Saturday, December 26, 2009

Daybreak Blues

What happened to the blue in Daybreak?

Maxfield Parrish painted his most famous work Daybreak in 1922. It appeared on several editions of art prints and on the cover of Coy Ludwig’s book from 1973. That book, and all those prints in antique shops have a decidedly “Parrish Blue” tonality (bottom image).

The painting was sold in the 1920s and was hidden for almost 50 years in a yacht near Boston. It surfaced again in the 1970s, where it was sold a couple more times.

In all the reproductions since then, it has a much more golden and violet appearance (top image). I saw the original when it was shown at the Norman Rockwell Museum retrospective exhibit, and it really did have that golden/violet appearance.

If you put the two side by side, at first glance it seems like a difference in color balance. Maybe all the early prints were wrong, printed from a transparency that was shifted too far into the blue range.

Maybe all the warm colors faded out of all the reproductions. Maybe the “Maxfield Parrish blue” was never there in the first place, and was a consequence of bad reproduction. But Parrish personally approved the early reproductions.

Let’s see what happens if you adjust the color balance sliders in Photoshop and make the “new” version more blue. Now they look a bit more similar, but there’s a crucial difference.

The value organization of the new incarnation is completely different from the original version. Note how the milky cerulean water above the sleeping girl’s legs now is a very dark value (1). Where that promontory in the valley used to separate from the shore (2), now the shadows are much darker.

And where the values of the blue shadows in the far mountains used to be much lighter than the bending figure, now they’re darker and more uneven (3). Changing the color balance sliders shouldn’t affect the value organization.

Is it an alternate version? Parrish dealer Alma Gilbert, in her book on Parrish’s Masterworks, referred to rumors about a second version, but put them to rest with the confirmation by Parrish's son upon seeing the painting again in the 1970s.

Was the painting overcleaned? Did some overzealous conservator remove some blue scumbles or glazes? Normally when you clean a painting, you take off the smoky, dirty, orange and brown layers and the yellowed damar varnish, and the painting gets more of a pure blue look. It doesn’t usually get more yellow, brown, or violet. The cleaning of the Sistine Chapel (above) intensified all the colors.

Parrish’s unfinished paintings showed he worked in various layers of pure pigments. Here he left a blue underpainting of a tree, with warm glazes applied on the tree at left. Did he work warm-over-cool glazes and cool-over-warm semi-opaque milky glazes on Daybreak? Is the new edition of the painting closer to the original product of Parrish’s hand? If so, how can we account for all those early approved prints?

I don’t know the answer. It’s kind of a mystery. Maybe some of you can shed more light on it.
-----------
There’s a good discussion of the painting at Jim Vadeboncoeur’s site here and here.
Sistine Chapel restoration controversy.

18 comments:

Larry said...

If it's not due to the magenta and black fading faster than their yellow and cyan counterparts, then my guess is that when originally printed in CMYK, the Magenta was left out totally and the black was printed light. That would give the best explanation for not just a shift in color but the subtraction of intended values. Try this, in Photoshop, convert the image to CMYK, Image-Mode-CMYK, then in the Channels pallet drop out the magenta by clicking the eye to the left of the pallet. close but there is still value in the lake, try highlighting the black withing the same channels pallet and adjusting the value, Image-Adjust-Brightness/Contract.

Larry said...

On closer examination, The drawing on the mountains on the far right and the architectural detail on the base of the right column also appear different.

Mary Byrom said...

I agree with you Larry. Its clear to my eyes they are two different paintings.

Andrew Wales said...

Interesting painting detective work once again!

Have a Maxfield Parrish Blue Christmas everybody!

Mary Bullock said...

Something else that has to be considered is Maxfield's technique - it was unlike most oil painters. He would not mix his colors - but painted in clear, pure glazes - allowing the eye to mix the colors. But here is the kicker - between each layer of colored glaze he would put a layer of varnish. It took him a long time to complete a painting this way, but he worked on many paintings at the same time - during the drying time on one, he would be working on another. So perhaps all these layers of varnish that are "unreachable" below the glazes, could explain some of the color changes?

Jim Serrett said...

A interesting post.
Could it be that the original printed images were done from a piece he developed in his monochrome underpainting technique.
And later after the publishing finished the work with more warm colors and values.

Tristan Alexander said...

