Hacker News new | past | comments | ask | show | jobs | submit login
The Day I Saw Van Gogh’s Genius in a New Light (asada0.tumblr.com)
233 points by dicemoose on Dec 12, 2011 | hide | past | web | favorite | 65 comments

I suspect that this whole exercise is moot. However one adjusts their color spectrum, I'm still staring at tiny digital reproductions of paintings on a fairly cheap LCD monitor.

One of the things I've learned by reading a book on art, and then wandering around various art museums looking at the very art that was pictured in the book, is that paint is different from ink or pixels. You can look at a poster print of a famous painting all you want, but you still haven't seen the painting. Paint has color and texture and light-catching qualities that defy easy reproduction.

For example, I wrote an essay on Vermeer when I was in high school. Then I went to see some Vermeers. And, frankly, "View of Delft" is not very interesting in print, even if you see it reproduced in the correct size, but in person "View of Delft" is amazing. It has atmosphere. And across the room from "View of Delft" in the Mauritshuis is "Girl With A Pearl Earring", an absolutely ubiquitous work, reproduced on book covers and posters everywhere, but I can't look at those posters anymore, they are hopelessly flat and dull. The actual "Girl With A Pearl Earring" is far more beautiful.

> I'm still staring at tiny digital reproductions of paintings on a fairly cheap LCD monitor.

If you've seen the originals, there is a 3D quality that is also lost. He had a way of sculpting the paint on the canvas, which I've never seen in any reproduction.

Yes! I noticed that too when looking at a Van Gogh in person. He had built it up using a lot of paint - almost like what you see when a small child uses way too much finger-paint. The physical texture of it was striking.

Not to mention the paints have translucent and reflective qualities, all of which are lost on a 2d print or monitor.

This is moot for another reason. Your eyes take in the full spectrum of color and produce 3 scalar values for red, green, and blue. One can take a work of art, illuminate it with pure white light (containing all colors, not just red, green, and blue), determine these RGB values that your eyes will experience, and store them in an image file. You can takes this file and reproduce the work on a computer monitor, on a printer, etc. It will look the same as the work of art under white light. As soon as you illuminate the work of art with different light, all bets are off. The RGB values do not contain enough information to determine what the work of art would look like under arbitrary spectrum. For extreme examples consider the following effects: http://en.wikipedia.org/wiki/Fluorescence http://en.wikipedia.org/wiki/Phosphorescence http://en.wikipedia.org/wiki/Iridescence

Conclusion: illuminating a reproduction under different light, or processing a reproduction (an image file) in a computer, will not necessarily give you an accurate representation of what the original would look like. We need more information to know if what they saw in that light room is actually what the original would look like under that light. The same goes for the examples on the page.

A great example of this for me was that modern art painting that is just a specific shade of purple or violet (I forget the name, might be "shade #456" or something). It would look like total BS on paper, but displayed in the gallery it had this very strange effect, like a black hole, or weird dead zone, a matte presence that was a distinct experience.

This is precisely right. And a great deal of modernist work is just like this... for reasons that may be obvious if you think about them: In a world with ubiquitous photography, the details that don't show up in photos are even more special, and art galleries have the time and space to focus on them.

Those paintings that seem to consist of just one black line on a field of white? They make sense in person. (Hint: Just because it's notionally a field of white doesn't mean it isn't made of seventy-five overlapping layers of slightly different shades of white paint. Just because it's notionally a black line doesn't mean the line is uniform.)

Right. I'm curious where he got the images. Certainly you just have to google image search "van gogh starry night" to see the huge array of color balance "interpretations" you'll find on the net.

That said, I did think it was interesting how for instance in the sunflowers, there are a few highly-saturated parts (the centers of the flowers) that pop in the originals, but blend into maybe a more coherent image in the version with the color deficiency simulation.

van Gogh chose paints for their immediacy of color, sometimes without understanding the problems of hue shifting and degradation displayed by certain pigments. This causes a huge headache for curators and in many cases the changes are irreversible. For example his chrome yellows contain sulphides which have significantly darkened and browned through exposure to UV. He also used various red lakes that are prone to fading and discoloration.

van Gogh famously wrote in a letter to his brother Theo: "Paintings fade like flowers... All the colors that impressionism has brought into fashion are unstable, so there is all the more reason to simply use them too brightly - time will tone them down only too much".

Any theory of color vision deficiency that attempts to reconstruct the color balances that van Gogh actually saw should take into account the hue/value/chroma of his paints such as they possessed when originally applied, and also consider that van Gogh intentionally adjusted his aesthetic to render color schemes in expectation of future pigment degradation, and that these adjustments cannot have been an exact science.

What I noticed: In each version, my focus is drawn to completely different locations.

- In The Harvest, I immediately look to the farmer in the newer version.

