I can highly recommend John Berger's legendary documentary series Ways Of Seeing. The first episode explores precisely this issue - how our perception of painting is affected by reproduction. It has perhaps the most arresting opening scene of any documentary series.

https://www.youtube.com/watch?v=0pDE4VX_9Kk

My favorite painting in the world is Jackson Pollock's Mural, his first major painting. It's owned by my alma mater the U of Iowa. The canvas is 9x20 iirc, just huge. They used to display it in a large central gallery, where you could look at it from more than 50 feet away, or get your nose right up to it.

One of the great things about this painting is how it is a totally different experience at different distances. From a distance, the sense of right-to-left motion is palpable. Get closer, and figures seem to pop out of it. But really close, you can no longer see the entire painting, but you can see individual brushstrokes (mostly 3-6" house painting brushes, I think). The subtlety and detail of each stroke is extraordinary. There are patterns across the canvas that are a result of large-scale patterns of strokes - it's almost like cursive handwriting.

Another of my favorite examples of the "reproduction doesn't do any justice" is Mark Rothko. Like Pollock, he's often held up as example of an artist whose work "doesn't look like anything" and "my kid could do that." And it's true, if you're looking at a poster of a Rothko, it just looks like some large splotches of color.

But, in a gallery, people encounter the Rothko and just stop. Many of them will stare at those large splotches of color for 1, 2, 5, even 10 minutes. Turns out there's a whole lot going on there that just doesn't translate into print.

Sadly, a lot of the techniques that abstract expressionists used don't translate well into the future, either. Layering just oil paint, that's pretty durable. Layering different kinds of paint, not so much. I've seen a fly dislodge a flake of paint from one of Rothko's paintings just by landing on it.

You might be interested by this video about the challenges faced by conservation scientists with Rothko murals : https://vimeo.com/111469325

I had a similar experience seeing Summertime Number 9A in the Tate. It just looks like meaningless squiggles when reproduced on a postcard, but in person you can see the structure, the layers of line and shape.

You may be interested in this video by the Getty Conservation Institute on the conservation of Mural:

https://www.youtube.com/watch?v=EWAbVpkV0jQ

Ah, yes, I watched this series a few years ago and just realized that it may have inspired my post above a bit.

PS: I just shamelessly throw in another excellent BBC documentary series here: How Buildings Learn, by Stewart Brand [0]. I sometimes see it mentioned here on HN, but it cannot be mentioned too often. If you haven't seen it, please please do.

[0] https://www.youtube.com/watch?v=AvEqfg2sIH0

Thank you, that is truly excellent and has spawned a productive rabbit hole exploring topics I've been considering.

Oh, yeah, and waves in the water of those hyper-detailed Italian 16th century paintings, or tree leaves -- when lit correctly, you get specular highlights that seriously look like sunlight glistening on there. Blows me away every time.

Those paintings are quite low-resolution on average, but it took me a while to figure out how they could add so much resolution where it mattered, and turns out it's because they have an entire extra dimension to vary things in.

> (I also distinctly remember being shocked by the size of some paintings, especially the crowning of Napoleon (https://en.wikipedia.org/wiki/Coronation_of_Napoleon_I#/medi...) This painting was printed in many of my history school books, but I was not even remotely aware that is nearly 10 meters wide.)

And yet, it's the postal stamp-sized Mona Lisa that gets all the attention...

Sure, it's a good painting, but I personally find it to pale in comparison to the other works on display at the Louvre.

The painting opposite of Mona Lisa is the The Wedding at Cana. It's one of the most impressive paintings in Louvre (I think it's the largest). I have wondered if they placed it there as a joke.

There is over 100 people in the painting being busy and only person looking outward directly towards Mona Lisa is Jesus looking surprised. Mona Lisa is looking directly at Jesus and smirks.

Hah, I will now believe that this was indeed placed there as a joke.

There are actually a few more people than just Jesus looking out towards Mona Lisa, some with amusing expressions. It is just a rather busy painting, so it can be hard to notice.

There's a place for both. Rothko's Seagram Murals are bland and unremarkable in the pages of a book, but become utterly overwhelming in person due to their sheer scale. Conversely, Sir Charles Bell's watercolour sketches of battlefield wounds are so affecting on part because of their small size. When stooping over to peer at these pocket notebook sketches, you are transported to a field hospital in Waterloo, standing over the shoulder of Bell as he quickly but expertly records the carnage of war in intimate detail.

It was not to say that small paintings do not have their worth. My point was more that Mona Lisa, specifically, seems... unremarkable in comparison to the other works.

I'm noticing an increasing amount of rude comments on HN these days. Please, keep it civil.

I read a book about the Revolutionary War called "Washington's Crossing" by David Hackett Fischer. The wrap-around cover of the book is, appropriately, the iconic painting "Washington Crossing the Delaware" by Emanuel Leutze (https://en.wikipedia.org/wiki/Washington_Crossing_the_Delawa...).

I learned that the actual painting was nearby in NYC at The Met. When a friend visited, I wanted one of my goals was to go to The Met and see the painting. I entered the American Wing close enough to closing time that I worried I may not find the painting in time. I started wandering, turned a corner, and sat dumfounded starting at the very thing I was looking for, sitting 21 feet in width at the end of the hall.

Extremely interesting, thank you! But my point still stands. Consider this image in ultra high definition:

https://artmyn.com/partners/sothebys-newyork/Michele_Mariesc...

If you drag the image around, you notice that the water has different highlight points at different angles. If you would stand in front of this painting, your natural body movement (of even a tree before the window) would constantly change the lighting and almost certainly result in a very natural-looking water surface. I cannot imagine reproducing this digitally without it feeling "artificial" or "forced".

I wonder how well this sensation would be conveyed by viewing a lightfield capture in VR? This is meant to be the sort of thing that that's good for.

It's not impossible, computers have been modeling 3D fluids for a long time. It might be computationally intensive, but definitely not impossible.

He's talking about reproduction, not modeling. You can model the painting's physics all you want, you still won't be able to make a faithful reproduction on a flat and gamut-limited LCD screen.

With head tracking it should be possible to get close. The thickness changes are small, so the 3D effect is mostly from the uneven light reflection revealed by head movement, not stereopsis.

The texture of oil on canvas is extraordinarily complex and subtle. You would need billions of pixels and polygons and trillions of rays to even begin to capture the heavy impasto of Van Gogh or the coarse diffusion of Bacon.

> You would need billions of pixels and polygons and trillions of rays

So give it 5 years, and you'll have it on your phone...

If there's one thing I've learned over the decades about computer technology, it's to never say "impossible" or "never"...

Eventually, what you once might have thought to be only in the realm of "big and expensive" inevitably becomes "cheap and commodity" in a far shorter time than you expect.

That was one of the rather unexploited features of the Amazon Fire Phone, which I loved for that very reason. Unfortunately, it didn't see much use, but you could take pictures that could capture depth (created by rotation) and it was so cool.

Plus you can prerender 99% of the lighting.

Or even just take ten million photos from almost the same spot, presenting the one that matches the user's subtle movements.