• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Putting MP issue into perspective

Asher Kelman

OPF Owner/Editor-in-Chief
Tom,

That's good to know!

Why can't you buy the magazine and send it to me!

Anyway, it's good to see comparisons in real life situations. 4MP pics have been used for billboards, but wont necessarily cut it for Lancome!

The public is not a measure for judge the quality of antibiotic choices, or high off the ground SUV's for your kid's they all seem reasonable.

With pictures, what counts, to me at least, is how it enables or enhances the ability to deliver an image or print for the purpose in hand.

If one has to make a living, quality might matter in getting a sale where subjects or detail rich, such as landscape, or the file faces demanding art editors at Vogue or the like or whether you can focus fast enough and track a player in a game.

If one photographs the shot that records the shooting of a movie star, heaven forbid, then, it will sell, no matter how many pixels.

One has to look as to who is the customer that pays the photographer. It doesn't matter what the public wants as far as sales are concerned if they don't buy your product!

For detail rich landscapes and architecture, the 1DsII is for many people just marginal.

So it all depends what the Practical Photography article used as example files for printing and how they were processed.

Maybe we should have some tests with real life jobs?

asher
 

Asher Kelman

OPF Owner/Editor-in-Chief
Tom Yi said:
...........Remember, using a bigger sensor with higher resolution will show the softer edge of the image circle that a cropped sensor will not have, so there is some trade off as well.
I would guess that softness of Canon lenses at the edges is not an issue for most work except architecture.

In fact, the edges may need to be softened anyway.

We'll see what happens when and if a 1DsIII comes out this Photokina and photographers vote with their wallets.

Asher
 

Alain Briot

pro member
Tom Yi said:
Another way to look at it is that at 12x18 prints, about 44% of viewers could tell the resolution diff between 16.7 vs 8.2 MP sensors.
If you routinely print at this size or bigger, I'd say that it's a significant thing. If you routinely print at 8x12 or smaller, the sensor issue maynot be a readily visible thing.

In this regard the question is: what maximum print size do you intend to make? For 8x10's a 1DsMk2 is a luxury. For 40x50, it's insufficient.
 
Tom Yi said:
Another way to look at it is that at 12x18 prints, about 44% of viewers could tell the resolution diff between 16.7 vs 8.2 MP sensors.

Kind of funny consider that doubling the number of pixels only yields 1.44 times as many pixels per inch/cm in linear measure. 44% of viewers could spot a 44% increase in linear pixel density (which is the only measure that matters for edges in terms of pixels [this assumes the same or a very similar lens is used]). :)

Makes me wonder if 50% of the viewing public could spot a quadrupling of pixels (32 MP, 50% increase in linear pixel density) in a 12x18 print? This is probably a likely number consider that 19.44 MP is all it takes to get a 1:1 300 dpi 12x18 print.


a silly dash of numerology,

Sean
 
Tom Yi said:
Ah, the getting what you need vs. getting what you want issue.
True, I think for 8x10 prints 6MP is probably enough.

I have been happy with 8x10s from a 5 MP camera (with good technique and appropriate subjects for a digicam). Albeit, a 4x5 crop from a 5MP 3x4 image is likely similar in resolution to a 4x5 crop from a 6 MP 2x3 image. The 5 MP has roughly 97% of the linear pixel density in an 4x5/8x10 crop/print. Whereas in an 8x12 crop one only gets roughly 86% of the linear pixel density. Yielding the fact that consumer digicams have all the resolution they need for 8x10 prints (but the lenses may be awful and user skill is often lacking).

enjoy,

Sean

-------

Calculation notes:

5000000/4/3 = 416666.666 pixels per unit area

6000000/4.5/3 = 444444.444 pixels per unit area

ratio = .9375 (square root transforms area ration to linear ratio)



-----------------------------------



5000000/3/2.25 pixels per unit area
6000000/3/2 pixels per unit area

ratio = .741
 
I should also note that while linear pixel density correlates well with edge sharpness, 2-D textures correlate with the areal pixel density (94% and 74% for 8x10 and 8x12 prints assuming minimal cropping).
 

