Where is Canon?

[Donald]

... as such, they give far too much weight to irrelivent factors (such as ridiculous pixel counts), and completely ignore other factors that have 100 times the impact on their photography (lighting, composition, technique, sharpening).

I'd just like to add that my dear old 10.5MP, 1600ISO (expandable to 3200) 40D is doing away just fine and seems to manage to produce half decent images when I push the button! No-one's complained so far!

[Manfred M]

I agree.

The absurd thing is that people go totally obsessive over diffraction and then completely ruin the image through resultant factors like insufficient DoF and totally inadequate (or non-existent) sharpening which contribute 20 times as much image degradation.

If folks want to shoot at a particular aperture to minimise diffraction - and that aperture still gives acceptable DoF - then go for it (no harm), but apart from that, JUST FORGET ABOUT DIFFRACTION. It's occurring at a level you'll only see at 100% magnification - and bears no relevance in the real world for all intents and purposes.

Hell, if I need to shoot at F32 then I shoot at F32 - and no client has EVER come back and said "Hmmm - nice photo, but there's too much diffraction at that aperture for my liking".

Totally agree Colin. I am still trying to figure out why people are obsessing over diffraction and all of the other minutia.

Lighting, composition, technique and post-processing (I put proper sharpening into that grouping) are what makes or breaks an image. You are quite right that no one is going to complain about diffraction, but if a piece of clothing looks be be the "wrong colour" all hell breaks loose.

I'm not sure where this obsession with these technical details comes from, but I think we can blame the camera review sites. I find that they report on things that are easy for them to measure (and repeat across a good many different makes and models), as opposed to things that are meaningful in real life photography. At that point the people with little or no understanding of the physics learn a new buzzword and we waste a lot of time arguing about it.

[Ady]

Just to be awkward I DO produce commissioned work and short run prints for sale at sizes of up to 50"x70" and a few extra pixels would be nice on occasions as a 22MP image with no cropping at all works out at about 80PPI at 50"x70".

I'm afraid I think the that the whole viewing distance thing is little more than a useful bit of misdirection, with the exception of instances where you can't/aren't interested in getting close, such as billboards. One may suggest that an appropriate viewing distance of a 50"x70" print is further than the distance available in the average living room but it seems that folks who commission a work of that size expect to be able to get up close and inspect every minute detail.

That is also one of the reasons I enjoy producing high resolution work at that size. To be able to look at a huge print of a storm wave from the other side of the room and take in the power and majesty of the whole, but also to be able to get up close and see every eddy and ripple and each drop of spray is something I love being able to produce, and it seems to be something that others are willing to pay for.

So a 30 or 40+MP FF sensor would certainly be welcome if not essential. I haven't dumped my Canon gear for anything else so it's obviously something I can live without, but when it's here and if the IQ is good enough I'll be putting my order in.

Cheers,
Ady

[ajohnw]

I agree.

The absurd thing is that people go totally obsessive over diffraction and then completely ruin the image through resultant factors like insufficient DoF and totally inadequate (or non-existent) sharpening which contribute 20 times as much image degradation.

If folks want to shoot at a particular aperture to minimise diffraction - and that aperture still gives acceptable DoF - then go for it (no harm), but apart from that, JUST FORGET ABOUT DIFFRACTION. It's occurring at a level you'll only see at 100% magnification - and bears no relevance in the real world for all intents and purposes.

Hell, if I need to shoot at F32 then I shoot at F32 - and no client has EVER come back and said "Hmmm - nice photo, but there's too much diffraction at that aperture for my liking".

I thought Manfred's post might come up if I used reply with quotes but taking both.

I don't get wrapped up in diffraction effects. I'm probably more aware of them due to interest in both microscopes and telescopes. The main reason I don't get wrapped up in them is that black and white contrast falls to about 7% at what is usually used at the diffraction limit. In other words about 7% of what ever contrast goes in actually comes out of the optics. As I understand it the generally accepted maximum contrast reduction for good photography is 80%. Often referred to as MTF 80. So basically in this respect it doesn't make much sense to consider diffraction limits at all other than the fact that they are a real physical limit full stop so there is zero advantage in having pixels that will record it. What is far more important is or was when quoted is the resolution in line pair per mm over a wide range of stops. These days we get line pairs per picture height hopefully at 80% contrast. That could be used to determine max print size but given software it's a very woolly area.

