Tim’s Vermeer…from a painter’s point of view

June 21st, 2014

After I posted various reports about the documentary film Tim’s Vermeer, a few readers encouraged me to give a scholarly assessment of Jenison’s claim that Vermeer had used an optical device called a comparator mirror as an aid to his painting. Given my limited knowledge of the use of optics in seventeenth-century painting, I found it more appropriate to examine the issue from a technical viewpoint, since I am by profession a painter. The fact that I have studied Vermeer’s painting technique and attempted to emulate his manner for over 40 years, I hope, might give me a discreet edge over non-painters in evaluating if Jenison’s device is or is not compatible with what we know of Vermeer’s pictorial strategy and technical procedures.

Following some lively discussions with Mr. Jenison on the finer points of Vermeer’s painting procedures, I was able to meet him in Texas and experiment with the comparator mirror on the premises of a full-scale mockup of scene of Vermeer’s Music Lesson which Jenison had built in order to test his hypothesis by paintings his own Vermeer.

My first attempt to use the comparator mirror was frustrating. Not only was I unable to produce acceptable pencil outlines of a black and white photograph with which Jenison had used in his first experiments, I was utterly incapable of matching on paper any of the photograph’s tonal values. To use the comparator, at least as I was attempting to use it at the moment, one is constrained to work within an extremely small area of the drawing, along a thin edge where the image of the comparator mirror abuts on the drawing below and the two can be compared. Initially I found this procedure mentally and visually stressing, and at odds with my experience in conceiving and making paintings.

With a little more practice I was able to produce a few acceptable contours, even though they lacked any sort of artistic quality. However, seeing that I am not particularly skilled with a pencil, I though it best to test the device with paint and brush with which I have greater familiarity, even though on first consideration the oil painting technique seemed even more at odds with the mirror’s limitations than with dry drawing.

Surprisingly, I made rapid progress with the oil medium. Although with a certain fatigue, I learned to define first simple and then complex contours with a fine-tipped brush and began, even more surprisingly, to marvel at how it was possible to match with utmost precision both the chromatic and tonal values of my painting with those of the mirror in a completely objective manner.

Having made substantial progress in coordinating mind, eye, brush and mirror after a few painting sessions, I started afresh and began to depict a small portion of Jenison’s Vermeer mockup Vermeer room following what I have come to understand of Vermeer’s multi-step painting technique. Beginning with a schematic line drawing which served to fixed the most salient contours of the scene, I first underpainted the lights and darks with monochrome brown (raw umber plus black) and white paints without, however, systematically consulting the comparator mirror. I was, in fact, interested in testing how close I could get to the correct values on my own.

Once the underpainting was thoroughly dry, I began to apply the final colors over it using thick opaque paint in the lights and thinner paint in the shadows, according to seventeenth-century prescription. In order to render a given passage I first mixed, as all painters do, the proper paints on the palette attempting to match them as closely as possible to the color and tonal value combining what I perceived in nature with I had learned through practice. I then applied the mixture to the canvas and compared the values of my paint to those of the corresponding passage in the mirror. I sometimes discovered that both the color and tone of my mixture were very close to those seen in the mirror, but just as often I was struck by how poorly I had interpreted nature notwithstanding my decades of experience. In a back and forth manner I was able to register the erred values of my work with those of the mirror and return to painting. Once the proper values were firmly in place, I freehanded most of the modelling as I would have done without using an optical aid, taking care to verify the accuracy of my progress via the comparator mirror at regular intervals. The comparator mirror was also of help in verifying difficult contours and defining the smallest details that I had been unable to capture by freehand.

In any case, once I had registered the values of my painting with those of the mirror, the passage appeared much more true to life (painters simply say “right”).

Although the set of mirrors and lens (Jenison’s used a double convex lens of the camera obscura in coordination with a concave mirror and a comparator mirror) requires periodic adjustments in order to view the different areas of the scene, this does not unduly interrupt the painting process once one has acquired the necessary skill maneuver them.


Given Jenison’s complete lack of painting experience, he painted his Vermeer employing the comparator mirror, as would be expected, in the most literal of manners. He painstakingly matched what he saw in the mirror with paint applied directly, alla prima, forgoing any sort of layering techniques that we know Vermeer and his more accomplished colleagues sometimes employed. This aspect of Jenison’s approach provoked considerable criticism, including my own. It was reasoned that Vermeer could not have used the comparator mirror because Jenison’s essentially paint-by-numbers technique, and the consequential one-layer paint structure gotten by such an approach, is completely at odds with the multi-layered structure of Vermeer’s paintings.

According to my experience the comparator mirror neither dictates nor limits the painter to any fixed procedure or techniques, including those used by Vermeer. Certainly, it would interfere no more with the creative painting process than a systematic use of the camera obscura.

If it is used in a “painterly” manner, as any experienced painter would be naturally inclined to do, the comparator mirror opens the possibility to study color more precisely than can be done with the camera obscura alone and allows the artist to match with remarkable efficacy the illusive tonal values of nature, which in effect are crucial to Vermeer’s unique brand of realism. Furthermore, I discovered that the erred tonal values of my monochrome underpainting did not compromise the rendering of the proper tones and colors of the final paint layers. The aim of seventeenth-century underpainting, as I understand it, was not to establish the precise tonal values of the final work from the very beginning, but rather to approximate the distribution of darks and lights thereby creating a sort of compositional blueprint which provided a solid base on to which the successive layers of colored paint could be applied in a more efficient manner. Although I used the glazing technique in only one passage (red madder over an underpainting of vermillion), it was evident that with some practice it would be relatively easy for any practiced painter to anticipate the tonal and chromatic values of the colored underpainting so they might eventually match those made visible in the mirror once the passage had been glazed with the final color.

The use of such a simple device as the comparator mirror in tandem with the camera obscura lens, in my opinion, is technically compatible with Vermeer’s known painting techniques (to be distinguished from his “pointillist” mannerism), and it is in line with what Lawrence Gowing appropriately called the artist’s “optical way” as well as the artist’s search for absolute tonal authenticity.

550 Responses to “Tim’s Vermeer…from a painter’s point of view”

  1. Louis R. Velasquez

    I am uploading 2 line drawings. They are self explanatory.
    Both are based on Fig. 60 in Prof. Steadman’s book.
    I took liberty to make some changes and make some guesses.

    One diagram shows the projected image into the Camara Obscura
    As it would appear against the vertical back wall.
    It would be inverted upside down and mirror reversed.

    The second one is how it would appear in a horizontal position
    on the canvas on the flat horizontal table- after being projected with a flat mirror..
    I did not want to have to redraw all the perspective lines.
    So this one is just the cropped image.
    Of course both projected images would be in color.

    Vermeer had to straighten out all distortions caused by the lens.
    Horizontal lines above the center point become smiles.
    Those below become frowns. Vertical lines left and right of center
    become barrel like as they bulge.
    He used a straight edge ruler to establish the correct lines with paint.
    I believe he did the corrections outside the Camara Onscura.

    I made a criss-cross of the painting to show its center point.
    The letter ‘A’ is that point.
    Prof. Steadman has said Vermeer liked to offset the
    ‘vanishing point’ slightly off center.
    I had a poor quality reproduction and had difficulty
    having the perspective lines meet at one vanishing point.
    So you see two of them /// one is labeled 1 and the other is 2 .
    Vermeer stuck a pin at his vanishing point to get accurate perspective lines.

    I believe Vermeer experimented in his painting procedures, like all creative artists,
    and painted his monotone, in both positions, vertical and horizontal.
    He solved his mirror-reversal in various ways.

