Sunday, 19 March 2017

Riddle from the Past

André Kertész: Boat at Sunset on the Danube. Esztergom, 1917 (1967)
This photo by Hungarian photographer André Kertész taken exactly 100 years ago. Kertész became world-famous for his innovative and ground-breaking photographic techniques and the new perspectives he used in photo-journalism. The image above has created much discussion recently in the Hungarian halo-observers' forum, and generated some international correspondence. 

What can we actually see in this photograph? At the very first superficial glance this is like a solar halo with its reflection in the river. The light also becomes diffuse around it, so had the photo been modified, this would really be a masterpiece of manipulation from 1917. But Kertész was known for his documentary style not for remaking his photos.

However, if it was a halo, the image must have been taken by a lens with ultra wide focal length, which would not be commensurate with the boat in the foreground. The size of the feature could indicate a corona, but would a corona appear like this in a photograph made with a gelatin silver photographic process and the camera Kertész was using at this time: a Voigtländer Alpin 9×12 cm with a 6.3 lens?
All ideas are welcome. 

Nicolas Lefaudeux has given an explanation for the riddle in the comments section. The image (see below) he found of the same Kertész photo with but a different development confirms his point.

The photo appeared in in 2015.


  1. At first, I thought this might be a lens artefact. Many lenses of that era were uncoated so it is not inconceivable that it could produce lens flare similar to what is seen in the image. However, after a cursory glance, you notice the reflection in the water. This is definitely an aerial phenomenon.

    As Ágnes mentions, if an extremely wide angle lens was used, a 22 might be rendered smaller in the frame relative to the foreground. However, I don't think that is the case here. Back then, they were mainly standard lens sort of guys. Also, I'm not sure whether the Voigtländer Alpin he used was a fixed or interchangeable lens camera.

    Darkroom manipulation was very primitive in those days. To produce an effect like we see here would be difficult by burning and shading the print. The image we see here does not bear the hallmarks of such manipulation so I think we can reasonably discount that.

    With regard to the film stock used by Kertész, I wonder whether it was panchromatic or orthochromatic? Back then, I think panchromatic film was available but it was expensive and I'm not sure whether it was in universal use. If the film used was orthochromatic, it was less or completely insensitive to red. This increased sensitivity to the blue and green end of the spectrum meant that objects of those colours appeared lighter and those red, darker. If this photograph was taken on orthochromatic stock, could we then infer something about the phenomenon?

    Lastly, I notice that it appears darker immediately within the object. Could this be a second ring or area, dark because of its red colouration?

  2. I don't agree that the fact the reflection in the water means it is aerial phenomenon.
    It can be other effects that can create a ring of light around bright objects, hence around the sun AND its reflection on water.
    The darker inner area is not consistent with usual flare, that are on a general basis defocused images of the diaphragm. They are also aligned with the optical axis and the source, not around the source. So for me it is not usual flare.
    Atmospheric phenomenon would be something we don't know, no phenomemenon has this aspect, not even some kind of corona, they are always brighter around source, not a dark inner area. And montage would be too perfect, I agree.
    So for me, it is something else. I would say fog on an internal lens surface, or fungus on a lens surface are reasonnable suspects.
    Such stuff on a lens surface cause aureoles around bright sources, so around sun and its reflection, with darker inner area. They are usually a few degrees wide also.

    1. I suppose fogging or fungus on the lens could be responsible but would either of those be able to account for the fact that the reflection is a little broader and more diffuse than the ring around the sun and that it also appears slightly distorted by the ripples on the surface of the river?
      Also from what I understand, internal fogging or haze can take several if not many years to develop in a lens and would affect a more widespread area of the image. In the Kertész print, this does not seem to be the case. Plus the camera model he was using would appear to be relatively new at the time this photo was taken. If fungus or fogging were the culprits, I wonder whether any other of his images from this period taken with this camera exhibit the same problem?
      A couple of other thoughts occurred to me. Could the effect have been caused by grease or oil smeared on to the front of the lens eg a fingerprint? Likewise, could it be some weird aperture reflection off the thick base of the plate?
      Over the years, I have looked at a great number of vintage prints and I can honestly say that I have never seen anything quite like this. As to what this effect is, I have absolutely no idea.