- In Starry Night, I immediately look to the church in the new version, instead of the moon.

- Likewise, in The Road Menders and The Sower I look at the people more, and the trees and sun less.

- In the Cafe Terrace at Night, I immediately look to the server in the new version, instead of the yellow lighted wall in the original.

I noticed I measure the person and their posture much more readily. I can now, in my mind, recall the posture of most of these people: the server is erect, the sower is slightly bent and tired, the farmer in the harvest is steadily working, if slightly hurriedly, and in The Road Menders the two nun-looking characters in the back are somewhat tired as if on a weekend.

What I learned from this isn't "van Gogh had a color deficiency," but the way in which color can affect the mood of a painting in really subtle ways I hadn't considered.


Maybe I am a little color blind and never knew it- because while I can see a difference between the two, it's not all that striking- the only difference I see is that oranges become yellows and deep blues become a little lighter. Am I the only one?

Something to keep in mind: the gamut of your LCD is a pretty crude mapping of the actual paint colors.

In the best case, on a wide gamut panel, these are very rude approximations of what you might see under the unnatural lighting experienced by the author. It should not be taken as any kind of diagnostic.

If you have e.g. a 6 bit TN panel, all of these are going to look very similar, before anomalous vision is even involved.

This is really interesting. I wish the pictures he provided were a little bigger though. I still don't know that this really improves on the originals. To me, part of the genius of Van Gogh is his unusual use of color. The "improved" versions look nice, but they also look sort of boring compared to the old versions- they are more realistic, but they also loose a lot of the whimsicality and surreally of the originals (two of my favorite traits of Van Gogh). I guess my problem with this article isn't as much the new versions of the paintings, as the authors assertion that they are superior to the originals.

Concur wholeheartedly. Van Gogh's unusual color choices are a feature, not a bug. The author asserts that the muted color pallet for 'The Harvest' imbues the painting with an autumnal glow. For me, it removes the element of the painting that I find most interesting. The other comparisons fall along similar lines.

Of course, this is art, and it's not truly possible for either of us to be 'wrong'. Taking a new perspective on Van Gogh's art is itself interesting, so this is a great read, even if I disagree with some of his conclusions.

Yes, I think it would be quite interesting to have a museum exhibition where a high quality print of a original Van Gogh is displayed next a high quality print of the filtered version. I would defiantly pay to see that

I wonder if any other famous artists saw their work differently than the rest of us?


If you have problems with this word then it may help to think of it as De[finite]ly - and tie that in your mind to "finite" meaning "measurable".

(Sorry to be a spelling nazi but the use of the 'a' in defiantly indicates this is not a typo.)

> I wonder if any other famous artists saw their work differently than the rest of us?

Ludwig van Beethoven's hearing started to deteriorate early on and eventually he became completely deaf though he continued to compose.

Of course there has to be a substantial difference between losing a sense or never to have had it but we don't know whether Van Gogh's altered vision was there from birth or the alteration happened later in his life.

I agree with both of you, most of the pictures lose a lot of character and uniqueness in the "corrected" versions; though one or two of them gain an incredibly realistic feeling of ambiance.

Better or not? Shouldn't even be the question - but it is an interesting theory and result!

There are some who argue that Beethoven's metronome was broken, and his deafness was part of the reason he didn't realize it. Here's an article about it: http://www.nytimes.com/1987/02/19/arts/critic-s-notebook-pon...

Along those lines, John Eliot Gardiner conducted Beethoven's 9 Symphonies with Beethoven's metronome markings and with instruments that would have been used at that time period. It's a very different sound that I'm a huge fan of, though I think it wasn't particularly embraced by traditionalists.

I don't think that most of the paintings look better or more interesting in the protanomal simulation, some of them actually have lost their magic and look dull. I find this most visible in the self-portrait, of which the author writes: the man whom one cannot approach easily.

The tinge of green in the original portrait makes this man much harder to approach, because it adds the psychotic, unusual aura to the picture. This man has left the world of commonly shared human experience.

The author remarks that in the picture The Road Menders the trees have a strange color and after conversion look more solid, giving some depth to the road. So the picture looks more like what we would expect, in other words: we see the usual.

But this is contrary to the kind of psychotic perception of the world, which I associate with van Gogh's paintings, at least with those paintings that we would call typical van Gogh (there are paintings from the earlier dutch period, which have a different character): Reality is distorted, colors deviate from the ordinary, everything is flowing together and swirling, the wheat, the sky, the world. Proportions are deranged, look at the painting of his bedroom for example:


Theo van Gogh, his brother, wrote 1889 in a letter to his future wife a characterization of Vincent van Gogh: "As you know, he has broken since a long time with everything what we call convention. The style of his clothes and his behaviour show that this is a special human, and since many years people who see him will say: This is a madman."