    I believe the simplest way to solve the mirror-reversal is to project the image onto a horizontal position. Once the monotone was finished being painted – with paint, not drawn with pencil- inside the camara obscura, the colors were applied in studio light with the canvas placed vertically on an easel, EXACTLY as the mirror reflection in the Music Lesson describes his work method to be. Unfortunately, I am an “idiot” with analytic geometry, and I am unable to
    Calculate the exact position of Vermeer’s easel, his vertical painting and stool, and box, as they should be included in these two diagrams.
    Louis R. Velasquez

  2. Louis R. Velasquez

    second diagram

  3. Louis R. Velasquez

    sorry–just pretend the diagram is inverted–as it would be inside the camara obscura on a vertical wall=

  4. Louis R. Velasquez

    Here it is—INVERTED and MIRROR REVERSED as seen vertically–inside the CO

  5. tim jenison

    Louis, this is why you must come to San Antonio to experiment with the machine. It would save so much time. Heck, I’ll buy you a ticket!

    The “seahorse smile” curvature is not in fact present in the camera obscura projection. It appears only when the concave lens is used. The simple 17th century lens projects remarkably straight lines (in lens jargon, it is “rectilinear”). Tomorrow I will try to upload some photos that show this.

    A more fundamental problem with your “semi-dark camera” theory is that you seem to be saying that you could paint a tonally accurate underpainting under the projection. It is not possible to make a tonally accurate underpainting on top of a projection in either a totally dark room, or your “semi-dark camera”. Many have tried.

    We talked about this briefly in the film with the quick demo showing blue and white cards under projected light. People assume that a camera obscura image can be easily turned into a painting. Intuition tells us that we should be able to paint on top of a lens projection.

    Daniele Barbaro was maybe the first to express this idea in 1568.

    “Seeing, therefore, on the paper the outline of things, you can draw with a pencil all the perspective and the shading and colouring, according to nature, holding the paper tightly till you have finished the drawing.”

    But this seductive idea doesn’t work well at all in practice because as I mentioned earlier, the paint and the projected light modify each other. Shapes are easy to trace on a projection, but duplicating values and colors is nearly impossible. The comparator mirror solves the problem completely, which is the crux of my experiment.

    The diagram you drew represents the geometry of “The Music Lesson” using a large 45 degree mirror. The comparator mirror used to make an underdrawing gives exactly the same results. What’s your point?


    HI TIM
    Welcome back.
    You have have posted three long letters. I will comment only on a couple of points.
    T: Heck, I’ll buy you a ticket!
    L: You are too generous! I’ve made my point very clear Tim. Once you paint the city of Delft..under the very relaxed conditions I stated…then….I will go to Texas. :), and I will buy my own ticket. Glad to see you are still talking to me.

    T: By the way, just to correct a minor point made by Louis Velasquez, I did not use a large 45 degree plane mirror to make my initial under drawing. I used the small comparator mirror without the concave mirror as I pointed out in my post on October 1st, 2014 at 10:31 pm.
    L: Several persons mentioned the RAZOR concept- the easiest simplest method- Not long ago you posted a photo of you holding the large flat mirror, inside the Camara Obscura at a 45 degree angle. One sees the image, projected to the horizontal position on the table…..the image is quite clear..but still fuzzy..but you admitted you never had heard about a second lens , the FOCUS LENS….this second movable lens brings that fuzzy image into clear focus….by moving it back and forth. ..like a telescope or microscope, both known to Vermeer.
    IN ANOTHER post, Jonathan wrote and I think you concurred, that drawing with the comparator mirror is quite quite frustrating , difficult, and requires much concentration with constant adjustments of your ” apparatus”.
    Does not the RAZOR tell you to choose the easy simple method of simply tracing an INTRICATE outline of the image as it lies there on that table? Why do it the hard way, as you now say you did it? And why did you even post that photo with the flat mirror? More, why are you holding it with your hand? I would think it would be lodged in tight by some mechanical apparatus like the monstrosity your lens is encased in so you do not accidentally BUMP either.

    T: The “seahorse smile” curvature is not in fact present in the camera obscura projection.
    L: That is not what Steadman’s book says…..and was it WICK who also posted his recognition of this bending of horizontal and vertical lines in the optical effect? Look again at Steadman’s photos of the reconstructed studio ….he used a simple lens….the foreground tile is woefully distorted , and Steadman says ( paraphrased).. Vermeer was right to trim it out , out of the bottom edge of the Music Lesson painting.

    T: The diagram you drew represents the geometry of “The Music Lesson” using a large 45 degree mirror. The comparator mirror used to make an underdrawing gives exactly the same results. What’s your point?
    L: The razor concept..KISS… Keep It Simple S—–d



    To your comment:
    T: A more fundamental problem with your “semi-dark camera” theory is that you seem to be saying that you could paint a tonally accurate underpainting under the projection. It is not possible to make a tonally accurate underpainting on top of a projection in either a totally dark room, or your “semi-dark camera”. Many have tried.
    L: You, my friend, are not an artist! And your comment proves it.
    I have no desire to ever paint what you called a , “tonally accurate underpainting under the projection.”. I, like Vermeer , am a creative painter….there is NO NEED for a ” tonally accurate” monotone..not ever, not never, to be painted! You made a pencil drawing..where is the tonal accuracy of that? I have painted so many paintings over the years..there is never any NEED for the monotone to be tonally accurate. But you are the one who said you were an ” idiot” at painting. Your comment proves your point, And I say that with respect. Good try, but no cigar , my friend. The projection serves ONLY for me to make a foundational image… Of my chosen decided composition..it has no bearing on the colors I will apply on top of that brown monotone. The is no need for value gradations in the monotone. All that’s needed us FLAT paint. .if you see that small self portrait I posted..I used three basic FLAT values of tempera paint to underlay the foundation for the oil colors that I applied. The tempera dried in less than a minute…I was able to proceed with oils instantly.
    Cordially, LOUIS


    BTW TIM, That small self portrait I posted…using three flat values…was done without a Camara Obscura projection….lIiF I had made that self portrait using the Camara Obscura…I would not have used all three values..I would fave used just one..any one…of the three. The way Vermeer did in that photo you posted of The girl with earring…one flat brown tone….. You don’t see Vermeer concerned with a ” tonally accurate” monotone, do you? and like Vermeer, ..I would not have been any too careful of getting down the exact contour lines or shapes. He made many corrections in the color over paint stage. Yes..you are NO artist. Do not be confused by Jonathan’s post of his monotone of the girl reading the letter. It gas tonal variance..But it was not painted inside the Camara Obscura. One more thing TIM, there is even a process to paint a flat brown image the way the image of the Shroud of Turin, is observed to be…that image is in the NEGATIVE. An artist can paint, in the negative fir good reasons of vibrancy of the final coloring. Areas ( shadows in reality) are left white or pale to reflect from under to give the dark oil paint over layers a vibrancy from beneath..like what Van Eyck did…. And areas that in reality are in LIGHT, are painted in dark tones, so that opaque scumbles and impasto oil pain to top gains a depth and the creation of ” optical grays”. Sorry to go over your head TIM, but you are a video man…not a painter.
    Cordially, LOUIS

  9. craig

    Does anyone have a good color image of Steadman’s reconstruction of the Music Lesson? If so, could you post it here?


