported from raw to 60-70 megapixels. When Alamy first set a minimum file size, cameras were 6 megapixels and the required file size was 17 megapixels, involving an enlargement from approx 3000 pixels long dimension to approx 5100, or 1.7X linear. Despite the heavy AA filters and often quite inaccurate focus of these early DSLRs, these scale-up image were accepted - subject to QC - and many remain in Alamy's files. I have a few thousand myself. When I return to look at these, they are softer to a degree I can hardly believe, compared to my submissions today. With the camera I'm using now, I can reduce files by 0.5X linear, one quarter of the data size, for Alamy's current file size lower limit of 24MB. With the choice of cameras we now have available, every image you submit to Alamy should be as sharp as they require it to be, at 100% view.

I understand from my own experience with D800/E and agree with almost all of this and do pretty much the same. Leaving DoF differences between 36MP and smaller sizes aside and considering a shot of a flat plane only, the large pixel size image will be just as sharp as a smaller pixel size image, assuming a decent lens etc etc. Experimentation is required when considering depth of field but it is definitely possible to produce an image from a D800 that is just as sharp as that from a D700 when viewed at full size if the differences in DoF are taken into account.  

 

What I find confusing about David's post is the stated need for tilt adaptors and focus stacking. I know David is talking about a different camera but I am confused about this. Is this specific to the Sony camera or a general principle?

 

Maybe I'm misunderstanding David's post, but I have found that, with a little experimentation on the D800/E, I can use hyperfocal focusing with wide angle and standard lenses at f11 (for optimum DoF, best edge-edge sharpness and to avoid diffraction effects) and get everything pinsharp front to back, viewed at 36MP and 100%.  I have needed to find new hyperfocal distances as the old tables and barrel markings don't work at 36MP viewed at 100% on screen. With a 24mm at f11 and a D800, it's possible to get everything from just over 3 metres to infinity more than acceptably sharp without any capture sharpening or any other form of sharpening or downsizing. With a 50mm at f11, everything from 9 or 10 metres to infinity is pinsharp if optimally focused. These near-distances are about twice the near distance for hyperfocal focusing using the traditional tables. If one wishes to get a closer distance to infinity in focus, then downsizing becomes necessary. 

[David Kilpatrick]

There's the difference - I am looking for much greater depth. The two-exposure focus composite I posted is sharp from 2ft to infinity and was shot at f8. A tripod was not an option as the camera was two inches about the water surface with the rear screen extended to waist-level position, and for this I wanted to keep down to ISO 200. Three metres is distant for me. I use wide-angles in a very specific way, often with small foregrounds.

 

 

This is on the Fuji XE-1 I think, using the 14mm (21mm equivalent). The RX10 can achieve this result easily. The A7R obviously at full frame is not ideally suited to extremes like this, but I'm looking for d-o-f more like 1m to horizon with a 17mm or 20mm. To achieve this, I would use the 17mm tilt-shift on Canon (Nikon's shortest option is 24mm) but on the A7R I have a wide choice of wide angles and tilt adaptors. I would not use the technique all that often and it can produce some ugly imaging quality.

 

Many photographers would not need to consider this as they don't use such close foreground detail.

[Bill Brooks]

Why get everything in focus from close foreground to infinity?

 

The viewers attention is drawn to the sharpest thing in the picture, so the subject of the picture should be rendered the sharpest. Everything else should be a gradation to slightly softer through to out of focus completely, depending on their importance to the main subject.

 

David DFYCBX is a successful image with a great treatment of foreground, but I think it could be stronger if focus did not extend to infinity.

 

Forgive me as I am not familiar with UK sports, but to me that spikey thing in the foreground is the subject and probably denotes the sport. In order of importance. The spikey thing, the grass denoting a playing field, the player on the right, the players on the left. The trees, houses and clouds in the sky to me are distracting and should be softened until they only suggest the environment.

 

Therefore I think hard depth of field should extend only to the spikey thing, the grass, and through to the player on the right.

 

If an illustrator were drawing this image the distracting elements would only be suggested with a few brush strokes, and the image would be made stronger.

 

The classic photograph of a close flower in front of a distant mountain is a good example of extended depth of field that will work against you. Most photographers will move heaven and earth to get everything in focus from inches to infinity. If the subject is the flower in the mountains, wouldn't it be better to get the foreground flower critically sharp and the distant mountains softer so that they still read as mountains, but soft enough that they do not distract from the flower? It is distracting to have trees on the mountain just as sharp as hairs on the flower. With everything sharp the viewer has too many subjects and you loose a sense of scale.

 

The cover of the December issue of Outdoor Photographer is a case in point. A technical tour de force, but I think the tack sharp frost in the foreground and tack sharp mountains in the background, fight for the viewers attention. The top mountain subject is a great picture on it's own. The bottom frost picture is also great on it's own. But bring them together in one image?

 

http://store.wernerpublishing.com/Outdoor-Photographer--December-2013_p_222.html

 

Photographers can learn a lot from illustrators

 

http://www.amazon.ca/Color-Light-Guide-Realist-Painter/dp/0740797719/ref=sr_1_2/182-2313763-9735619?ie=UTF8&qid=1390252633&sr=8-2&keywords=Gurney

[MDM]

There's the difference - I am looking for much greater depth. The two-exposure focus composite I posted is sharp from 2ft to infinity and was shot at f8. A tripod was not an option as the camera was two inches about the water surface with the rear screen extended to waist-level position, and for this I wanted to keep down to ISO 200. Three metres is distant for me. I use wide-angles in a very specific way, often with small foregrounds.

 

 

This is on the Fuji XE-1 I think, using the 14mm (21mm equivalent). The RX10 can achieve this result easily. The A7R obviously at full frame is not ideally suited to extremes like this, but I'm looking for d-o-f more like 1m to horizon with a 17mm or 20mm. To achieve this, I would use the 17mm tilt-shift on Canon (Nikon's shortest option is 24mm) but on the A7R I have a wide choice of wide angles and tilt adaptors. I would not use the technique all that often and it can produce some ugly imaging quality.

 

Many photographers would not need to consider this as they don't use such close foreground detail.

OK David - that makes sense now. I'm often looking for a realistic representation of a scene and use a 50 mm more than anything else. I've adapted my style to the D800 so I've sacrificed close foreground at the expense of sharp background. In contrast to what Bill says in his last post, I want maximum sharpness to infinity - I may want to use images for scientific illustration,  I also want to be sure that they pass Alamy QC and I simply like maximum sharpness as a matter of taste.

 

In cases where sharpening is permissible (not for Alamy of course) as in making display prints, for example, then it would be possible to increase the apparent depth of field even at full size. And where sharpening is not permitted, the option of downsizing is always available. I can get from about 0.7m to infinity sharp with the 24 if I downsize to 12MP size without any sharpening. Using a camera with a  36 MP sensor (and good lenses) allows these sorts of choices to be made. The pros greatly outweigh the cons as far as I am concerned.