Whether it adds any thing to the article but Van Gogh was on drugs with the known side effect of mucking with your yellow and green perception.

He is specifically mentioned in this article on adverse wise effects of digoxin: http://en.wikipedia.org/wiki/Digoxin#Adverse_effects

Interesting, as there does seem to be a consistently stronger vibrance of yellow monochromatic hues in the protanomal paintings.

In programs like Photoshop you can preview your design using color blindness filters (View > Proof Setup > Color Blindness). Another trick is to grayscale your design to check the contrast. This will help a lot to create a design that is also accessible for colorblind people.

Contrast is the key here. Red for example is a very dark color. That's why they use to print secret documents on red paper. When you would copy such document with a copier it would become completely black.

Vincent van Gogh is arguably the most famous example of a painter associated with crazy; almost everyone has heard of (the rumour of?) the artist that cut off his own ear. And certainly it is not only the colors that are outlandish, boldly dissonant and flirting with the insane. Just look at the strokes, the use of perspective and the general composition of the paintings. Realism is not all there is to it! Perhaps even anything but.

Some of the altered pictures look nice in the same way that greyscaling a photo creates a pleasingly different expression, but generally, as for myself, I get impression that content has been stripped.

I'm color-blind and I see absolutely no differences between the original and the modified images.

So I assume this is working.

Same here - I actually checked the comments to be sure this wasn't a joke, actually. It's just a series of identical pictures to me. I guess I'm a born Van Gogh fan.

(I did a wall mural once involving a large green sea monster; I had a friend tell me when the shadows were getting red so I could compensate back to "normal-brown" - they all looked fine to me.)

I'm not color-blind and I must say modified images look more realistic to me (e.g. normal skin color instead of sick-person greenish). Overall colors are lighter to my eye as well.

I can honestly say that this is the most accurate simulation of my color blindness I have ever seen. Usually I can tell a difference between the two if I look back and forth, and almost always when I flip between the images quickly. But in this case I popped them open in two tabs and switched back and forth repeatedly, and could only notice minute changes if I looked very carefully at specific parts of the painting.

Likewise. The left and right versions look close to identical to me.

I was going to comment that the article was quite far fetched, since I saw almost no difference... and then I read your comment.

Edit : Oh shit. Hi Greg. Fucking genetics. It's Dad fault.

As am I, although I saw significant difference on the Harvest. The rest I couldn't really tell, but they were much busier color wise.

In some of the converted pictures two colors that were different in the original become identical.

If Van Gogh saw colors the way they are in the converted pictures why would he use two different colors that would look identical to him?

Oil paintings aren't painted in a single go. There's a lot of layering involved.

So perhaps a better question is, how would you mix the same color twice, if you couldn't tell a range of colors apart?

This is what makes me question the colour blind theory. Although they are very strong in places, the colours all work together.

If he was colour blind surely there would be patches of green and patches of non-green instead of smooth colour. How could he possibly match the same colours over and over again each time he mixed his paints if he couldn't tell the difference?

I thought about this for a while.

I'm not an expert in oil painting technique, but I think one possibility is, you tend to overpaint an area at a time. So: mix some paint, layer up part of the canvas. Come back later, mix some paint, do it again.

If that's how it works it would explain a bunch of things in the comparison images. Like: how come the bridge is yellow, but its reflection in the river is bright green? Maybe they were painted at different times.

The colors don't necessarily look identical to him. If you have trouble see in green, and two colors differ by how much green is in them, you might perceive the two colors to be closer together than someone with normal color vision perceives the two colors.

I think that's what you see in a lot of the pictures. Look at the green in the coat in the self-portrait. What appears as distinct lines of green in the normal color version becomes muted in the modified color version and blends in much more with the other colors, but with just enough difference to lend a rich texture to the coat.

I'd blame that on the author's image processing.

It's also several times more common to be green-deficient than it is to be missing that cone entirely.

Let me ask this in another way, why would you not use two different colors as one, if they look identical to you?

Why would you separately put two "different" colors of paint on your pad, when as far as you can tell they are identical?

I can understand mixing them, or using first one then the other (after you run out of one) - but shading the image partly with one color and parts with another?

The only conclusion is that they were not identical to him, and this software is not correctly showing us what he saw.

Simple: You mix the colors one way, paint for a while, go off to some other part, mix the first color again, and paint again. If you can't tell the differences between two colors, it's easy to come up with two similar/'identical' colors that are really quite different to someone with a different color perception.

While this article doesn't remotely prove that this is what's happened, it seems like a possibility to me.

The effect where colors with different spectral power distributions matches under certain conditions due to the observer or the light source, is called metamerism.


I'm colour-blind (red-green) and have, since I can remember, always had a huge affinity for the paintings of Van Gogh. I didn't realise that other people considered his paintings to be unusually toned or that he may have been colour-blind, so I found this article fascinating.