  3. Just a minor point to add: one of the Hungarian members of our forum (György Répás) has called attention that Kertész had two cameras at this time. Earlier he had bought a portable ICA camera. or So he had an older camera model, too, and he might have used this one.

  4. Fogging or fungus would indeed give a broader and distorted reflection because the reflection of the sun itself (=the bright source) is broader and distorted.
    And it would also affect the whole image, but only be visible as a weak ring around any bright source.

  5. Grease or fingerprint would not create such perfect rings. You need to have more a network with more or less constant width to get a nice ring. The effect is diffraction by a grating. This is what happens in case of fogging when you have a "network" of mini droplets.
    Fogging would be temporary, so it would not be on other picture. And lastly, to see the effect on pictures, you need to have a superbright light source in the frame, which is not the case of many vintage pictures.

  6. Nicolas, the fogging or hazing that I have in mind is not the temporary sort you mention but when the lubricant or cement which holds the elements of a lens together evaporates over the course of many years and leaves a powdery deposit on the surface of the glass. I presume this would lead to an overall degradation of image quality especially contrast. Could this be the cause?

  7. possibly, powdery deposit could do the same. in the case of the picture, the loss of contrast would be pretty low, even barely noticeable or invisible in regular picture, as the effect is only visible around bright source with "contrast" of many order of magnitude (sun is overexposed to 10^4 at least on such pictures).

    here is an example of what fogging gives around bright sources: a few degree wide, perfectly circular, with dark inner area. the sharp edge of the case of the picture would be due to the spectral response of the system taking the image (film sensitivity + glass transmission that usually decreases abruptly at short wavelengths).

  8. Okay, you are persuading me Nicolas! I suppose one has to invoke Occam's Razor in difficult cases like this and not resort to "multiplying entities beyond necessity". The cases of fogging I have seen present exactly like the example you shared but these have affected the whole of the image eg if you breathe on the lens. The Kertész print certainly doesn't seem to be affected in such a way. Do you think fogging could produce the sharp ring (as you say, given the spectral response of the system) but without affecting the rest of the image adversely?

  9. How about testing it? Give diffraction grating corona image a treatment that imitates the properties of the old film.

  10. I wouldn't know where to begin, I'm not a technical person, but sure if someone else wants to that would be great.

  11. simulation should be doable. take the sensitivity spectrum of film, multiply by absorption of a heavy flint glass, multiply by solar spectrum, and then make a rotational symmetry of the whole stuff.

    meanwhile, i have also found this image with different developing:

    the ring in the reflection on the water is smooth and continuous. if it would have been a reflection on an actual phenomenon in the sky, it should have been discontinuous and broken in multiple parts, like the image of the sun, and like the reflection of the cloud.
    if it is from diffraction corona (or something else) inside the lens, created by the broken reflected image of the sun on the water, you indeed expect something continuous and smooth (convolution of the effect with the elongated and broken source of the sun reflected in water), exactly like it is here.

    so this is definitely no atmospheric phenomenon but something inside the lens

  12. Thank you Nicolas, you are right and I am now satisfied. This version looks like it is one of the original prints from the negative. The vast majority of fine art photographers return to individual images over their careers and produce different interpretations of them. On the site where you found this image, there is also another iteration from 1980 which looks different again. Had this 1917 version not come to light, I would still have had my doubts, but this one confirms your position beyond a shadow of a doubt. I think this has been a really instructive and profitable exercise. I wonder whether Ágnes could update the post to include this image just in case the link becomes lost or broken in future?

  13. got a nice case today of this effect in the high humidity environment of a bath room. see image here:

    not as sharp as the author case (which is due to the very different spectral response of digital camera vs film), but note how the effect only shows up around bright areas and leaves the rest of the image rather unaffected, whereas the whole outer lens was covered with condensation.

    also, in Kertesz case, the more i think of it, the more i believe the author indeed breathe on his own lens to experiment what this effect would look like in such a backlit scene...

    1. Thanks for the comment and the photo, Nicolas. I'm convinced.