Looking at the book "Masterworks" that shows the one picture he left unfinished after strating to rework it. The book shows the original print and the unfinished reworking (which is a more modern better photo) and between them there seems the same sort of color shift and weird detail change and yet it is the same picture. I think it was simply some odd effect of color change that happened in that time when printing his art. I have a very good print of "Daybreak" (an uncut puzel actualy)and it has stronger color than what yopu are showing in the blue version. The water area is darker and not milky at all. Some contrasts seem odd compared to better photo of the original, but I think in the end it was something odd and specific having to do with the printing process of the time.

BobN said...

It seems possible that MP revised the work after it was photographed for printing. He was very familiar with the relatively new 4-color process, and adapted his color choices and methods with reproduction in mind. He considered Daybreak his masterpiece, so I wouldn't be surprised to learn that he reworked it after the printing process was done.

Zaknafain said...

The skyholes in the trees dont match.

Theres a big one in the red version on the left edge thats missing in the blue one.

Julie Baroh said...

I have one of the original prints (my great grandmothers)that has gotten very little direct sunlight. We had it cleaned and restored and the colors are rich with blues and greens, no violets at all. It's more like third image in the post. The yellows are more golden than Hansa yellow.

When selling the prints, keep in mind they backed them with non-archival cardboard, very acidic, so that could explain the paler colors of some of the early reproductions. According to my great grandmother's diary, the prints sold were pre-framed, often in pairs (I have Garden of Allah as part of the pair). Garden of Allah was displayed in an area that got sunlight and therefore is slightly lighter than Daybreak.

BobN said...

In fact, the last example in the post is a reworking of "October". According to Ludwig, "After publication, the artist removed much of the original painting with the idea of making some major changes in the composition. The panel remained unfinished at the time of his death."

He also removed figures from several paintings after they were published, because he liked them better as pure landscapes.

thokitts said...

Part 1:

When I look closely at both images I can't find any difference in the contours of the figures. Or differences in the 'touch points' between the contours of the figures in the foreground and the landscape in the background. In addition, the patterning of Parrish's complicated foliage -- well, the leaves and flowers look absolutely identical across both images. This leads me to believe we are looking at one piece of art where the color has changed over time. Not at a painting which was reworked, or two different versions of the same painting.

However, I do see pronounced differences within the shadow masses of the painting. Most notably, in the amount of detail articulated in the distant mountain shadows, and in the differences found in the core shadows modeling the two pillars. (Examine the left pillars in each image and you will see what I mean.)

From here on I speculate, but all the extra detail in the later image makes me consider yet another possibility -- that there maybe some form of pentimenti involved. Parrish painted this image in 1922 and at that time he was known to like working on panels covered with a lead ground. In time, all linseed oil lead grounds shift towards the yellow, and darken with age, and that would affect any transparent layers laid down on top. Parrish may have also sandwiched semi-transparent layers of white between his glazes to create or accentuate the atmospheric perspective. Those layers of white would become more transparent over time. Such changes would reveal darker colors and texture underneath, if they were in fact originally there. (Actually, all oil colors increase in transparency over time to some degree, not just white paint. It's an effect caused by the oil itself, and the effect is responsible for the unintended pentimento you occasionally come across in an old master's work. Can start to appear within 50 years, or take as long as 200. Depends on the pigment and chemistry involved.)

thokitts said...

Part 2:

I understand Parrish was, and still is touted as a painter who used only transparent glazes all the way down to the white of his substrate, but that level of purity seems rather extreme, especially since Parrish was known to be experimental with his materials. (He was also known to paint in his figures on top of a completely finished background. That alone would require some blocking out of the background with opaque white paint.) Parrish often brought in tools and techniques from other artistic disciplines (printmaking, for one) Few painters are as rigid with their creative processes over their entire lives as many biographers and technique-hounds like to imply, so it seems conceivable that Parrish might have stippled or pounce on a veil of white into the shadows of those distant mountains -- and other areas as well -- in an effort to modulate the amount of texture and detail he needed, before glazing again over the area in question. It's a technique that predates Parrish by centuries, and if the veil is laid down very thin and smooth, it will be difficult-to-impossible to distinguish it from the base white of the substrate below. Besides, the purpose of such veils are to control or diminish contrast (detail) -- not to eradicate it entirely. Or call attention to itself.

Such veils would be susceptible to pentimento over time. Especially if they have been ladened with a lot of oil, or over-painted with a lot of glazes.