  10. W. Wick


    Thank you for your comments. I agree that a drawing with paint could go undetected. I believe other drawing mediums can go undetected too, particularly against certain colored grounds. Conservators call this “fugitive media”. I think you have to allow Tim’s graphite pencil lines, since the same could be made with fugitive media. I think one of the difficulties in painting a grisaille under a projection is controlling the contrast, because the light of the projection would falsify the contrast of your painted highlights and shadows – though with practice, I think it could be done. So yes, I agree a projection could be used for a sufficient monochrome underpainting.

    Some further thoughts: In my collage, you’ll see the Canaletto painting of Campo Zanipolo, Venice (1741), and the corresponding reference sketches believed to be done on site with a camera obscura. He would likely have used a set-up similar to yours, though probably a one-lens system. It would have a 45 degree angled mirror, horizontal table, darkened by a tent (there are many period engravings to be found of this style camera online). Canaletto’s CO sketches seem less confident than his beautiful freehand sketches, but that wobbly quality is something I recognize from years of tracing on a ground glass (for reasons other than making paintings). One tends to shut down the “thinking” part of the act of drawing as you struggle to find the right contours and contend with the visibility of your marks within details of the projection. Note there is no light or shadow information in the sketches, no ground details, no moving objects such as people or boats. Color notes are written in his hand in the margin. And though he made a 4-sketch panorama (presumingly by swinging the camera), he only recorded about 50% of the scene. The painting itself was made from a completely reconstructed perspective grid, as it’s clear he changed many of the proportions of elements to balance the composition. It would appear he transferred the sketch information to the painting by eye or measurement, rather than tracing literally line for line.

    Obviously, unlike Canaletto’s super-wide views, it is possible for Vermeer to capture the entire schematic of the Music Lesson and most of his other paintings with a single projection (if not with the same lens). But to what extent would he be satisfied with that projection? We have few references to actual objects like we see here. For example, If you look at the bottom panels, you’ll see a closer view of Canaletto’s single-projection view (the sheets are 6″ x 9″). It is the most complete sketch of the sequence, yet in the corresponding painted section next to it you’ll see the heighth of dome is elongated, the proportions of the plinth has been changed, the size of the windows diminished (not to mention surface detail added not seen in the sketch). So even though this segment is a relatively complete and coherent drawing, it would have been premature for Canaletto to paint a grisaille at this stage, even if this represented the complete painting. He still wanted to square up the image and refine the schematic first. True, Canaletto can’t move a building like Vermeer can move furniture, but in my own studio experience, especially with narrative scenes, you can’t always get “the projection” you want, especially with your back up against the wall. Mind you, I’m not trying to disallow the CO, I’m just trying to give Vermeer the flexibility Canaletto had.

    With perspective construction, Canaletto achieved very convincing, deep perspective, aided by the light colored paving–stone strips, which can still be seen today in the Google satellite view. The final perspective of the overall scene is impossible to achieve optically, and the sunlight and all the highlights and shadows are a complete invention, as sun could never be seen at this angle from the North. Incidentally, this painting was acquired by Windsor Castle from John Smith, who also owned the Music Lesson at the time this one was painted.

  11. Louis R. Velasquez

    WICK, FACINATING!!!! simply fascinating! Thank you. Creative artists will always find the way…of that … there is no doubt. I propose that for Vermeer to complete the minute micro-fine details of the virginal…he first—finished the painting to his satisfaction. Then, he took that virginal outside on a nice bright sunny dry day…and took a portable CO with him. Once aligned to get the brightly sunlit virginal projected into his CO, he simply traced the lines==or–painted them carefully. Either that, or he used various small flat mirrors INSIDE his studio, to bounce bright sunlight onto the details of the virginal–making the details bright enough to see and trace. ITS ALL PURE INVENTION!
    That’s how creative artists ..CREATE!
    cordially, Louis

  12. tim jenison

    This picture was taken by taping some graph paper on the camera obscura screen. I used a Canon S120 to take a time exposure. The lens of the camera bends the graph paper a bit, but you can see that the 17th century lens keeps the lines straight. No seahorse smile.

  13. Jonathan Janson

    I fear we are getting sidetracked in minutiae and questions completely irrelevant to the core of the debate, even though they may be undoubtedly interesting for a minority of us.

    For a moment, can we please take stock and examine a few problems which seem to have been lost in the “fog of war”?


    1. In your opinion, precisely in which paintings did Vermeer use the comparator mirror?

    2. And for each painting, can you please tell us what evidence proves or suggests he did so?

    3. For what reasons and how extensively did Vermeer use the comparator?

    4. Can you substantiate with objective evidence what you maintain to be Vermeer’s superiority in evaluating and rendering tonal value? Or is it essentially a subjective evaluation?

    I believe this last question echoes a question Craig asked various times, Jon answered, but you did not (if, naturally I have properly read all your comments).

    Craig’s question was:

    Is there some sort of objective standard by which we can tell when a painter is using a comparator and when he is not?

    Naturally, you could respond to each question with a small book, but for the sake of new or confused readers like myself I hope you can reply to each in a few lines so we know exactly the principal points and overall extent of your hypothesis?

  14. Jane Jelley

    I think it is good to get back to the core arguments.
    I would like to say also, that the picture Tim just posted of the harpsichord as seen through the camera obscura alone, looks to me to be much more like the painted result Vermeer actually achieved: not exactly exact and detailed, but soft and slightly fuzzy. It would be possible to trace this, as it is, through the camera in dim light, and make judgements of placement and tone, but the result when ‘printed’ would be more impressionistic than Tim’s precise rendering in his picture.However, once out of the camera booth, the painter can always use direct observation on top of first tracings to ‘correct’ or sharpen up the result.
    I do not think that Louis’ ‘portrait’ had relevance here, not only because I did not enjoy the content, which I thought lacked respect; but because the underpainting he demonstrated had little tonal polarity. There is no evidence at all to suggest that Vermeer used tempera in the first stages of his painting. Analysis has found only oil based media.

    I think that it is very unlikely that Vermeer would take the harpsichord outside as Louis suggests. It was far too valuable and expensive and would had to be treated with great care, particularly if it was borrowed. The weather is changeable in Northern Climes. It might be sunny one moment, rainy the next. You would not want such an instrument to get wet.
    The lower layers of painting did not need to be detailed, and it is possible to see that Vermeer’s underpaintings lacked line. So a ‘fine’ underpainting such as Louis suggests is not necessary, and also does not match Vermeer’s known technique.

    Jane Jelley Oxford UK

  15. tim jenison

    OK. I’ll try to keep it as short as possible. 5 volume book to follow later.

    1. In your opinion, precisely in which paintings did Vermeer use the comparator mirror?

    The one I studied in depth and have the most confidence in is “The Music Lesson”.

    For a variety of reasons I have the least confidence in:
    “Saint Praxedis”
    “Christ in the House of Martha and Mary”
    “Diana and her Companions”

    2. And for each painting, can you please tell us what evidence proves or suggests he did so?

    As I said in the film, there is no proof Vermeer used optics. There is no documentary evidence.

    My project was to test the usefulness of the comparator mirror in making a 17th century painting. As such it is akin to the field of “experimental archaeology”.


    Some of the evidence I see for Vermeer’s use of a comparator mirror in “The Music Lesson”:

    There is strong correspondence in values and dynamic range between Vermeer’s painting and the physical room, and therefore my painting.
    Using the comparator is consistent with Steadman’s argument and very effectively solves problems with it.
    The comparator was highly effective in rendering the details of the virginals which could not be seen in the camera obscura.
    Curvature found in the virginals decoration in the Vermeer is consistent with distortion caused by the concave mirror in my optical setup.

    3. For what reasons and how extensively did Vermeer use the comparator?

    I would speculate that Vermeer saw it as a powerful tool to achieve realism.