Andreas Kanon

New member
MP count is overrated in my opinion.
I have enlarged slightly cropped prints from my D70 (6.1 MP) to 20x30" and no problems whatsoever.
What people often seem to forget is at what distance the viewing takes place.
 

Gary Ayala

New member
I think that the "crop" factor is an important play here. There was a time when I used to shoot full frame (even filed out my neg carrier for those full frame egotistical borders). But now that I'm an ol' fart, I seem to be cropping everything after the fact. So for moi ... the more MPs = the more I can crop = the more I can be lazy. But I am digressing ... I think for us that do not shoot set-up shots (studio or landscapes), due to the crop factor, the more MPs (all else being equal) the better.
 

Dierk Haasis

pro member
Andreas Kanon said:
MP count is overrated in my opinion.
I have enlarged slightly cropped prints from my D70 (6.1 MP) to 20x30" and no problems whatsoever.

I wholeheartedly agree, since the days of 4 MPx cameras we are in an area of possible quality where more pixels will yield larger files but not necessarily bigger pictures. To get a considerably bigger 'negative' we need four times the number of pixels as has already been written - that puts a Canon 1DsMk2 just at "twice as good as" a Canon G2. Actually it is even worse, at least for this specific combo, since the sensor sites in the dSLR have a wider pitch resulting in less detail per area unit!

This example shows that even the crop headroom is a non-argument. In the good old times of film, the recording area between a 35 mm negative and the smalles middle format negative (4.5x6) was huge from the start, and the actual light recording entities (the film grain) was not spread wider with larger film formats but remained the same. thus it made sense to crop after the fact instead of getting it right from the start. Even then only one format was regularly cropped, the square 6x6, which neede to if one wanted to create a dramatic format.

Since digital images are much smoother than film (made up of randomly scattered grains of varying size) they can usually be blown up on step more than film (as an estimate) without loss. With a tool like QImage you can go much higher. Setting the relative resolution correctly - most tend to put it too high - even ,ore can be done without interpolation.

At 300 dpi a G2 photo blows up to ca. 8" x 6", roughly A5; at 240 we get 10" x 7" , slightly smaller than A4. Even 200 dpi would give a very good print. On a practical note: I've printed G2 photos up to A3+ ( ca. 13" x 19") without problems on an Epson 2100.

Not to forget, as Andreas pointed out: The larger the print the further away you stand to look at it. Unless you are a PixelCounter.
 

Alain Briot

pro member
Harvey Moore said:
One factor seldom seen in this type of discussion is image cropping headroom

It can be an expensive proposition ;-) With the 1DsMk2, tilting the horizon slightly, say 10 degrees or so, will cost you 2 to 3k in deleted pixels by the time you level the image and crop out the corners...

There's a lot to be said for composing the image as precisely as you can in the field when using a digital camera. Best to use all the pixels you paid for!
 
Dierk Haasis said:
Not to forget, as Andreas pointed out: The larger the print the further away you stand to look at it. Unless you are a PixelCounter.

Not quite true. Throwing out pixel peeping, I view prints at noseprint distance because I enjoy having fine detail tickle my eyes. On a wall, 3.3 MP will yield a nice looking 8x10, but no fine detail will be present in a print. This "appropriate" viewing distance reminds me of going to galleries and seeing boring (i.e, dull subject and 3rd rate composition) B&W photos blown up to huge sizes with platinum prints and called art (for the process, not the actual lack of artistic skill that is shown in the composition).

The correct viewing distance is the distance from which the observer chooses to observe.

No complexity, no sophistry, no manual to read, just a picture to look at. Never forget that in art there are no rules, just results. Either it works for the viewer or it does not.

In short, your ideal viewing distance may be not be mine. Who cares. If the goal is to view the composition as a whole, then a reasonable range of viewing distances could be prescribed. But I am more detail oriented than that. The composition as a whole is judged before my mind even takes in the colors fully (visceral reaction). But the details that tickle the eyes can hold my attention much longer. But this is also what grabs my eyes in the 3-D universe too.