So in a nut shall when considering the purchase of a new camera purely for more pixels I consider diffraction limits and if anywhere near it loose interest. I've seen comments that Pen's have diffraction effects at F8, 48mp at full frame - personally having use much slower stops on all sorts of things I feel that the comments are a load of rubbish. Pens - no choice in pixel counts nor with the EM-5. In practice camera lenses produce a certain sized circle of confusion that usually crisps up when it's closed down some what rather than a diffraction spot and in practice that aspect is what matters.

My main interest is shots on a PC screen etc or even an iPad anyway.

John
-

[Colin Southern]

Just to be awkward I DO produce commissioned work and short run prints for sale at sizes of up to 50"x70" and a few extra pixels would be nice on occasions as a 22MP image with no cropping at all works out at about 80PPI at 50"x70".

I do them too

The problem you have there though is that it isn't just "a few extra pixels". To be able to do what you're wanting, you're going to need around 180 PPI (some say much higher) - and if a 22MP image (say, Canon 1Ds3) is giving you 80 PPI then you're going to need 111MP to give you 180 PPI - and I doubt you'd even find a MF camera that'll give you that. Going from 22MP to - say - 32MP doesn't make a lot of difference.

I'm afraid I think the that the whole viewing distance thing is little more than a useful bit of misdirection, with the exception of instances where you can't/aren't interested in getting close, such as billboards. One may suggest that an appropriate viewing distance of a 50"x70" print is further than the distance available in the average living room but it seems that folks who commission a work of that size expect to be able to get up close and inspect every minute detail.

In my experience, people can generally be divided into 2 groups. The first are photographers, where the minimum viewing distance is limited only by the length of their noses (although some try to simulate getting even closer by using magnifying glasses), and the 2nd group comprising of "all the rest". And the photographers are a pretty small group.

In all my years of printing, one of only two things happen when I produce a large print for a customer: either they look at the whole print and go "wow" (the majority), or some do come a little closer to look at some detail - but - (and this is the important part) - they're always quite accepting of whatever detail they can or can't see. There's always going to be some limitation on detail; even with a gigapixel photo of a city you're never going to be able to zoom in to see how clean the drinking glasses are on their kitchen tables, so in that regard "it is what it is".

At the end of the day, the 3 most important things to remember are "It is what it is", "it is what it is", and "it is what it is". If someone wants to invest in a 120MP sensor to produce images with twice the print resolution over competitors with a 32MP camera then he's of course welcome to "go for it", but in the same breath, he needs to gain a measurable commercial edge to get a return on his considerable investment and - frankly - I think he'd struggle to do that. At best I think if would be a very high end / niche market.

[Colin Southern]

That could be used to determine max print size but given software it's a very woolly area.

Hi John,

It never determines the maximum print size -- it only ever determines the viewing distance for a given size, to an individual's personal standard.

This is where people keep putting the cart before the horse; some would say that the "maximum print size" from a 18MP camera is "18 x 12 inches"; it's BS - you can print as large as you like. It simply comes down to how close do you want to be able to get before the degradation and lack of detail becomes an issue for you. If someone is one of the (photographic) minority who feel that a 8 foot wide print is ruined because they can't see all the detail when pressing their noses up against it then they're very much SOOL ("Seriously" Out Of Luck). Thankfully though, the vast majority just stand back and look at the whole thing. NEVER ONCE has somebody not bought one of my large prints because they couldn't see enough detail; EVERYTIME they stand back at a distance and admire the colours, overall composition, and "majesty" of it.

[jad]

I can see there are many good and informative opinions about the direction FF sensors and pixel count. I have many years of working with large format cameras and making large display prints for exhibitions. I started the switch to digital in 2008 with the purchase of the 5D11. It is a very impressive camera but in my opinion it is just mid way on the technology curve of what future digital cameras might be. I didn't want my post to focus solely on pixel count and should have emphasized the need of sensor technology a little more. I could be happy with a 28mp camera with an improved sensor to eliminate digital noise. There is one big difference with the old film technology and the newer digital. With film I didn't need to upgrade my equipment to improve my print quality. I simple relied on the release of new films, papers, and chemicals to get better prints. Today it is solely with the equipment and that usually requires a costly upgrade of camera,lenses, printers, software,and computer etc. to improve picture quality. It was often said that the camera is just a tool. I don't think that applies much anymore. It's just an upgrade like a computer. I worked with a used Linhof for nearly forty years that never needed upgrading. That was a tool.
http://johndoddato.blogspot.com/

[ajohnw]

Hi John,

It never determines the maximum print size -- it only ever determines the viewing distance for a given size, to an individual's personal standard.

This is where people keep putting the cart before the horse; some would say that the "maximum print size" from a 18MP camera is "18 x 12 inches"; it's BS - you can print as large as you like. It simply comes down to how close do you want to be able to get before the degradation and lack of detail becomes an issue for you. If someone is one of the (photographic) minority who feel that a 8 foot wide print is ruined because they can't see all the detail when pressing their noses up against it then they're very much SOOL ("Seriously" Out Of Luck). Thankfully though, the vast majority just stand back and look at the whole thing. NEVER ONCE has somebody not bought one of my large prints because they couldn't see enough detail; EVERYTIME they stand back at a distance and admire the colours, overall composition, and "majesty" of it.

It an interesting area Colin. There is a web calculator about that uses the old 10in / 250mm viewing distance metric assuming good eyesight. The outcome is often disappointing so it also offers a size for poorer eyesight ! It's a fact that viewing from a more realistic distance has a dramatic effect on the ppi that are needed to match 20/20 vision. In real terms given up sampling / conversion to printer dpi etc and all the the other things that go on the only important thing after a fashion really is the level of detail in the photo even to the extent that a certain minimum size may even be needed to even show it. Where detail is low cheap upscaling compacts show that much bigger than expected prints could be made that way - even viewed from the end of the nose. It all depends on the actual detail in the shot.

Perhaps the best way to look at max print sizes is to remember 20x16 exhibition prints from film days in both colour and black and white. They didn't take an awful lot of effort really. Viewing distances for something that size is probably around 3ft, or even more really. The actual detail level needed for that is probably about 80dpi - what ever dpi it's printed in. if 80 dpi actually captures the detail and it was printed at 600 dpi in principle it would look fine from 5in not that many can focus that closely.

I've probably taken some liberties with distance versus eye resolution but hopefully some will see the point. In real terms as Colin points out it all depends on viewing distance. That will also limit detail that can actually be seen as well.

John
-

[Colin Southern]

It an interesting area Colin. There is a web calculator about that uses the old 10in / 250mm viewing distance metric assuming good eyesight. The outcome is often disappointing so it also offers a size for poorer eyesight ! It's a fact that viewing from a more realistic distance has a dramatic effect on the ppi that are needed to match 20/20 vision. In real terms given up sampling / conversion to printer dpi etc and all the the other things that go on the only important thing after a fashion really is the level of detail in the photo even to the extent that a certain minimum size may even be needed to even show it. Where detail is low cheap upscaling compacts show that much bigger than expected prints could be made that way - even viewed from the end of the nose. It all depends on the actual detail in the shot.

Perhaps the best way to look at max print sizes is to remember 20x16 exhibition prints from film days in both colour and black and white. They didn't take an awful lot of effort really. Viewing distances for something that size is probably around 3ft, or even more really. The actual detail level needed for that is probably about 80dpi - what ever dpi it's printed in. if 80 dpi actually captures the detail and it was printed at 600 dpi in principle it would look fine from 5in not that many can focus that closely.

I've probably taken some liberties with distance versus eye resolution but hopefully some will see the point. In real terms as Colin points out it all depends on viewing distance. That will also limit detail that can actually be seen as well.

John
-

To be honest John, I often feel that folks - literally - "lose sight of the big picture". I like to use the example of carpets; if I were to pay you a visit - and your lovely wife, being the house-proud woman that she is, vacuums the carpets so that they look nice. Do you think she'd scream "foul" when I arrived if I were to get down on all fours and inspect the carpet with a magnifying glass? She would of course have been expecting to have the entire carpet observed from a distance - not inspected up-close with a magnifying glass. Same with images - in my opinion anyway.