    I believe Vermeer would have used the comparator in all major stages of the painting.

    The underdrawing would be done in a dark camera obscura using the comparator without the concave mirror. This allows the entire surface to be seen and outlines to be drawn or painted. It is too dark to be very useful for details or tone matching. Alternatively, a large mirror could direct the camera obscura image onto the canvas for basic underdrawing. I used the former for the simple outline of the room and windows, and the latter for the outlines of the harpsichord. I preferred using the comparator because the large mirror is awkward to work around, especially at the top of the painting.

    As you found, the comparator is useful in the monochrome underpainting stage. It’s interesting that Vermeer’s underpaintings were apparently highly similar in tone and detail to the finished painting, and not necessarily monochrome. According to Melanie Gifford on p192 of “Vermeer Studies”, Vermeer began “The Music Lesson” with “an extensive colored underpaint”. (Gifford credits Viola Pemberton-Piggott of the Royal Collection for this information)

    4. Can you substantiate with objective evidence what you maintain to be Vermeer’s superiority in evaluating and rendering tonal value? Or is it essentially a subjective evaluation?

    The close tonal match between the real room and Vermeer’s “The Music Lesson” is a start. This type of experiment could be done on certain other paintings by Vermeer and others, using miniatures à la Steadman or 3D software to model and render the rooms for comparison to the paintings.

    For example, I am pretty sure the de Hooch painting you cited earlier,


    could not tonally map to a real scene because the dynamic range is far too high, simultaneously showing a brightly lit exterior and a dimly lit interior.

    5. Is there some sort of objective standard by which we can tell when a painter is using a comparator and when he is not?

    Not with 100% certainty. Because the artist is free to depart from slavish copying in the mirror, there is a lot of ambiguity. It’s a matter of likelihood, not proof. It’s a bit like trying to determine if a modern painter started with a photograph.

    That being said, there are many different types of clues, not only in Vermeer’s work but others, that a comparator was used.


    Please do not post any further questions to me. I will no longer be responding

  17. Jane Jelley

    Are you able to qualify your last sentence and tell us what sort of clues you see in Vermeer’s work and others that confirm that a comparator was used?

    On the same page of Vermeer Studies you quote from (P192), Melanie Gifford also says that for, ‘A Lady Writing’, “Vermeer began the face with a loosely brushed, rough-textured underpaint, laying out highlights and shadows….in an exaggerated contrast of light and dark. He built up the final painting in a number of discrete stages.”
    Further down in the same column, Gifford says that Vermeer “created differences in tone between direct and reflected light”

    Can these observations from technical examination fit with the use of the comparator? Although Gifford also tells us that some of Vermeer’s underpaint had some hue, would you not agree it had a particular function as a ‘tonal map’ in Vermeer’s painting?


  18. Jane Jelley

    PS In her notes (p198 Vermeer Studies), Melanie Griffiths acknowledges that she is not making a clear distinction between a monochrome underpainting and a ‘dead layer’, which would have paints made with earth colours (that could subsequently be overlaid with more expensive, brighter pigments).

  19. tim jenison

    I’ve been discussing quite a few of the clues here on the blog. It really would take a chapter or two to lay them all out and explain them properly. There are arguments pro or con about each of them, as we have seen here. To me, as an amateur detective, they add up, others disagree.

    Steadman writes about the evidence for the camera obscura, and both our theories build on that. But proof is a pretty high standard. As I say it’s a matter of likelihood.

  20. W. Wick


    As always, thank you for your willingness to put up with us. As you know, I’ve expressed some skepticism about Steadman’s argument, but let’s just say I’m on the fence about certain aspects of it and leave it at that for now. I have no problems assuming a camera obscura could have been used to get the basic outlines of things – as you can see I’ve been looking into Canaletto’s use of the device.

    Picking up on Jonathan’s request to get back to the subject of tone, I’d like to mention I’ve noticed it’s difficult to meaningfully assess the tonal differences in side by side comparisons. Inspired by Craig’s observation that the window well nearest the viewer was among the brightest surfaces in the picture, I decided to make an experiment by converting both Vermeer’s original (taken from the NGA site linked on the home page of Essential Vermeer ) and your painting to grayscale to eliminate color differences and concentrate on tonal values and, in effect, study the LIGHT coming through the window of both versions. To further understand how yours differs, I made a tonal adjustment layer to bring your version as tonally close to Vermeer’s as possible. You can see those adjustments turn on and off in the gif, which in turn helps us instantly see the COLLECTIVE differences between your light and Vermeer’s. We can also better appreciate how darkening the periphery of the picture helps guide the eye toward the center of the picture, whereas in your unadjusted version, the eye gets hung up in brighter details around the edges. In photographic terms, this is not a bad feature; the light looks beautiful coming in the windows, the detail of the painting on the far right is great. But Vermeer wasn’t making a document, he was telling a story, so I think it’s safe to say he controlled the tonal scheme with artistic intent. An analogy can be found in Jonathan’s earlier experiment of putting back the “missing” white tiles to illustrate how Vermeer may have intentionally de-accelerated what Jonathan called the “violent” regression of perspective, a trope often used by earlier artists. So if we can think of Vermeer as not just an inventor, but a sophisticated artist (as I think we must), we see how he would have a design strategy for tile, and a tonal strategy for light. For me then, the questions this raises for the tile and tone are similar: does Vermeer experiment by laying down actual tile, trying the checkerboard first, then replacing every other one with black, or does he just draw use the grid system and paint them in? Does Vermeer shutter the windows here and there to try to control his tonal scheme with light, or does he just depart from a his basic natural-light set up and make up the rest?

  21. Jane Jelley

    I want to thank Jonathan for hosting this discussion which has been quite fascinating and frustrating in equal measure: fascinating to see how many different approaches people have taken in commenting on Tim’s experiment, and frustrating because of my inability to follow all the arguments, because of gaps in my knowledge.

    I think that the thread may have finally come to the end and that we will now have to agree to disagree with our opinions. But what a tremendous amount I have learned in the process. It has been a very good journey in excellent and learned company. Thank you too.


  22. craig

    W. Wick,

    Do you think you could do a similar side-by-side with Tim’s painting and Steadman’s photograph? I was hoping to do something of the sort, but struck out in my attempt to find a workable version of Steadman’s image.

  23. tim jenison

    There lots of loose ends in the thread here that I meant to address. I have been slammed at work and in personal matters and I’ll be travelling a lot in the next few weeks, but I’d be glad to check in and answer any dangling questions.

    I’d also like to get into the “seahorse smile” issue and post some images.

    And Jane, I’d love to discuss your paper privately with you at some point.

  24. tim jenison

    Here is Steadman’s miniature. Copyright Philip Steadman.

  25. W. Wick


    Your pictures are way better. : ) Seriously, what I like about your experiment is that you used natural light on a full scale set, made a plausible 17th century lens, and, possibly for the first time in history, produced an accurate full color painting by way of an optical method without using photograph as an interim step. I don’t know if I’ve seen even one CO tracing from Steadman. You are at the forefront of this idea – no doubt about it. What’s more, you “radically tested” this with cameras rolling. Genius and courage too!

    But of course we need to look at all the clues your experiment produced. I have attached two full color images: Vermeer’s from the NGA site; yours from the Vanity Fair piece. I have not manipulated them, except I bumped up the contrast of the NGA’s Vermeer very slightly for monitor viewing (otherwise it looks unnaturally flat in the computer environment). For this comparison, no more blinking lights, just the two side by side.