This makes little sense to me.

A partially color-blind artist has no reason to amplify is dimmed color to "correct" his perceptual deficit. His brain is calibrated on his usual input. Even if his perception of (say) red is diminished, an image with the red channel amplified will look more red to him too.

A color blind person does not perceive colors dimmer or lighter, but perceives them wrong. So for example, Van Gogh would see a wheat field and pick a color that matched the wheat to paint it, according to what he saw. Turns out that this color would be orange, because to him orange and wheat both looked the same.

As a color blind myself, I see absolutely no difference between the original and the modified images. They are 100% identical to me. So I already see them as Van Gogh did. But if you can see a difference then this proves that this experiment works.

Even The Harvest was the same for you? That one I could see a large difference. The rest were more or less the same for me though.

Looking closely at the Harvest I see that that the field is different indeed. Orange in the original, and "wheat" on the modified. But that wasn't immediately apparent at first.

The Harvest is different from the others, the wheat goes from intense orange to yellow. All others are about green being replaced with yellow.

I am partially color deficient, although I don't know the classifications given in the article. I was told it was "medium red-green" several years ago. The modified images look the same to me, except for the last two. In the Cafe Terrace, the sky looks more purple to me in the original, and in the self-portrait, the background swirls look more purple to me in the modified version.

I have normal colour vision and I experienced the same thing with those two pictures, a purple to blue shift. (I also saw differences in the others.)

I'm struck by how bland some of the altered images are! Sure, still impressive but there is a certain quality of the spontaneity of the abrupt introduction of new or contrasting color that is lost is some of the transformations. As my personal preference I had to admit I much prefer the original although that's not to say they aren't magnificent works of art regardless.

A (professional) photographer friend of mine showed me one of his autumn landscapes. When I was not really impressed (I only saw some bland, kindof greenish kindof yellowish trees), he remembered: "Oh, I forgot you have that sad color filter on"

If I don't see the difference between these images (ok, some parts are slightly brighter), does that mean I have color deficiency too ?

Yes. Or you have a really really bad computer display. Welcome to the crowd.

When you see a Gauguin, you think, This man is living in a dream world. When you see a van Gogh, you think, This dream world is living in a man.

—Adam Gopnik


A beautiful, inspiring critique if you have the time for it.

With the exception of the last example, the two versions of each painting look identical to me. Of course, I am looking at them through a 4+NS cataract, in an eye (the only one I have) that also has a damaged macula and other significant retina damage - but I suspect the cataract plays a bigger role.

I'm having said cataract removed this Friday, so I'll have to check this again next week and see how things have changed. I'm very curious to see what the difference will be. The progressive worsening of my vision over the past year has been so gradual that it's sometimes hard to quantify just how much things like my ability to see color and contrast has changed.

"The Harvest" and "The Starry Night" both look under-illuminated. The rest look more or less the same.

Not sure whether that's my deuteranomaly talking or if it's his small pictures.

This is such a bizarre exercise, albeit interesting in terms of trivia, i guess.

"""Finally I feel that van Gogh’s astounding qualities are available to me."""

I've often felt that nerds (including myself here) shouldn't be allowed near artistically expressive mediums and equipment, such as cameras, because we nerd them up way too much and miss the whole point. And this quote just sums up what i mean. :S A minuscule colour shift reveals Van Gogh's genius? NOPE.

Someone who isn't me once ate some funny mushrooms and Van Gogh's paintings appeared to him as though they were made with magic pixie paint that actively ebbed and flowed right before his eyes. That was the day he saw Van Gogh's genius in a new light. Does that mean Van Gogh only painted after eating similar funny mushrooms?

He mentions this color vision deficiency simulator app: http://asada.tukusi.ne.jp/cvsimulator/e/index.html

It's interesting to look at real life through color deficient eyes.

"protanomal simulation" sure looks like an unsharp mask (with some contrast added) to me

OT but I really like that blog theme, very clean and easy to read

I have normal colour vision, but the Van Gogh self portrait in the Musee D'Orsay is one of the most beautiful paintings I've ever seen.

I think the modified versions look worse in every case, and I question the author's sanity.

You can only appreciate Van Gogh if you can make his color choices more "normal"? So you'd only appreciate Picasso if you could use photoshop to reconstruct his paintings?

Take the work as it is. Attempt to understand it on its own terms. That is how you grow as a person.

Hm. But to folks with different color vision, they 'are' different paintings. Neither view is 'normal'. An attempt to reconcile views on the same painting is entirely rational.

All the examples look mostly the same to me. I am mildly color-blind.

Anyone else think this article is a plug for his app?

So you argues he painted wheat the color you see wheat, even though he does not see it in real life the same color as you see?

Now I'm confused.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact