Do pixels matter? What 20" x 30" prints reveal -the saga

Started Jun 22, 2013 | Discussions thread
plevyadophy Veteran Member • Posts: 4,258
Re: Print

Mako2011 wrote:

plevyadophy wrote:

Mako2011 wrote:

cpkuntz wrote:

I could see the difference in the whole image as well as in these crops. Maybe you couldn't but overall the image on the right is sharper and therefore more striking. Pixels matter.
Just out of curiosity, Mako, which cameras were used to make these?

D70 and D7000. My apologies for not having a more appropriate camera for the comparison. Again, at 8x10 it's very hard to tell one from the other in most scenes. Also here the comparison is same subject and side by side. Put the prints in different rooms and very few will pick up on the difference

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Are those cams even of the same generation?

If not, then I don't think it's a good comparison.

Why? There is no low MP current gen camera to compare to. The original premise was that lower MP bodies can't be distinguished from newer gen higher MP bodies at 20x30. What bodies would you use to compare?

I am sure there is something that can be compared. And I wouldn't use "low MP count" as a definition because what is low MP is very subjective. Rather I would opt for one "lower" MP count cam and one much higher. So use, lets say, the current MP daddy, the Nikon D800 and then compare it with another Nikon cam with say half as many pixels (and even that introduces a slight problem because the D800 sensor, as far as I am aware, is designed by Nikon themselves but fabricated by Sony, whilst all of Nikon's other sensors are both designed and fabricated by Sony; and I did say that it would be a better test if the only major variable, as far as the sensors are concerned, is the difference in pixel count).

The thing I noticed on the blow-up of the dog's nose is that the left piccie had major probs dealing with the white highlight. That doesn't seem like a pixel issue to me but rather one of dynamic range.

Yes, DR will be another advantage of newer gen. That's another discussion that's interesting.

My take on it is that more pixels do make a difference just in the same way that a larger engined car makes a difference but ...................... just as with a larger engined car, a higher pixel count doesn't often make a difference;

Not a question of how often...a question of can it be noticed. I can easily notice the engine size dif when compared side by side. (at 20x30)

I think many of us can ............... if we put our "experienced-gear-head-nerd-photographer" head on. For example, I recently went to a Sebastio Selgado exhibition; absolutely amazing, probably the best, certainly one of the best, exhibition of images I have ever seen. I noticed when I put on my "experienced-gear-head-nerd-photographer" head that I could easily see noise and possibly which images were shot on cameras that are not anywhere near today's state of the art. But once, I relaxed, and viewed the images as they should be viewed (at an appropriate distance and taking in the image as a whole) it was simply impossible to tell the quality of one image from another; in fact his work is so damn good, that one is captivated by the majesty of the imagery rather than noticing any of the technicals (in fact I would go so far as saying that if someone went to his exhibition (the current Exodus exhibition) and came away having only or mostly picked up on technical aspects of the imagery, I would immediately make a call to the emergency services in fear that planet earth had been infiltrated by alien beings from another planet because ............................ surely such a person is not human! )

you have to be using the tool in the specific circumstances where the differences matter to see the difference (e.g. prolonged driving at high revs and fine detail blown up for large display)

Agreed...that's why here and in the original thread, 20x30 was stipulated.

Depending on image content i don't regard 20 x 30 as particularly big; it's roughly A2 in metric paper size. I think it is the border though; it's the tipping point at which things BEGIN to make a difference. Any larger than this, and in my view, the pixel count does matter.

I do think that at this size, depending upon the image content, one would see no OBVIOUS difference between say a 12 Mpix shot and a 24 Mpix one (that didn't require me to wear my "experienced-gear-head-nerd-photographer" head). For example, I am a people shooter and I doubt this paper size would result in much difference between these two pixel counts. However, if I was shooting a fine detail macro project and I knew that cropping into the image to pick out extra details would be often required for the project, I don't think I would trust the 12 Mpix cam for such a job.

I think your comparison would be more meaningful if you compared cameras of the same generation, using sensors from the same sensor manufacturer and with the only difference being the pixel count.

Result would be the same...I suppose you could crop using equivalence to compare in that case.

I doubt very much the result would be same. A 16 Mpix Canon 1Ds Mark II barely outperforms a modern Olympus OM-D, or put another way a modern tiny sensor is now nearly as good as yesteryear's premium full frame sensor cam. Thus in your comparison I think it would have made more sense to compare cams of the same generation.

Plus looking at images on screen is different (more flaws show up) from what one sees in a print (prints are more forgiving).

Not really. Prints can often more telling depending on the paper/processed used and size printed. Can also be less noticeable, as you point out, if one decides to print on canvas.

I have seen a blow up of an ISO 60 image (I think that was the ISO) from a little Canon G10 that was being displayed at a trade show by Canon reps on their wide format printers stand. I stood there in disbelief; I even went so far as to demand to see the original raw file (which the rep duly showed me). The image of course was shot at low ISO and in broad daylight. However, I was amazed at the quality of the print which was (if my memory serves me correctly) larger than this discussion's reference size of 20 x 30 inch.

Agreed, people are often surprised at how large you can print when you do the enlargement well and also print using the best techniques.

Maybe if the same subject matter was taken with a high end higher pixel count camera, and the two images displayed side-by-side, I may have spotted a difference;

That is actully my only point....side by side (same subject/printed large) it is easy to tell.

Hmm, I wonder. When we say "easy" I think we should be careful. For example, I often hear people say driving is easy. And I always challenge them by breaking down the individual components/skills that it takes to drive a car and then ask them if they thought they had those skills, or did they have those skills, before they passed their driving test. After my little exercise they are not so keen on insisting that driving is easy; it's simply easy to them NOW, given all the skills they have NOW absorbed and which have become second nature. I feel the same happens with photography, in that our brains are tuned to home in on something if we have some experience and then we make out that what we have honed in on is easy to spot.

in any event, I stood there for some time staring at the picture looking for some obvious flaw, before eventually admitting defeat and giving up.

Well done prints are indeed something to be admired.

The problem with that is ........................ if we believe that, then we automatically give away our age because ....................... the younger generation couldn't give a toss about prints as most of their perception of imagery is what it looks like on a digital display. I like to pretend that I am younger than I am so ssh! don't mention that I like prints.

Oh by the way, I noticed that you didn't address the last paragraph of my previous post :

"An interesting article on this pixel thing was written some time ago and makes for interesting reading, here it is: You've Got to be Kidding! No - I am not."

Any comments/observations from that link?

-- hide signature --

My opinions are my own and not those of DPR or its administration. They carry no 'special' value (except to me and Lacie of course)

Keyboard shortcuts:
FForum PPrevious NNext WNext unread UUpvote SSubscribe RReply QQuote BBookmark MMy threads
Color scheme? Blue / Yellow