    First, look at Vermeer’s original with the knowledge that the tonal map of his painting is very different than yours, resulting in a more subdued, quieter setting. The eye easily finds the center of interest, where the two instruments positively glow. However, not only is the tonal map different, but the hues are different too. I think Louis was right to bring up Monet’s haystacks. Vermeer is using something akin to a “golden hour” palette adapted to a north-light interior setting. Not that he painted outdoors, but merely that he was using an optical trick to set the mood and give resonance to the instruments. Move your eye over to your version and the instruments blend into the background, as if the north-light color temperature has been too literally rendered. In the next two posts, I’ll zoom in for some other clues.

    Here is the source of the pictures:



  26. W. Wick

    Isolating the central area of both pictures, in yours, the eye still gets hung up in the periphery: the light on the wall, the bright reflection in the mirror, the bright yellow blouse, the higher contrast tile. Yes, there could be condition issues, and you could be forgiven for not being the best fashion stylist. I also appreciate that you “painted what you saw” and that was essential to the experiment. But still, I think you have to agree that Vermeer had total command of the tonal scheme. Once again, you can see the “golden hour” palette: the blue toned shadow on true wall behind the virginal in the corner sets off the golden hue of the instruments, the blue tone in the tile contributes to that too – as if reflecting the sky. As Jonathan points out, that blue tone drops out of the tiles toward the back wall. We can only speculate on that, but in the overall scene, you’ll see that a section of wall under the windows is warmer too, as if Vermeer had seen a back wash of warm light reflecting of the instruments. If so, it’s not utterly convincing, but maybe, for reasons we may never know, it was to him.

  27. W. Wick

    Cropping in around the viola da gamba (as I’ve mentioned earlier in the discussion) you can see how Vermeer’s instrument looks like it would make a beautiful sound. This is so essential to the success of this picture. The top looks curved and taught, the color has a mellow, gold-toned glow, much like what you associate with the acoustics of such an instrument. The contrast between the hues of the instrument and cool tint of shadows are what sets that off. Lighting is also key, as the top has a bright region that gives shape to the curvature, sweet highlights along the edge and tailpiece, and very a distinct shadow nearest the viewer on the floor. By contrast, your viola is flat looking – a circumstance of poor lighting in that area of the scene. It looks heavy, yet its ambiguous shadow does not properly anchor it, so it appears to float. This kind of problem often arises in studio photography. You get all the objects in the right place, but light doesn’t behave nicely in all the places it needs too. So what do we do? If we can’t judiciously add a spot of light to pump it up, we might move the instrument to another location that better shows off its features. But Vermeer can move the instrument anywhere in his studio and paint under a different light source, and no one would be the wiser. Indeed, the highlights on the instrument differ from yours; the shadow may be invented.

    Finally, the tiles in your version continue to distract the periphery, even in this close crop. In Vermeer’s the whites tiles are darker; the black tiles lighter. This gives him room in his dynamic range to add shadows, even on the black tiles, to anchor his objects. The viola is one, the shadow under the virginal is another. I would venture to guess this is invented too.

    Taking all of the above on the whole, it’s very possible that there’s a lot more invention in Vermeer’s color and tonal scheme than meets the eye. Once Vermeer starts the underpainting, the CO projection is more or less useless. His canvas becomes the roadmap. I’m not sure about the comparator. It has it’s uses, for sure. But I’m skeptical he would paint absolute values, only to cover much of them over for the above described effects.

    I don’t have the answers, but I do know your experiment has added hugely to this debate, even if we don’t agree it’s for the same reasons. It’s a great intellectual puzzle. I thank you for contribution to it, and all your thoughtful and kindly replies to our comments. I have to thank Jonathan too, for hosting this forum. I’ve learned something from everyone here. Maybe we should all meet up some day! Windsor Castle, anyone? This great puzzle has many more surprises in store, I’m sure.

  28. W. Wick

    The two posts above are part of a 3 part post to Tim, however, the first post appears to be held up in moderation. It should make more sense if part 1 appears. But if it all
    just sensless, that’s OK too! ; )

    Sorry for any confusion,


  29. Jon Boone

    I’m looking forward to what Tim does by way of tying up his loose ends, some of which must involve more clearly explaining how his informed perceptions about the linear quality of Vermeer’s optical representations manifest themselves with Tim’s optical system. Here’s Tim’s quote that is so provocatively compelling, at least to me:

    “Vermeer, if he used a comparator, is more accurate than digital or analog photography. Since he is directly converting light intensity and color to paint color, it is a highly linear system. It doesn’t look like modern photography. It’s better. It’s like a digital camera without the processing combined with a perfect ink jet printer.

    “Throughout the range from black paint to white paint, the image is highly accurate, or linear. We never see images that are linear. But again, our retinas don’t really care. It looks like a picture. But that linearity is what I noticed when I looked at “The Music Lesson”. I’ve worked with analog and digital images all my life, and I was the right guy in the right place at the right time to see it.”

    Of course, knowing more about the evidence that emerges unbidden via his use of the comparator, such as the virginal smile, is also important.

    Finally, I’m appreciative of W Wick’s latest remarks as they express his concerns about Vermeer’s artistic deployment of optical devices. I especially found his comments about the differences between the way Vermeer and Tim painted their respective viols very stimulating. And perceptive. As a great artist, Vermeer would have rendered such an instrument in a way that would have honored its reason for being, providing a visual image that compatibly resonated with the best sounds it could make (recall that I think The Music Lesson is in part Vermeer’s reification of music in visual form).

    However, I’m not as convinced by WW’s comments about shadow patterns throughout and beneath the viol. In the first case, I see few differences between those shadows in the two paintings of the instruments themselves. In the second case, the shadows beneath Vermeer’s viol are indeed more conducive to a credible image in space than the shadows under Tim’s viol.They look like “real shadows” that would realistically anchor an object in real space. They don’t look like something Vermeer “invented” to achieve more artful artifice. The lack of “realistic” shadows under Tim’s viol seems evidence of Tim’s lack of experience as a painter more than it does to support a thesis that Vermeer “invented” the shadow.

    As Tim points out, most of Vermeer’s paintings took place over many weeks or months. From the forensics, Vermeer added the viol well after the painting was underway. As the painting progressed, he could have altered the lighting to comport with his artistic intentions at virtually any step along the way, imitating the color values accurately with a comparator–values that were slightly different than they were when he began the painting, something that Tim couldn’t control for.

    Nonetheless, the painting’s roadmap would still, in the final analysis, be his projected image as established by his sensibilities, even if those images changed slightly over time. I don’t think Vermeer would have wastefully covered up accurate tones. Quite the contrary. Given years to hone best use of the optical devices Tim thinks Vermeer likely deployed, as do I, Vermeer probably had a clear sense of when to use his comparator efficiently to mimic the tones he sought to convey. As an artist, I think it also probable that Vermeer controlled for the amount of light he desired to make his point, then painted the resulting values with a high degree of accuracy.

    As I’ve said from the get go of this discussion, I believe it will be skilled painters who will use Tim’s optical equipment, hitched to an informed awareness of Vermeer’s painting techniques, stylistic preferences, and acute sense of design–to make the kind of optical/dimensional works worthy of the master’s hand. For these developments, I’ll stay tuned….

    Meanwhile, I hope Tim will use this experience to help prepare for the ardors of writing his book.

  30. W. Wick

    The first of my 3-part segment (November 4th, 2014 at 1:00pm) has cleared moderation.

  31. craig


    Thanks for posting the Steadman photo. That’s a big help.