That said, a blurry 16x24 is okay viewed from 3'/1m away. But it will not be a spectacular noseprint viewing shot. But it can still be a great print.

one opinion,

Sean
 

Andreas Kanon

New member
6 MP still holds way enough detail for a 20x30 enlargement.
Even if I stand close and view the details.
So saying that you need 10+ MP to make enlargements interesting is incorrect in my mind.
If the original is well exposed and good technique is used 4 or 6 MP can get you a long way.
More MP will give you more room to work with but doesn't by definition hold more details just because it has the capacity to do so.
 
Andreas Kanon said:
6 MP still holds way enough detail for a 20x30 enlargement.
Even if I stand close and view the details.
So saying that you need 10+ MP to make enlargements interesting is incorrect in my mind.
If the original is well exposed and good technique is used 4 or 6 MP can get you a long way.
More MP will give you more room to work with but doesn't by definition hold more details just because it has the capacity to do so.

With a sharp lens, MLU, and remote shutter release one can get a dead sharp image. Printing that 8 MP image at 12x18 shows softness, but an acceptable amount. At 20x30 same image would be soft (like an 8x10 from a 3.3 MP camera. But then I am nearsighted. Also, personal perceptions of what is sharp and what is oversharp hold sway here. I am not saying you need massive MP to make a large print, but I am saying a 20x30 print will show softness unless you use an ancient 50 DPI plotter to print it. I look at posters off of offset presses (same size range) and they have never been soft as they simply lack the resolution to be soft. Instead the fine details are ink dots rather than image detail. It is more a matter of what you want from a print than what is right or wrong.

In the end, it is always up to the viewer to judge what is right or wrong for them.
 

Dierk Haasis

pro member
Sean DeMerchant said:
The correct viewing distance is the distance from which the observer chooses to observe.

No problem with that. The 'sophistry' you invoke is an after-the-fact explanation - a rule not a ruling. Usually people instinctively use an appropriate distance all by themselves, without prompting, without knowing anything about rules. Only those interested in certain phenomena, like technique or technology, will consciously[/i] change their viewing distance to a (ridiculous) distance.

Reducing ones viewing distance to 10 cm or less, perhaps with an 8x loupe, for a 3 x 3 m Pointilist picture is anything but appropriate. Standing 10 m from a postcard-sized portrait of a lady is equally ... un-appropriate. Just watch how non-geeks adjust their viewing distances in art galleries, when looking at differently sized prints, or in the cinema [there's a good reason most of us view people who want to sit in the front row as "mentally challenged"*].

Sean, like you I inspect prints for minute flaws, especially when thinking about a new printer. No printing process can reproduce details I need to close in for - and if they could it wouldn't be worth it. Just print a closer crop larger to see the detail. I remember a story by a German photo journalist decades ago about one of his teachers shooting in MF or LF, making a very large print of part of the original, and then cutting out the relevant portion to the size he wanted it.

When the detail has informational content and you need to go close up, there's something wrong with the print in the first place. One rule of thumb to test this: Do you have to ask others to close in? That is, does a viewer has to change his instinctively sought position? If so, crop the picture to the real image within it and print that.




*Before someone takes exception, I try to use double-quotes consistently for irony. I am not making fun of actual ill people here but of political correctness and what it has done to language. What we all think in reality when hearing someone to want a front row seat in the cinema for a Cinemascope movie is: 'idiot' or 'don't you have glasses' or 'his physiotherapist must be rich'.
 

Dierk Haasis

pro member
Sean DeMerchant said:
[...] get a dead sharp image

Which achieves what?

I'd rather view images with content than a dead sharp one with an evenly spread histogram and other technical problems solved. May be the reason I don't like advertising photography, I don't like (German) contest photography, I don't like stock photography. Not all but most of them are hollow shells, posing as interesting because technically and formally they are "correct" [= adhering to rulings*].

A few months back I went it studio portrait, not a matter I'd see myself as proficient in, but a freind asked me for a new picture of her for job applications. The result was quite nice, capturing her love of life and her natural friendliness. Unfortunately a lot is wrong with the photo, it's slightly blurred due to camera shake, the dark background does not set her (she's a blonde) free as I thought it should, some specular highlights are a bit distracting (particularly in the teeth). The last one is not really a technical mistake, I like these highlights as they bring life into the portrait.