I see far too many people who ruin the image trying to save the pixels; they get motion blur because they won't use higher ISO modes because of the "noise" - they get insufficient DoF because they won't shoot past F8 because of the "diffraction" - and they won't print an image shot with a 36MP camera bigger than 12 x 8 because it starts to lose detail when viewed under a microscope. My personal feeling is that on a scale of "OCD" to "Real World" it scores pretty close to the "OCD" end of the scale. Striving for perfection is admirable, but folks also have to be accepting of the fact that that's just a photographic utopian dream; past a certain point we just have to let it go or it's just too darn restricting. Something is usually better than nothing.

[nimitzbenedicto]

To be honest John, I often feel that folks - literally - "lose sight of the big picture". I like to use the example of carpets; if I were to pay you a visit - and your lovely wife, being the house-proud woman that she is, vacuums the carpets so that they look nice. Do you think she'd scream "foul" when I arrived if I were to get down on all fours and inspect the carpet with a magnifying glass? She would of course have been expecting to have the entire carpet observed from a distance - not inspected up-close with a magnifying glass. Same with images - in my opinion anyway.

I see far too many people who ruin the image trying to save the pixels; they get motion blur because they won't use higher ISO modes because of the "noise" - they get insufficient DoF because they won't shoot past F8 because of the "diffraction" - and they won't print an image shot with a 36MP camera bigger than 12 x 8 because it starts to lose detail when viewed under a microscope. My personal feeling is that on a scale of "OCD" to "Real World" it scores pretty close to the "OCD" end of the scale. Striving for perfection is admirable, but folks also have to be accepting of the fact that that's just a photographic utopian dream; past a certain point we just have to let it go or it's just too darn restricting. Something is usually better than nothing.

Exactly !!!!

( OTOH, some strange native-English speaking people who minutely pore over every single pixel, who spend hours and hours and hours and hours and countless hours and more hours critiquing/ editing images, spewing out advice/tips/secret formulas/etc., spend money on expensive softwares, buy tutorial books/DVDs, expensive lens, cameras, etc.
but can't even take a sec , perhaps too lazy ( or don't know how to spell) to correct their typos?!

but let one pixel out of line or color or WB or exposure -
Boom! they are quick to point out to the photographer their mistake!!!

Jz seying...

)

[Glenn NK]

I think there's an even BIGGER gap between what people need and what they THINK they need.

I think this is absolutely correct - and is likely the crux of the problem in 99.5 percent of cases.

Glenn

[Colin Southern]

I think this is absolutely correct - and is likely the crux of the problem in 99.5 percent of cases.

Glenn

But I wonder if the sales and marketing folks are aware of this?

Old adage I guess - selling people what they lead them to believe they want and need - not what they actually want & need.

[Glenn NK]

Colin:

Isn't that what salesmanship is all about?

Glenn

[Colin Southern]

Colin:

Isn't that what salesmanship is all about?

Glenn

Sales and marketing - yep - absolutely.

It's the ethical & moral side I wrestle more with ...

[jcuknz]

I am probably wrong in my estimation but I blame the current emphasis of technical as a result of glossy paper being the usual surface for paper in the early digital days ...I also know from my TV days where I was an operator instead of a technician that they simply want technical advances for itself so it looks good on the oscyliscope or CCU and the rest as Manfred listed #23 is far down the list.

The head of technical services at WNTV-2 apparently often said ' I can send a tech out onto Lambton Quay and he gets perfect exposures ... why cannot you news togs? ' A troll if ever there was one Then a jounralist accused me of being artie when I organised a tree over a building I was shooting ... I knew the white sky would cause telecine to bounce so I reduced it as much as possible.

The only time I had a technician get enthusiastic about what I had done was when on reading a Kodak Newsletter I played with controlled fogging of 16mm film to reduce the contrast range to what telecine likes. Latter the whizz ex-maintenace guy and colour processing supervisor Ray Englefield adapted the DNTV-2 machine to do this and we had film sent from other TV stations producing programmes for this processing treatment as producers heard about it and saw the results on air.

It is good they have this attitude because it leads to all the improvements we are given to use so I am not knocking them just understanding their position until it goes to far like another situation where we had a projector capable of recording magnetic sound which the technicians told us we couldn't use, "not up to standard". But when we wanted this service we sent the film and audio to an outside firm using the same machine LOL Crazy! Insisting that a Film Editor should not do it but a Sound Technican I could understand but the total ban .... silly.
I have another choice story for another time

[jcuknz]

I am sure diffraction was covered when we had our weekly two hour theory session back in 1952 but it wasn't until I joined CiC last year that it became a 'serious topic'. True I was not doing photography all that time but other related and unrelated employment ... even now I am not sure I'd recognise it if I saw it ... something is sharp or it isn't ... too simplistic for some I guess.