    I’m also trying to review some of the loose ends. Way back on October 10th, you posted a quote from the Burlington Magazine and addressed it to me. Was this in response to my comment to Wick about Vermeer and the ladder or was it in response to something else?

  32. Jonathan Janson

    To all:

    Thanks for the words of appreciation. It is a pleasure to read your comments.

    I assure you that I have learned as much as anybody else, and for this thanks again.


    PS I am not trying to wind things down so don’t feel obliged to put on your coats and hats yet. Don’t hesitate to let me know any suggestions you might have to improve the discussion.

  33. Jonathan Janson


    Very interesting comments. Thanks.

    However, I think it best not to tamper too much with the image of the NG because that might just compound the uncertainties which already plague our observations.

    The NG has probably the most advanced digital imaging programs of artworks in the world and we have no reason to believe that their shot of the Music Lesson was realized with any other methods than those used form their own collection, which were produced with the most exacting criterion. Let’s stick with that, unless we know for sure they used a different procedure.

    I hope Tim can let us have his best image, as I do not know the quality of Vanity Fair one but it does not look good to me.

    It does not seem a good idea for Tim to do any reciprocal tweaking to make the color charts of the two paintings match. If he has to tweak something perhaps he ought to tweak his painting to match the NG image and not vice versa. If that is not possible, the comparisons we can make are of limited value.

    In any case, had WW not tweaked the NG image, his objections would have probably been more strengthened than weakened, accepting of course that the Vanity Fair images is reasonably accurate.

  34. W. Wick

    JJ, Very good point, but as you say, my tampering did not disfavor Tim. I assure you, however, that my move was very slight. I was very impressed that the NG color chart was dead on when I read the Lab values of the grayscale reference, but I thought Tim’s black point may have been pushed a little more, which is why I made the slight move on NG’s file. Tim may benefit further if he reshoots his with an X-Rite color checker, but it will not help the differences in hue, nor change the global tonal map differences very much. For the monochrome, I used the same basic “desaturate” method in Photoshop for both. While it’s hard to know what’s going on “under the hood” in Photoshop with respect to converting different colors of each painting to monochrome, I don’t think those anomalies would come close to explaining the strong, localized tonal difference first pointed out by Craig.

    I’ve been asking myself ever since, why did I not see such obvious differences, even as I’ve picked up on many other smaller ones from the very start? One reason is we don’t look at a picture all at once, even though we think we do. We scan it and assemble the information in our brain on the fly. I think this discussion illustrates that very clearly, as we are all learning from each other, many new observations about these pictures as we gradually drill down forensically over time.

    Another reason is we tend to accept Tim’s painting as “true”, because it IS true within itself, that is, we trust that all the lighting effects are internally conforming, just as we would assume light in a photograph would be (we are also trusting Tim, as I think we can). But when our eyes move over to Vermeer’s painting, this also seems true within itself, so nothing seems amiss. But from experience in the studio, I know how easily one can “cheat” light and shadow without setting off any alarm bells in an unsuspecting viewer. Or a suspecting viewer! I also think about the many times I made deliberate lighting changes, only to later be confused about which was which, until I reexamined both versions carefully all over again – and noticed that, say, that I had turned off a side light.

    Think about those “spot the difference” games that are so popular. Two identical pictures are presented side by side, with a dozen or so changes added to one of them. We can’t see them right away, because in essence the changes are camouflaged among the great majority of similarities. This is what’s happening here. The differences between Tim’s painting and Vermeer’s are camouflaged among the overwhelming similarities, many of which have been achieved through reverse engineering a painting, including the faux painting of walls. We tend to be impressed with certain conformities of light between the two, confirming that absolute values have therefore been achieved. We can spot differences here and there, but no alarm bells go off. But when you switch the differences on and off, as in my monochrome experiment, the alarm bells go off, and we see all the changes at once. The tiger has moved in the grass, betraying his camouflage. Tim’s Vermeer needs a lot more faux painting.


  35. Miss J

    Thank you Jonathan for requesting the entries be more to the core of the argument. I appreciate the extraneous quips, ‘cuts and jabs’ to Tim being discouraged.

    As to the differences to the highlights to the viola on Tim’s painting vs. Vermeers:
    If you look at the square PANES to the top window on the far left of the painting of which there are 6 from top to bottom, you see that Vermeer’s panes show top pane as white (cloud), next 3 are blue (clear sky) and the bottom 2 are white (cloud). Tim’s panes look more uniform where you don’t see distinction of white/cloud and blue sky patches. Although one can’t expect the blue sky and cloud patches as they appear in Vermeer’s window panes to be the same for Tim when he did his painting, could it be that Vermeer’s viola at the time and sky formation when he painted it, just happened to be in the right place to catch a selected highlight that way? Could Vermeer, at times, have seen the same ‘flatness’ Tim has to his viola BUT also noticed there were SOME days it caught the highlight shown on the painting and waited for one of those days to copy it or added it from memory?

  36. Jonathan Janson


    Thanks for the exhaustive reply.

    Good thinking, our eyes adapt not only to reality, but as your comparison shows to pictures as well, perhaps even more so. This had escaped us all, I believe.

    Although I may have overstated the case in respects to tweaking the NG image, I was more than anything else trying to get at one of the problems which keeps me itching: consistency. Do we really have the kind of comparative material with which we can formulate more than educated observations?

    It seems to me that we attempting to draw conclusions about a very hard thing using not only just apples-and-peaches type of comparisons, but apples (Tim’s photograph of his painting) peaches (Tim’s reconstruction of Vermeer’s room) oranges (the colors and tonal values Vermeer’s actual room which no longer exists) and watermelons (photographs of Vermeer’s painting) comparisons.

    Which of these comparisons are reliable and which are not? As we know, Tim used one photograph of the Music Lesson (the old one he got from my website) to reconstruct the colors of his room but another (the new on the NG website) to compare his picture with. Is this a reliable comparison whatever results it might yield? The only thing we do not have, but is indispensable as an element of comparison, is Vermeer’s real Music Lesson, which may no longer even be like it was originally painted.

    Yes, paintings are documents as Hockney claims, but they are not always reliable documents, not at all. They often tell us things that false.

    Please forgive me for insisting on a point I have already brought up other times but if we were conducting Tim’s experiments with Vermeer’s Procuress or Woman in Blue, any sort of scientific measuring of their chromatic and tonal values would have lead to wildly differing conclusions had they been taken before or after their recent restorations. It unimaginable what differences there would have been had they been taken when they were fresh off Vermeer’s easel. This I say not because Vermeer’s pictures were originally better or worse artistic statements, that’s not the point, but because we are essentially arguing about differences so slight in tone and color that they go up in smoke or match more closely with a flick or two of a Photoshop lever. With so many varialbes in our equation, can we truly understand which are of these ephimeral tones were captures by the comparator and which were not?

    I have serious doubts about the condition of the Music Lesson itself, which as it would seem, was painted with a full measure of ultramarine blue.

    Look what can happen. As you can see in the image below, the ultramarine blue doodles on the floor tiles under the virginal of Vermeer’s own Lady Seated at a Virginal got ultramarine sickness and are now actually lighter than the base gray tone of the tiles in the deep shadows. The figure-ground relationship has been essentially reversed!

    Since there seems to be lots of ultramarine in the Music Lesson (recent analysis has show it is even in the window leadings, the brown wooden beams and in the shadows between the gentleman’s fingers!) we should know if it is behaving correctly or not because it can change tonal value very significantly. The chair seems to almost to fluoresce and I cannot for the life of me see why the wall underneath the virginals is so bluish while it is not so in the expanse above: it is a different hue. I suspect there is something wrong with one of them. Varnish, restoration, retouching or ultramarine sickness? Who knows. If it were the upper section (which I suspect to be the case), would it ne enough to effect Tim’s original premise?

    Obviously, all these wobbling uncertainties are not Tim’s making and it seems to me that he has done everything in his power to minimize them. For this he deserves much credit. But they remain (obviously, in my opinion).

  37. Jonathan Janson


    For your information, I believe that Desmond-Shaw of the Queen’s Gallery told Tim that they suspect the side wall directly to the left of the spinet was overpainted. If this is so, it would confirm your puzzlement: “in the overall scene, you’ll see that a section of wall under the windows is warmer too, as if Vermeer had seen a back wash of warm light reflecting of the instruments. If so, it’s not utterly convincing…”

    I admire your ability to relate optics, aesthetics and narrative discourses.

    Unfortunately, the vast part of the art history community is utterly unaware of the problems that are being hashed around here. Paintings are still subject matter and historical context, with a bit of style tossed in to make things more interesting. Some would all do very well to pay attention to Tim’s discourse even if they may not believe Vermeer used the comparator or a camera obscura.

    As we say, a theory does not to have to be correct to be useful.

  38. Jonathan Janson

    Although being a painter I should have, I had not thought of WW’s idea about how the eye moves around Vermeer’s paintings respects to how it moves around in Tim’s. It would be interesting to make records of how our eye moves around each painting mapping the principal points of fixation and their sequences.

    I would imagine that the patterns of the two pictures would be rather different, and herein may lie one of the reasons why Vermeer’s painting is a masterpiece: it actively orchestrates of viewer’s attention. The optical device cannot tell us either where to look or why we should be looking because it is utterly passive. It tells us only how we look and so it may, as WW’s tonal comparison would suggest, come into conflict with the “where” and “why” which have been determined actively.

    The reason why the deployment an optical device, whatever it may have been, is so successful in Vermeer’s paintings is that it was used very selectively.

  39. W. Wick

    I appreciate Jonathan’s points about the discrepancies in the various digital images and condition issues of the original painting. I’ve been aware of this, but if some are already claiming Tim’s experiment a success, they must be doing so by using what they consider favorable comparisons to one or more of the many digital versions of the painting. That’s one reason I decided to go ahead with the side by side comparisons in the earlier posts, just to see if we can shake out any other clues. To illustrate Jonathan’s concerns, and with an eye for a level playing field, I’m posting 3 versions of Vermeer’s painting, along with Tim’s, all of which come from the web. The version at the top left, pictured without a frame, is the most super-saturated and contrasty. It’s from a New York Times article about the movie, credited to Queens Trust (there are many similar version of this one online). My guess is this image was generated from an older, film-based photograph of the painting. Next to that, is probably a derivation of that same file, from Essential Vermeer. This looks similar to what Tim was using as his target, judging by publicity stills from the film (Tim may want to clarify). Below left, is from the newer, high resolution version on the National Gallery website, with no further adjustments from their baseline setting. This may be after a recent cleaning, and may be a digital capture as well. To the right of that, Tim’s comparator-painted version from the Vanity Fair story.

    Many things could have happened to the top two renditions in the course of being digitized and shared. Because the bottom two have color charts attached, we can determine they have not been drastically altered. So yes, we are dealing with comparisons of peaches and pears, and maybe a couple rotten apples. As it turns out, it may not matter too much. What’s important is that the newly available NGA rendition, which on the surface is quite different from the top two, in fact shares the same DNA. That is, the lighting and color patterns of the brighter older versions can easily be derived from the newer, darker one. A simple increase of exposure and contrast and you can match either one very closely almost instantly. You cannot do the same with Tim’s because the lighting pattern and palette is all wrong. If Tim’s reference for his experiment was the Essential Vermeer rendition, it is nevertheless way off on color palette, even if you forgive the blouse and ignore the chair. He would be better off trying to match the new NGA rendition, where the instruments are less saturated – not to mention truer to the actual painting in question. But there, the bluish tone of “black” tiles and shadows is more evident. I believe Vermeer’s complimentary blue shadows and golden wood tones were intentional. I don’t believe Tim can match this with the comparator in a room lit with the color temperature of north night unless he has an unsually painted room and weirdly painted instruments.

    Vermeer’s palette could have been inspired from anywhere: a shaft of golden light coming through the window of his studio or somewhere else in the house. It could come from another painting or a walk outdoors. Jonathan, no doubt, could make a better painted document of Tim’s set-up, but there would still be no life in it if he referenced the comparator alone. Or maybe there would be, just another kind of life. Tim would be better off shelving the comparator for now and look into Vermeer purely from an aesthetic point of view for a while; from the notion of a storyteller in full command of his painterly craft. He can always bring back the device, but he would be a better investigator if broadened his understanding of how Vermeer differed aesthetically from his contemporaries. I can assure him, the wonders will never cease.

  40. Jonathan Janson


    Although it is possible that I have not understood your last comment correctly, it doesn’t seem that you have straightforwardly addressed the fact that, as WW comparisons would seem to demonstrate, in too many cases the tonal values in Tim’s and Vermeer’s paintings are hardly as close as they might appear. If one believes WW’s comparisons are in fact technically legitimate, they do not harbor well for Tim’s hypothesis.

    Why is light of the wall of the forward window well in Tim’s so much stronger than Vermeer’s, as WW pointed out? If the two rooms were structurally the same, illuminated in the same manner and both artist’s used a comparator, this does not seem possible. Can you justify such a significant difference in simple terms?

    It can also be noted that the tone of the most illuminated part of the nude back of Cimon in Tim’s right-hand picture-within-a-picture is far too bright in respects to that of the nearby wall. It seems to “glow” emanating its own light. In fact, it glows so much that its highest tonal value is not so distant from that of the fully lit side of the white ceramic jug, an absolute impossibility. Just in case, it is not a matter of hue, which can be excused since Tim worked from an old “burnt” image he found on my website a long time ago: it’s matter of tone, the crux of this long and very interesting discussion.

    On the other hand, look at Vermeer’s original. The same passage of the bound saint is decidedly lower in tone than that of the wall in every one of its parts, as it should be. This is because the picture-within-a- picture was not a pure white canvas (or even beige like Tim’s) but an ensemble of darker and lighter tones, all of which (except for a few brushstrokes of pure white paint) would have been darker than the tone of the wall.

    Naturally, we cannot know if Vermeer had indeed strove to maintain the exact tonal relationship between the wall and the picture-within-a-picture (if this were so the painting would have had to have been much darker than surviving replicas so my hunch is that there is some artistic license here) or if he exaggerated its overall darkness in order to make it very clear that the picture-within-a- picture in not a black-framed window looking through to a more luminous environment with floating saints in it.

    Frankly, I can’t see how such an anomaly could come through if Tim was using the comparator. I remember that even the smallest differences in tone between the image of the comparator and the corresponding passage of my e painting below were clearly evident, eerily so. The only technical justification would be that the local color of Tim’s wall was very very dark, which was not the case.

    I did not retouch the NG image or Tim’s from Vanity Fair. If we find Tim has a better image of his picture that is significantly different from the one used here, my observations may no longer apply.

    In any case, perhaps only when we look with sufficient care at our own attempts to emulate this master (obviously including each and every one of my own modest efforts) do we realize how complex yet how wonderfully simple is the art of Vermeer. In fact, the greatest satisfaction of having attempted to “paint my own Vermeer” for 40+ years was to discover how supremely talented this painter was. Art historians are often accused of overrating the technical sophistication and artistic stature of the painters they study. In my opinion the very opposite is true.