When printed the blur isn't visible, at worst it looks a bit like I used a very light soft-focus filter 70's style. I was very surprised when she came back congratulating me on the photo, telling me that all potential bosses she had sent her application to took note of it. Very unusual that somebody tells an applicant how good their photo is [at least in Germany, where professional studio portraits are a must for job applications].




*There's a difference between rules as observed explanations of the world, and rulings as prescriptions on how the world should be. The former are, to spell it out, descriptive, the latter prescriptive. I am all for the first, against the latter.
 
Dierk Haasis said:
No problem with that. The 'sophistry' you invoke is an after-the-fact explanation - a rule not a ruling. Usually people instinctively use an appropriate distance all by themselves, without prompting, without knowing anything about rules. Only those interested in certain phenomena, like technique or technology, will consciously change their viewing distance to a (ridiculous) distance.

Reducing ones viewing distance to 10 cm or less, perhaps with an 8x loupe, for a 3 x 3 m Pointilist picture is anything but appropriate. Standing 10 m from a postcard-sized portrait of a lady is equally ... un-appropriate. Just watch how non-geeks adjust their viewing distances in art galleries, when looking at differently sized prints, or in the cinema [there's a good reason most of us view people who want to sit in the front row as "mentally challenged"*].
I disagree with some but not all of this. Let me introduce one more variable here:
I am nearsighted.
And this significantly impacts my visual behaviors. Others are nearsighted too. Others are farsighted. Some are both (with my sympathy). Others have severe astigmatism. ....

Having always been nearsighted, getting in close is how I see anything.

While I do wear glasses, the early childhood development of the brain likely hardwired getting in close into my mind as I have always done that. The devil is in the details, but so is Nirvana. After years of photographic study and learning to see I can no longer trust my instinctive reactions to be representative anymore. Heck, I now need to watch an artisitic movie twice to see the plot as I get distracted watching the photography the first time through and often miss part of the plot. But once upon a time in the past I would watch movies and following critical focus through a film without ever once seeing the areas of defocus sharply (now I watch their choice of backgrounds and subtly appreciate the effort of having actors making their marks to create the photographers vision).

In short, getting close is how I see. Without getting in close, I am often left with visual confusion and I have always been this way. Without glasses, I cannot see the whole of a thing as it is blurry. Without glasses with my arm stretched out in front of me I can see the junction of my hand and wrist dead sharp, but my finger tips remain in defocus no matter how hard I try. But unlike many, I can focus easily 10 cm in front of my eyes with absolute clarity of detail.

In other terms, seeing up close is what seeing is for me. At a distance, I do not visually comprehend what I see fully.

As to fine details in a print, well, I find that having both a composition as a whole to grab the eye, plus the fine details to hold the eyes attention by tickling it is what makes an image work. With a 3mx3m pointilist image, I still need to see it up close and look at the dithering style before I have seen the work (I was this way as a child). There is no sophistry in this need, it is based in my physiology.


Dierk Haasis said:
Sean DeMerchant said:
[...] get a dead sharp image
Which achieves what?
It removes disability. Without glasses I cannot see the world sharp. With a sharp image slightly within arms reach I can see the world sharp in the image without my glasses.

Dierk Haasis said:
I'd rather view images with content than a dead sharp one with an evenly spread histogram and other technical problems solved.
Personally, I have a hard time seeing any art in a blurry photo. But then, I have to do is take my glasses off and the whole world is that way. A blurry photo that does not have exceptional compositional content simply shoves my visual disability in my face and that is not art.

Dierk Haasis said:
May be the reason I don't like advertising photography, I don't like (German) contest photography, I don't like stock photography. Not all but most of them are hollow shells, posing as interesting because technically and formally they are "correct" [= adhering to rulings*].
I personally enjoy a lot of the finer advertising work. Note, by finer I mean the ones where the details tickle the eyes and composition/graphic design works (unlike a lot which I find dull). But what holds my attention are the hyper-real details that tickle the eye and having more to see that I missed when I look at an image again. A blurry photo almost never has this quality as the more to see is blurred away.

