Do I need a D800E if I have an M9 and Leica lenses? I am confused.

[theguywitha645d]

"I would recommend you to learn about the concept of standard viewing distance. Of course it would be better if you read what I said--I did not refer to "experience," whatever that means."

I know all about  the technical definition of "standard viewing distances" but how often do you actually hold an 8x10 print ten inches from your face to view it and for how long?

Standard viewing distance is proportional to the size of the print and does not just refer to 8x10.

By "experience" I mean actually doing things in the real world - in this case looking at prints of various sizes in the way most people look at them. I suspect the average actual real world instead of pixel peeper distance is more like 15 to 20 inches for an 8x10 to 11x 17 inch print. Next time you are reading a magazine check the approximate distance you hold it from your face. By experience I also refer to how much a print dominates our angle of view and hence our experience of it as an object that we are taking in, or beholding.

When it comes down to evaluating prints at the end of the day  I'll come down on the side of sensual experience every time over a mere technical definition.

So, you really don't understand the concepts of standard viewing distance and its relevance. Basically it says that viewing distance changes with print size. So as the print gets bigger, the viewing distance increases. Most people don't measure the diagonal of a print and then place themselves in front of it, but they don't poke their face into the print either. What this comes down to is that quality perceived in a print is relative and based on viewing distance and the angular resolution of the eye. This is why pixel resolution is not a limit to print size because the viewing distance is going to be such that the viewer will be not be able to out resolve the print. And if you really understand the concepts, you will also know how to use the concepts for your own effect.

While is nice to discuss "sensual experience," it is really nothing that address the issue of print size. And you dismissal of "technical definitions" really show you don't understand what these definitions are pointing to nor how to apply them. BTW, pixel peeping is actually a closer viewing distance than standard which is a huge problem from which folks have made erroneous conclusions.

[BJL]

My observations of how most people (other than print-sniffers, and their offspring, the pixel peepers) view gallery prints of various sizes is that it falls between the two often stated extremes.
it is neither
- some fixed distanced like 10" or the viewer's minimum comfortable focusing distance, as assumed in arguing for a common PPI target for all print sizes
nor
- some fixed apparent (angular) image size, as implied by a standard like viewing distance equal to print diagonal size.

What I seem instead is that large prints are in general viewed from further away, but not so much further away as to completely neutralize the size difference: instead, viewing is often done at a distance that achieves a larger apparent size (angular size) with large prints. But, again in my limited experience, a great majority of even very large prints are rarely viewed from much closer than about the short dimension of the print, so about half the diagonal.


(For me, this suggests that about 4000 pixels along the short side, so a MP count in the low 20's, is almost always enough for the final displayed version. But I am a big fan of "capture oversampling" for flexibility, aliasing avoidance, easier demosaicing, and so on.)

[theguywitha645d]

The other observation I may make is that 6MP is probably perfectly good for 8x10" prints viewed at 25cm, so I cannot really see the reason any guy with a 645D would spend 10-20kUSD on an equipment, unless they would print large and view close.

Best regards
Erik

Because there is more to an image than simply resolving power. We can detect single lines thinner than the resolving power would suggest. Also the relationship of format size to contrast is important. If you think all images perceived are equal regardless of format or resolving power or contrast, or a myriad of factors that make up the final image, then I would say we have very different experiences. I bought the 645D primarily for what a large sensor would impart to the image--this is a personal choice to factors I like to see in my work. I really did not care about the pixel resolution--I was also thinking of a 22MP Phase/Leaf back. Others have found a different solution for their photography--I would not suggest any solution is "better" than another, just different.

My comment is simply in regards to is whether there is a real limit to print size. If the OP likes the images from his Leica, but is only thinking that he needs a D800 for larger prints, then I would say keep shooting with the Leica and enjoy large prints from it. The qualities imparted from the Leica are going to be more important than the number of pixels it has--although that is part of the look.

[theguywitha645d]

I do a lot of printing for researchers, artists, museums and such. I print all kinds of files at all sizes--nothing larger than 44"x144" at present. I have not found pixel resolution the limiting factor. And in cases of some really silly enlargements (web image to 6 feet), you can mask the pixels fairly well by resampling and other techniques. Making your picture big does not change it--if it sucks as an 8x10, it will suck at 20x30 just as much. A good image will remain good. Everyone that I have come to me for printing that believed pixel resolution was a limiting factor to prints size have never left holding that idea.

Another things folks sort of forget is viewers don't expect detail and sharpness to be endless nor absolute. So if they do stick their nose into a print, it is not a negative experience--they expect it. There is a robustness to our perception of a 2-D image--we don't find perspective strange when viewed from an angle, although a literal reading of perspective would suggest we should. We don't have an absolute expectation of quality--Holga images have regions that are perceived as sharp and detailed, even though by a comparative standard they would not be.

Life is short. Print big and have fun.

[Ray]

Another view is that for perspective correct viewing the viewing distance would be focal length times magnification. So I have a pano shot with 22 mm lens on APS C. My print is 50x100 cm, and it was shot with the camera in vertical position. So the correct viewing distance would be 500/23.5 * 22 = 468 mm or around 0.5 m. Looking at the image from 2 m is nice, but looking at 0.5m the viewer sees the image in correct perspective, it's like being there, an immersive experience.

Really? That's interesting, Erik. If there is a correct viewing distance dependent in part upon focal length of lens used, then surely one would have to admit that focal length of lens has some bearing on perspective, as well as position. Yet I've seen you comment to the effect that FL of lens has nothing to do with perspective.

How do you reconcile these two positions? 

[theguywitha645d]

Standard viewing distance is equal to the diagonal of the print and apparent perspective changes based on the angle of view of the optics.

Correct viewing distance is proportional to the focal length of the optics and print magnification. Apparent perspective will remain constant.

The standard and correct viewing distance of print from a normal lens (a focal length is equal to the format diagonal) are equal. The correct viewing distance for a wide angle is less than the standard viewing distance and, for a telephoto, greater than the standard viewing distance.

The object distance ratio when the image is taken sets the image size ratio of the image, the true perspective.

[Rob C]

I think I realise where the different points of view come from: we appear to be speaking about different types of perspective.

One perspective is that given by the subject to camera distance, which unless changed, is the same with whatever normal (as opposed to fish-eye, for example) lens we use, the only difference being magnification on the sensor/film will cover a smaller or larger area of the figure; the other perspective is the one that we create for ourselves once we make a print and put it on display. That perspective can look good or fairly awful dependent on the size of enlargement we produce. For example, a shot of a seated person made with a wide-angle lens on 35mm cameras, that may look good as an 8x10 can look strange when blown up to 5ft in height. In that case, a longer lens would have produced a better look had we still been hoping to make that 5ft print.

These effects can be seen in large fashion pictures used as point-of-sale bait. People sometimes appear to have oddly shaped proportions... and no, it doesn't always seem to correct itself by getting closer or further away - maybe the shops aren't long enough for that?

So in that sense, there are different things affecting perspective, but as I'm not that good at music, I shan't attempt to amplify.

Rob C

[madmanchan]

I thought the correct viewing distance had something to do with the length of one's nose ...   

[theguywitha645d]

I thought the correct viewing distance had something to do with the length of one's nose ...   

Or the number of pixels in your monitor. ; )

[theguywitha645d]

I think I realise where the different points of view come from: we appear to be speaking about different types of perspective.

One perspective is that given by the subject to camera distance, which unless changed, is the same with whatever normal (as opposed to fish-eye, for example) lens we use, the only difference being magnification on the sensor/film will cover a smaller or larger area of the figure; the other perspective is the one that we create for ourselves once we make a print and put it on display. That perspective can look good or fairly awful dependent on the size of enlargement we produce. For example, a shot of a seated person made with a wide-angle lens on 35mm cameras, that may look good as an 8x10 can look strange when blown up to 5ft in height. In that case, a longer lens would have produced a better look had we still been hoping to make that 5ft print.

These effects can be seen in large fashion pictures used as point-of-sale bait. People sometimes appear to have oddly shaped proportions... and no, it doesn't always seem to correct itself by getting closer or further away - maybe the shops aren't long enough for that?

So in that sense, there are different things affecting perspective, but as I'm not that good at music, I shan't attempt to amplify.