[ajohnw]

I probably sprout on about diffraction more than some due to interest in photography through a microscope where is does really matter. Not as many suppose though. Here it's a case that the contrast available at the diffraction limit is so low that it's of no photographic use at all on many subjects. In cases where the subject needs that level of resolution it's actually likely to spoil the shot so subject choice and level of detail has to be related to the optics used. Hence diffraction limits are irrelevant really here as well other than optics must be capable of resolving a lot more detail than the subject needs. Processing then adds it's usual woolliness to hard optical fact. Actually I find the thought that camera lenses are diffraction limited at fast apertures a little amusing

Personally I pixel peep for 2 reasons. Curiosity and a more practical one - to see how far I can push the effective focal length by cropping / how much I can crop out if I can't decide what to frame eg a flower bed maybe.

I do shoot at F20 at times by the way. In some cases that may have an effect. The infamous spot is then 27um dia then and contrast is useless. Tried to find a decent web page that might explain what this means but there is an awful lot of myth kicking about.

ISO I try to keep low but if needed will go as high as the camera can realistically go.

By the way if the typo's are mine I do check and make some really odd ones that might me missed. Prose - some don't like it. I know I'm right but can change my mind in a flash. Sorry and all that but that is me. I'm also not a dedicated follower of fashion. I firmly believe in trying things and making my own mind up. I feel the web makes that approach essential these days - and many books. Some one will probably decide shooting to the left is best one day and have photo's to prove it. Shooting to the right, good grief, shoot to suit. Some deep field focusing techniques really need a laser range finder and an iPad application to do the sums, why not look through the camera even though focusing screens aren't that good for this sort of thing :now eek: easy on mirror less but have to use the magnified view.


John
-

[xpatUSA]

. . . I'm not sure where this obsession with these technical details comes from, but I think we can blame the camera review sites . . .

It's just human nature, Manfred. Think "Hi Fi" back in the 60's, there were those who obsessed over Bode plots, rise times, "ringing", speaker polar radiation patterns, phase accuracy, room size, shape, acoustic properties and who religiously subscribed to Wireless World or some such mag. Then, there were those who just bought the best money could buy and listened to scratched-up 45s on it.

There were those who were deeply interested in Aviation and learned aerodynamics, knew of Reynolds number (instead of Rayleighs criterion) knew structural engineering, da di da. Then there were those just flew 'em.

Cars . . Boats . . Watches . . etc . . . etc . . .

Such folk are always in the minority and as such are always derided. Sour grapes, of course

[Colin Southern]

It's just human nature, Manfred. Think "Hi Fi" back in the 60's, there were those who obsessed over Bode plots, rise times, "ringing", speaker polar radiation patterns, phase accuracy, room size, shape, acoustic properties and who religiously subscribed to Wireless World or some such mag. Then, there were those who just bought the best money could buy and listened to scratched-up 45s on it.

The funniest part was though that in blind tests, they couldn't tell units apart.

[ajohnw]

LOL Some years ago I found some decent sounding PC speakers after a few tries. When these went west I tried to buy some more but Logitek had bought them up. Bought others. Crap so went for a pair of near field monitors with 5in speakers in them much better but still lacking in realistic bass. Tried a sub woofer but the bass came out from under the table the PC is on. Now have some studio monitors that have 8 3/4 in speakers in them. Much much better but another sign of wont change conservatism. It;s perfectly possible to get a better response out of 6in speakers but prejudice says they just wouldn't be any good. And of course the bigger they are the more they should cost as they must be better. Lots say so - must be true. It's even on the web etc.

The speakers sit on sand filled 2ft stands behind the monitor - Need that height to clear the monitor and leaves more space for clutter on the table, Sand - less chance of them falling over.

I used to buy Wireless World. It relates to my work and interests. When a US company bought it up and ended it immediately and replaced it with their own pretty useless rag assistants in Smiths, one of our mags and books chains, took several steps backwards when customers asked where it was. Many looked like they were going to get violent when they were shown the replacement.

John
-.