Is that the answer? Dunno. As I said, I'm speculating. But it is something to consider along with everything else that has been suggested. We'll never know without examining a paint chip under a microscope. But to do that would destroy the place the sample came from, and that is too high a cost to pay.

BTW, on a different note, the craquelure you so often see in Parrish's work is created by the intervening layers of oil and varnish pulling apart. It's unavoidable, since the two materials were sandwiched together and 'dried' at a different rate. Plus, linseed oil shrinks over time while a varnish does not. Despite what Coy Ludwig wrote in his 1973 monograph, a lot of Parrish's work is poor shape and was so then. Maxfield's heavy use of driers and siccatives, heat boxes, and summer exposure to direct sunlight, have left their toll.

Parrish gave us brilliant and beautiful work. Jewels right up there with any Vermeer. I love looking at a Parrish anytime I have the opportunity to see one. Yet I feel sad to see so much of them coming apart. Nothing lasts forever. Not even art.


Thomas Kitts

http://www.thomaskitts.com

James Gurney said...

Thomas, thanks, that's a really insightful analysis. I agree that it's the same painting. The colors remaining look like an underpainting.

If it's a transparentized scumble or glaze layer, it sure has lost a lot of opacity. That's why my first guess was that it was overcleaned. If the surface was cracked (and you're right: a lot of Parrishes are in bad shape), I could imagine someone wanting to strip off an outer layer.

Thanks everybody for all the comments. As some of you have suggested, the extreme difference between the two could have a lot to do with technical issues of photography and reproduction. Maybe the original photo/repro somehow exaggerated the cyan/white haze, which was never quite so pronounced in the original.

I guess the lesson is that you can't trust reproductions, and
as you say, Tom, paintings don't
last forever.

thokitts said...

James:

The other issue I didn't explore had to do with the lithographic technology that was available to reproduce MP's images in the '20s. I'm not a an expert on early pre-press processes but I have been around a lot of offset-lithography since the '80s as a graphic designer, and I know that back in the '20s the way plate separations were generated had more to do with a skilled handcraft than memetic cold technology. So it is entirely possible that the blue-green shift of the shadows came from a deliberate modification during pre-press preparations, decided by someone down the production chain, or, for what we see in Ludwig's book, a subsequent re-screening of an earlier print. Do we know anything about what source artwork was used for reproducing "Daybreak" in Ludwig's book? Did Ludwig have access to the original, or, as I mentioned is what we see his book just a re-screening of another print? That would bring in all sorts of other potentials...

It's certainly worth nothing that the beauty of MP's work comes through, regardless of the quality or accuracy of the reproduction, and perhaps that is intimately connected his lasting popularity.

Thomas

http://www.thomaskitts.com

sfox said...

When I was studying illustration at the then Academy of Art College in San Francisco in the late 1980s, I used to hit the galleries on Sutter St. at lunchtime. I walked into Blackwell's (I think it was. It's no longer there.) one day and, in a back room hanging on wall behind a desk, lo and behold, was "Daybreak" in all its original glory.

I was told that it was being held there while the owners were getting divorced and deciding on custody. The very nice gallery lady told me that I was welcome to come in and look at it any time I wanted, which ended up being six or eight times at least. Just "Daybreak" and me, all alone in that room. Sigh.

I was able to get right up to it and noticed in one corner that I could see the whorls of (probably) Parrish's fingerprints on the surface. Was he pressing down a ridge in the varnish? That's what I've always thought, but I don't know.

As to color, it was glorious beyond belief and seemed backlit like a giant transparency. It was jaw-dropping. I remember it being rich with pink and lavender shades, not even remotely like the yellow/blue prints I had seen up to then.

Having been a sign painter/graphic designer for some years by that time, I assumed that exposure to light had knocked out the red tones in the old prints.

We always double-coated the reds on signs that were going to have a southern exposure because that was the color that would fade the fastest.

I later recall reading somewhere that the painting had been sold at auction for $475,000. So I guess neither of them kept it.

Thanks for bringing back to me one of my favorite art school memories.

thokitts said...

sfox:

If you have a copy of the Ludwig book re-read the section on technique towards the back. In it you'll find it describing how MP used his fingers to wipe a glaze. Leaving fingerprints. I know you know what you saw. Just thought I'd mention it fits with the facts in that book.

Must have been nice to have Daybreak all to yourself.

Thomas

http://www.thomaskitts.com

sfox said...

Thanks, Thomas! I will go look at the book again.