Rob C

Even simpler than that. If I make an image at a certain view point, the perspective will appear normal (as if I were at that point) if I view that image from the same relative view point. When I view at a distance other than that point, then the appearance of perspective changes--if closer, then the perspective is weaker, if further away, the perspective appears stronger. At extremes we start to notice strange things: the Wide-angle Effect where round objects seem stretched away from the optical axis--it is not technically distortion because if you view the image from the correct viewing distance the object will then appear round (a similar thing happens with writing on roads that appear OK from a car, but not the sidewalk). It also happen when you use extremely long lenses, objects appear unnaturally flat and compressed.

Ever wondered why you could tell the difference among images photographed with different focal length, or to be more accurate, different angles of view? That is why.

[ErikKaffehr]

Hi,

No objections, but we often use perspective for effect.

The other point is that if we assume 12 mm on FF (widest I have) and print at A2 size (which I normally do) the "standard viewing distance would be around 20 cm, but if I use 800 mm on FF (the longest I have) 13.3 m.

One interesting observation is that in movie theaters the seating is fixed but they still use zooms...

Best regards
Erik

Standard viewing distance is equal to the diagonal of the print and apparent perspective changes based on the angle of view of the optics.

Correct viewing distance is proportional to the focal length of the optics and print magnification. Apparent perspective will remain constant.

The standard and correct viewing distance of print from a normal lens (a focal length is equal to the format diagonal) are equal. The correct viewing distance for a wide angle is less than the standard viewing distance and, for a telephoto, greater than the standard viewing distance.

The object distance ratio when the image is taken sets the image size ratio of the image, the true perspective.

[ErikKaffehr]

Hi,

What I wrote was that it is another view, never wrote that I share it ;-)

It's interesting if you happen to make a panoramic picture. You may use a medium focal length but achieve a very wide field of view.

Anyway, I admit that you have a good point ;-)

Best regards
Erik

Really? That's interesting, Erik. If there is a correct viewing distance dependent in part upon focal length of lens used, then surely one would have to admit that focal length of lens has some bearing on perspective, as well as position. Yet I've seen you comment to the effect that FL of lens has nothing to do with perspective.

How do you reconcile these two positions? 

[Ray]

Standard viewing distance is equal to the diagonal of the print and apparent perspective changes based on the angle of view of the optics.

Correct viewing distance is proportional to the focal length of the optics and print magnification. Apparent perspective will remain constant.

The standard and correct viewing distance of print from a normal lens (a focal length is equal to the format diagonal) are equal. The correct viewing distance for a wide angle is less than the standard viewing distance and, for a telephoto, greater than the standard viewing distance.

The object distance ratio when the image is taken sets the image size ratio of the image, the true perspective.

Oh my Gawd! For a couple of centuries billions of people have been viewing hundreds of billions of photos from the 'incorrect' distance, and therefore experiencing an 'incorrect' perspective on the compositional content of the images.

It's no wonder there's so much confusion in the world. 

[hjulenissen]

My observations of how most people (other than print-sniffers, and their offspring, the pixel peepers) view gallery prints of various sizes is that it falls between the two often stated extremes.
it is neither
- some fixed distanced like 10" or the viewer's minimum comfortable focusing distance, as assumed in arguing for a common PPI target for all print sizes
nor
- some fixed apparent (angular) image size, as implied by a standard like viewing distance equal to print diagonal size.

I think that this is often (generally?) true. People will move further away from a large picture than a small picture (viewing distance is not constant), but not so much as to negate the increased size (FOV is not constant either).

I am an amateur. I dont print images for pixel-peepers or theoretics. I print for my own sake (and those I choose to share my images with). If anyone are able to see loss of detail or the pattern of my inkjet at nose distance - good for them.

-h

[theguywitha645d]

Oh my Gawd! For a couple of centuries billions of people have been viewing hundreds of billions of photos from the 'incorrect' distance, and therefore experiencing an 'incorrect' perspective on the compositional content of the images.

It's no wonder there's so much confusion in the world. 

You are using the term "correct" incorrectly. It is only "correct" if you want to preserve the view as it was seen/made from the point is was viewed/taken. In fact, you would not want folks to view your images at the "correct" viewing distance if you think lens choice in regards to angle of view is important for the apparent perspective it gives. In those cases, it would be right to say the correct viewing distance would be incorrect for the right reasons.

[theguywitha645d]

Hi,

No objections, but we often use perspective for effect.

The other point is that if we assume 12 mm on FF (widest I have) and print at A2 size (which I normally do) the "standard viewing distance would be around 20 cm, but if I use 800 mm on FF (the longest I have) 13.3 m.