  41. Jonathan Janson


  42. Jonathan Janson


    Above you wrote that the dynamic range of your Vermeer setup was 7.6 f-stops. Is this measurement made with curtains? If so, did you make measurements of the dynamic range without curtains? I would be interested to see what the difference would be.

  43. tim jenison

    I’m back in Iowa tending to dad. Things aren’t going too well, unfortunately. Not sure when I’ll be back posting. Jonathan, that was with the curtains. I will take some readings without the curtains when I get back.

  44. W. Wick


    Sorry things aren’t going so well for your dad. It must be a difficult time for you and your family. There are times when priorities must be set, and this certainly is one of them. Our questions can wait.

    Take care,


  45. Jon Boone

    Hi, Jonathan. I’ve been traveling for the last week and chose to remain away from the Internet. But I want to respond to your comments posted to me on November 8.

    First, although I think Tim’s comparator/lens/concave mirror system sufficiently explains how Vermeer informed his visual style, I’ve maintained from the first that Vermeer’s artful designs and painterly skill were principal reasons for his overall mastery. Forensic analysis has long shown that Vermeer chose and generally applied his paints in a manner consistent with the techniques of his contemporaries. Because Tim did not apply his paint as Vermeer did, however, is not a reason to dismiss Tim’s general thesis. Quite the contrary.

    Like Tim, I’ve also maintained there are differences between Tim’s version of The Music Lesson and Vermeer’s, many intentional but some related to distinct variations of tone, a few of which seem curious. The most curious to me is the bright vivacity of the colors in the painting within the painting by Tim compared to the much more muted tones of that work in the original. You and WW noted a few other discordances. On the other hand, you’ve not mentioned the many extraordinary areas of concord, including the canny resemblance of subtle shadows, the image of the virginal, the clothing, the tapestry(!). Taken as a whole, the power of the resemblance between the two paintings should be, for any reasonable observer, overwhelming. Would that Van Meegeren had used a comparator. I join with Liedtke, as he challenged those who question the authorship of Girl with a Flute, when I ask that critics explain first the strongest areas of coincidence between Tim’s work and Vermeer’s before drifting over to focus on points of difference. The overall tonal match accuracy seems compelling.

    You know better than most how difficult it is to reproduce any painting with hi- fidelity even with today’s sophisticated reproduction methods, let alone paintings with nuanced color values. Until the advent of High Definition digital reproduction, most reproductions of Vermeer’s work did Vermeer’s craft little justice. Even some of these HD versions, as you have pointed out, miss the mark. As I look through the long history of bad Vermeer reproductions, I better understand why Van Meegeren was so successful.

    You also know that there are often striking differences between copies of original paintings made by the same artist–a common practice in Vermeer’s era and before. Ficherelli’s St. Praxedis is but one case in point.

    Finally, Philip Steadman realized first hand the way fugitive shadows affected Tim’s reverse engineered studio and his ability to recreate exactly the patterns extant in Vermeer’s work, for on some days double shadows under the virginal appeared–and on other days those shadows merged into one.

    One should keep in mind that Vermeer’s painting, like Tim’s, likely took months to complete, during which Vermeer made changes of both commission and omission–the positions of his models, the addition of the viol, the arrangement of and in the mirror. I’ve always assumed that Vermeer adjusted the room light and even the focus as any excellent cinematographer now routinely does for artistic effect–tweaking the work throughout until he was satisfied with the overall result. That meant there would be ineluctably resultant tonal inconsistencies within patches of of the painting that might seem discordant in their local visual neighborhoods but that, when seen in whole, enhanced the general harmonic effect. Nowhere is this idea more pronounced than in fugal musical scores.

    I think the way different way Vermeer and Tim painted the tapestry is instructive. Both are visually convincing (Tim’s version is not lifeless and mechanical, as some critics have alleged). That Vermeer artfully chose to elide many details, submerging them to enhance the overall visual effect (much in the way WW suggests for the rest of the painting) while Tim simply painted them in–distills the essence of my thesis. That Vermeer, as WW suggests, might have made similar adjustments of tone throughout the painting for artistic reasons, does not at all subvert that thesis. Again–quite the contrary.

    The eerie similarities of both Vermeer’s and Tim’s paintings to cinematic “stills,” which visual register does not appear in the paintings of other masters, remains nonpareil to me. I’m confident that Tim will eventually provide satisfactory explanations for some of the differences you and WW note; others may remain a mystery. But these will not, for me, be sufficient to toss Tim’s baby out with a few drops of bathwater. Really!? How can otherwise explain the figuration of the arch in the window of the Officer and Smiling Girl….

    In closing, let me once again urge you to use your painting skills and knowledge of Vermeer’s technique joined to Tim’s optical devices in your own reverse engineering experiment. Why not attempt a Girl with a Pearl Earring or even a Study of a Young Woman, if not a little street in Rome. I would love to see the result in an earnest attempt to honor Vermeer.

  46. Jon Boone


    I just read this piece, which sheds some light on Tim’s preliminary investigations.

  47. Jean Hémond

    Thank you Tim for your passion.

    I am a second career photographer and painter.
    I was was an Engineer Naval Architect before and painted from my youth as a hobby.

    I watched your documentary several times and got taken by it. Not so much about the controversy of mechanical or optical aid of painters.
    I for my part believe the innovation and result is what counts. I think “Art is a sublime lie”. And technology helps to achieve that.
    I can draw freehand . But nevertheless I very often use photos mine and others to assemble painted images. I usually change many things to suit the better composition respecting the light angles the perspective to dramatize the subject.
    I do that at times with photoshop and projectors.

    Your apparatus or Vermeer’s opens many new perspectives for me specially in now being able to compare tones.

    I am also aware there might be some tones variations with my studio lighting.
    I have in mind a photo of reflections in an autumn pond. It as a multitude of tones and variation of gold yellow greens and blues. That would be very difficult to render this image even for a very experienced painter. Of course there are short cuts.

    For me as I also have a minor problem with colours and tones” balfour’s brothers ” partial colour blindness it would of extreme difficulty but I think not impossible.
    I am presently assembling such a comparator to use with my computer screen and a canvas .
    But it might also be used with a Wacom tablet.I am also aware there might be some tones variations on the canvas with my studio lighting. will the screen light transpose to brighter vivid colours? I am in hope of that .

    I want to make a serious try at it overcoming my handicap.
    Thank again !

  48. Aaron Brown

    Hi everyone,

    I have a question about something that has confused me; please forgive a layman’s scanty knowledge of optics. Is the view of the scene through the lens/mirror assembly flattened? For instance, if you looked through the device at the white vase and the chair behind it, would there be a parallax motion between the two if you moved your head slightly? I don’t mean the view of the front mirror and the painting surface, but rather the view of the scene itself—Tim’s “Music Lesson” room.

    Again, I’m sorry if this question has been asked and answered already. If anyone knows the answer, please don’t hesitate to write.


  49. Firma List & Bedrog - BEELDHOUWER

    […] enkel aan te tonen dat zijn oplossing heel goed mogelijk zou zijn geweest in Vermeers tijd. (Lees hier een zeer uitgebreide discussie van Tim Jennison en critici over waarneming, optiek en veel meer, […]

  50. Anders Gudmundson

    I’ve used the simpler mirror setups or “Camera Lucidas” in a painting class last spring. Here is a video from Nyckelviksskolan in Stockholm:

Leave a Reply