Dierk Haasis said:
A few months back I went it studio portrait, not a matter I'd see myself as proficient in, but a freind asked me for a new picture of her for job applications. The result was quite nice, capturing her love of life and her natural friendliness. Unfortunately a lot is wrong with the photo, it's slightly blurred due to camera shake, the dark background does not set her (she's a blonde) free as I thought it should, some specular highlights are a bit distracting (particularly in the teeth). The last one is not really a technical mistake, I like these highlights as they bring life into the portrait.

When printed the blur isn't visible, at worst it looks a bit like I used a very light soft-focus filter 70's style. I was very surprised when she came back congratulating me on the photo, telling me that all potential bosses she had sent her application to took note of it. Very unusual that somebody tells an applicant how good their photo is [at least in Germany, where professional studio portraits are a must for job applications].
Absolute fine detail is often a negative in portraiture as enhancing wrinkles and blackheads is not flattering to the subject. And with images of humans, as opposed to things, I prefer that image be flattering to the subject rather than being technically perfect. But portraits of people are about people. With a landscape, detail of nature, or an insect I prefer to have the details sharp. I want to see the facets of an insects eyes. I want to see the shape of a water drop on a leaf without distracting specular highlights (usually, odd specular highlights could enhance image too).

As to portraits on job applications, those are taboo her in the US.

Dierk Haasis said:
*There's a difference between rules as observed explanations of the world, and rulings as prescriptions on how the world should be. The former are, to spell it out, descriptive, the latter prescriptive. I am all for the first, against the latter.
I prefer principles as a term to observed explanations as the term for this concept.

I do not hold with any rules at all in photography. There are some physiological principles I use, but these are principles and they have no direction expect correlating with the human visual system and human visual perception in specific ways. These principles have no right or wrong except in whether their serve your goals or go against them. With these principles I have never seen an incomprehensible image.

I have seen images that others claim break the rules and found them fully inline with my principles. But, what makes them special is they chose the negation of the more common usage of one of them. Instead of emphasizing the usual, they emphasized the unusal and created something special. But there are no broken rules. And if I ever see an image that does not follow these principles, then I will know that that image has something to teach me.

In summary, as a nearsighted individual, blurry photos without detail and viewing from a distance are both undesireable things to my physiological makeup. There is no sophistry needed as that is the way my body works.

enjoy your day, :)

Sean
 

Asher Kelman

OPF Owner/Editor-in-Chief
Sean,

I'm blury-eyed too, an sometimes even without Pilsner.

For some pictures, perfection and sharp is art.

For others, a blur removes the inconsequential and leaves the shapes, tones and hues to enticew us enough that we participate in all the rest of the image and its story by using our hard-wired circuits, experience and imagination.

It allows the viewer to be trusted by the artist, as if to say, I know you understand, because, perhaps, you have been there before.

Asher
 

Dierk Haasis

pro member
I am near-sighted, too (r: -6.25, 2.50, 160; l: -4; only the left eye is completely correctable by glasses). It doesn't change anything I wrote, nothing of which has any bearing on what I outlined. It's a bit like arguing that the world to me is flat because I don't experience it as an ellipsoid. This is very obvious when we compare Sean's way of dealing with hie nearsightedness and mine. The solution to this - crop to the detail you want to view and enlarge it - was also given.

Beside the singular decisions of individuals there's another exception, one I hinted at in my original post: an artist may put in - kind of an in-joke - something that is only visible upon very close inspection. The very nature of these in-jokes is to not be visible at normal viewing distances. The Sistine Chapel, I am told, has several of these.

A lot can be said about getting the techincal side just right, having enough headroom to work with in order to express what one wants to express. Anybody taking any hobby or profession serious will go through a phase where certain technicalities have to be mastered. The end of this phase is characterised by focussing very strongly on technique; in the next phase (if one reaches it) one inevitably gets over it, not thinking about technicalitoes anymore but putting them to use.

With the advent of digital cameras we had to go through a new learing phase. The still current discussion on megapixels, noise at high ISO settings, "accurate" colour and several other things is a symptom - we don't talk about creative decisions, we concentrate on how many pixels a sensor sports. Curiously only a few photographers care to learn about output, forgetting that output technology is the limiting factor ATM.
 

Asher Kelman

OPF Owner/Editor-in-Chief
Output is like growing up, setting up a store and putting your name on it and supporting a family.

Asher
 

Harvey Moore

New member
Alain Briot said:
It can be an expensive proposition ;-) With the 1DsMk2, tilting the horizon slightly, say 10 degrees or so, will cost you 2 to 3k in deleted pixels by the time you level the image and crop out the corners...

There's a lot to be said for composing the image as precisely as you can in the field when using a digital camera. Best to use all the pixels you paid for!

I agree with this statement, however, when chasing butterflys, dragonflys, children darting about, base runner sliding in to home plate, et al, cropping for composition and esthetics often becomes necessary.
 

Alain Briot

pro member
Harvey Moore said:
I agree with this statement, however, when chasing butterflys, dragonflys, children darting about, base runner sliding in to home plate, et al, cropping for composition and esthetics often becomes necessary.

Cartier Bresson was called "Mr do not crop"... his images were hardly of static subjects. We are talking of personal approach & skills here rather than of inherent necessity. Maybe you need to crop your images. However it would be inaccurate to say that photographs of subjects in motion need, by nature, to be cropped.
 

Asher Kelman

OPF Owner/Editor-in-Chief
Alain,

We hate to waste expensive pixels.

One can very well compose with a crop in mind but it is not always possible to shoot as printed.

There is, in fact, no particular reason why a specific framing has to be used for composition. Some pictures are intended for non-standard prints. An example would be a panorama shot with a WA lens.

So, re-thinking the great Frenchman, "Mr do not crop", we must admit that today, at least, cropping in the camera is not often possible. A better instruction would be "frame tight, crop tighter".

Asher
 

Alain Briot

pro member
Asher Kelman said:
So, re-thinking the great Frenchman, "Mr do not crop", we must admit that today, at least, cropping in the camera is not often possible. Asher

I don't think it was any easier in his time. In fact, provided that he only used a small number of lenses and no zooms, it may actually have been more difficult.
 

Asher Kelman

OPF Owner/Editor-in-Chief
Alain,

For most of my experience, only has either a 35mm lens on my rangefinder camera, a 50mm lens on my Pentax Spotmatic or the provided lens on my Crown Graphic.

I always understood that one frames with one's legs. I was poorly informed then and knew nothing of zooms!

How is it I managed?

We did have other lenses. Still, there was no obsession with film size, or lenses, just choice of camera.

Why does it have to be so different with pixels? The 1D may have been just right for sports and news, yet it was retired! Result, we need more storage!

In most applications, there are no better news or sports or event pictures, with more pixels!

Asher
 
Last edited:

Alain Briot

pro member
Asher Kelman said:
Alain,

For most of my experience, I only had either a 35mm lens on my rangefinder camera, a 50mm lens on my Pentax Spotmatic or the provided lens on my Crown Graphic.
Asher

Sounds like we agree that Bresson's approach is possible for anyone who is willing to devote the time, energy and concentration needed.


A.K. Corrected my quote for grammar!
 
Last edited by a moderator:

Erik DeBill

New member
Asher Kelman said:
Alain,
So, re-thinking the great Frenchman, "Mr do not crop", we must admit that today, at least, cropping in the camera is not often possible. A better instruction would be "frame tight, crop tighter".

And realize that all your hard work may still leave you needing to crop. Yesterday I took a shot of a trail through some trees. I used a tripod and mirror lock-up. I used a bubble level to get it true. Many of the trees tilted to the left and I carefully considered whether I should try to match them, or go for true-level. I opted for true-level.

When I got home and went to work on the image, I found that I had overlooked the tiny smidgeon of horizon where the trail opened out into a streambed. It didn't look right until I got that level. The trees still tilt. The image is 5 degrees off from true vertical. Next time I'll do better and look for things that might be construed as a horizon.

In the meantime, having 8MP to start with left me with 6MP in the final image - enough that I can probably print it at 8x12 without noticing an annoying lack of detail in places. I know I won't be happy with the results if I try to go for a 10x15 (the largest I normally print). If I'd started at 6MP I'd not even be able to go for the 8x12.
 
Top