You mean the correct viewing distance. The standard viewing distance would just equal the diagonal of the print area.

One interesting observation is that in movie theaters the seating is fixed but they still use zooms...

Best regards
Erik

Most photographers like the change in apparent perspective that the angle of view of the optics gives the image. So maintaining a fixed viewing distance allows a viewer to see the change. (I also think it would be disconcerting if the movie screen or theater seats moved around to try to maintain a correct viewing distance--I think the sale of buttered popcorn would fall off for a variety of reasons.)

[AJSJones]

I think the simple answer is that the viewer decides on the viewing distance(s) - the big picture initially and then perhaps gets drawn in close by the details, if they survive scrutiny  -once it loses detail they'll back off again..

[Ray]

You are using the term "correct" incorrectly. It is only "correct" if you want to preserve the view as it was seen/made from the point is was viewed/taken. In fact, you would not want folks to view your images at the "correct" viewing distance if you think lens choice in regards to angle of view is important for the apparent perspective it gives. In those cases, it would be right to say the correct viewing distance would be incorrect for the right reasons.

Oops! Confusion again. Consider your following statement: "It is only "correct" if you want to preserve the view as it was seen/made from the point it was viewed/taken."

I think we need to stress there's a clear distinction to be made between the perspective as viewed by the photographer with his naked eyes, and the perspective seen after raising camera to eye, or when viewing the camera's LCD screen in Live View mode.

If I've understood your point, you are saying that the 'standard' viewing distance of approximately print diagonal reproduces the perspective as seen by the photographer as he looked through the viewfinder, whereas the 'correct' viewing distance is one that compensates for the apparent change in perspective that has resulted from the use of a lens which is different to the standard lens, whether wide-angle or telephoto. Is this right?

To give a specific example to clarify the matter. Yesterday I was viewing the setting sun in a clear sky, from a roof-top bar. The sun looked impressively large and red. I raised my Nikon D700 with zoom set at 14mm, to my eye, and suddenly that large, fiery, red ball appeared pathetically small; hardly more than a mere speck within the composition.

Now supposing I make a modest sized 24"x36" print from such an image. According to you, the 'standard' viewing distance would be about 44". Call it 4ft. From such a distance I should see the same 'apparent' and 'fake' perspective that the photographer saw through his viewfinder, ie, a pathetically small sun completely dwarfed by impressively large objects in the foreground. Right?

However, if I want to see the perspective as actually witnessed by the photographer before he raised the camera to his eye, or the perspective that he would have seen through the viewfinder if the camera had had a standard lens attached; that is, if I want to see the one and only true perspective in accordance with the inviolate laws of Geometry, Physics and the origins of the universe, as opposed to the fake, unreal, trick perspective created by any non-standard lens, then I should view the 24"x36" print from a distance of about 10", from which position most of the content of the print is relegated to peripheral vision and cannot be clearly seen without turning one's head. Is that the idea? 

[theguywitha645d]

Ray, that is basically right. The usual example is given with an 8x10 image printed at 8x10 so the correct viewing distance becomes the focal length of the lens. Double the prints size, double the focal length for the correct viewing distance.

But whether you are simply looking at the scene with your eyes or through the viewfinder does not matter. What the idea of correct viewing distance is saying is that not only the view point (distance you are from the objects) is important, but their angular relationship to the viewer--are they 10 degrees from center, 50 degrees, and so on. So the correct viewing distance puts the objects in the image back into their angular position from where the photograph was taken. In extreme cases like with wide angle lenses, you can actually make the "stretched" objects at the corner of the frame appear normal by viewing the image from close enough--and you will have to turn your head or look out through the corner of your eye.

Folks will try to argue that since the image size ratio is fixed when the image is taken and so the view distance is irrelevant. But the image is a projection and were you view that projection will change how the image of those objects is perceived, see the attached image. The viewer position does change the angular perception of the blue and red object.

At a standard viewing distance, the angular relationship to the images is not natural if the correct viewing distance is different. We can basically perceive the difference in viewing the projection at a distance that is different to the one relative to the view point.

[hjulenissen]

What would the correct viewing distance/print size be for a 180 degree fish-eye image? I imagine that if it is rendered flat (as most prints are), you should print infinately large (or squeeze your noise into its center)?

-h