Slick and Rumpled Images

It’s funny how sometimes bits of info you read from totally different sources, sometimes sync up in a profound way. For example, sometime in August I have read an excellent article by Errol Morris titled Photography as a Weapon. The piece is a very interesting analysis of digital photography manipulation being used for misinformation and propaganda as exemplified by the fake Iranian missile launch photography which was used by most news agencies around the world. Morris comments on the fact that while doctoring digital pictures is very easy and detecting such fraud is often very straightforward the practice still and very effective. Even though we are accustomed to seeing various ‘shopped all around us we tend to automatically trust the little grainy pictures that show up in the morning newspaper. Any photos allegedly taken by journalists are usually given more leeway and avoid close scrutiny. The grainier and low resolution the image, the more realistic it looks to the layman’s eye.

While it is trivial to run analysis on a high resolution image to detect manipulation artifacts left by cloning and smoothing tools, working with low resolution pictures proves problematic. Often concealing manipulation is as easy as saving your work in a lossy JPEG format with highest compression. Not only does the compression conceal the manipulation evidence, but it also makes the end result look more fuzzy/grainy and thus “more authentic”.

I’m currently reading The House of Leaves (btw, only halfway through, but a review is forthcoming) and I stumbled upon a very similar discussion. Mark Danielewski framed it as “Rumpled vs Slick images”. Slick images are the ones we know from marketing campaigns and Hollywood blockbusters. They are the high resolution, high definition and high budget manipulations that look so crystal clear and realistic we are convinced they are fake. Not only that, we are also desensitized to these types of images. The author uses an interesting anecdote to illustrate this point by talking about an award winning independent movie titled La Bella Nicoise et Le Beau Chien. The film depicts a disturbing torture sequence, followed by a murder of an innocent little girl. All in high definition, with a perfectly steady camera, masterful composition, beautifully executed zooms, pans and cuts. The disturbing picture charmed the Cannes audiences, and critics world wide. Until of course it was revealed that La Bella Nicoise was not really a horror masterpiece but a real snuff film. Hundreds of viewers were duped into watching a real murder without ever realizing it. All because of high production values. It can’t be a real snuff film if it looks like a Hollywood production, right?

If the same events were filmed with a cell phone camera, without any stabilization, and with the obligatory heavy breathing and footsteps of the operator serving as an audio backdrop, then released as a highly compressed Youtube style FLV video it would likely spark at least some controversy. Inquiring minds would demand to know more about the picture and start digging. The grittier and more amateurish the execution, the more people would wonder whether or not they were watching a clever fabrication or a real event. But the slick, polished Hollywood like production slipped past everyone’s radar. It’s authenticity was verified through research, but which was the side effect of it’s success. If it was not immediately successful most people would not bother digging. The illusion of authenticity lies is in the “rumples” and creases – in the inaccuracies, and imperfections of your picture.

Ironically, if you think back to Morris’ discussion the “rumpled” images that seem to emanate realism are the ones that are the easiest to forge. The “slick” images which we automatically categorize as fakes are in fact the ones you could throughly analyze and prove lack of any manipulation with high degree of accuracy. Thus as we become more and more adept at producing “slick” images, it is increasingly easier to mislead the audience by creating false “rumples” in your source material. Danielewski seems to wonder whether or not this trend will destroy the role of photographic and video record as objective evidence. After all, we rely on our gut instinct to detect falseness in a digital image. Lately however this instinct being ruthlessly exploited and subverted. Hollywood movie makers themselves are now using amateur techniques to rub off some of the slickness from their films. Movies such as the Blair Witch Project or Cloverfield try to generate fake “rumples” to help their audiences suspend their disbelief. Snuff film makers use high quality process to make real crimes look fake. The line between authentic and fake is being pushed, pulled and bent continuously – for entertainment, political propaganda or just viral marketing. We are being trained not to trust the digital images all around us. How long until we lose that gut instinct that makes us trust grainy and gritty more than the clean, clear and polished?

Perhaps one day digital photography and video will become inadmissible as evidence in a court of law? Perhaps the next Rodney King video be dismissed by the jury, suspecting digital manipulation? Perhaps it will be rejected as evidence when specialized analysis is not able to produce a 100% verification of authenticity due to the low quality of the sample. When there is no longer any objective evidence we can trust, how will that change our news reporting, justice system and our daily lives?

In case you were wondering, La Bella Nicoise et Le Beau Chien does not exist. I looked it up and no such movie has ever been shown at Cannes or any other big movie festival. Why did I look it up? Because Danielewski made it look like it could have been real. He gave the exact dates when it was first aired, discussed it’s origins and cited critical works that referenced it. Which brings me back to Morris again. Using “rumpled” style is not enough. After all, the shaky-cam technique makes Cloverfield no more realistic than the recent remake of King Kong filmed with traditional methods. The context, and the way you present your digital image is equally important. People believed the Iranian missile picture was real, because it was shown in mainstream media and seemed to be attributed to a semi-reliable source. Danielewski’s book is a work of fiction, which pretends to be a real document about another document which may or may not be real. The author aptly mixes liberal amount of rumples with scholarly looking quotations, bibliographical notes, and citations to create illusion of authenticity. When he goes on a tangent talking about an obscure movie that was shown only once, along with quotations and referrences to obscure publications you may never know if he is actually recalling a real anecdote or making stuff up.

This entry was posted in literature. Bookmark the permalink.



2 Responses to Slick and Rumpled Images

  1. jambarama UNITED STATES Mozilla Firefox Ubuntu Linux Terminalist says:

    You said that slick images are likely to be automatically discounted – like the fictional movie you referenced. This makes perfect sense to me, life isn’t slick. Like when I see a gorgeous model, I assume she’s had cosmetic surgery, she’s been airbrushed/photoshopped, or both. The women I see every day don’t look like that. When I see car ads that have the perfectly smooth and uniform glossy car hood, I know they’ve been photographed in unreal settings, photoshopped, or both – not even the new cars I see look that good. So I’d say this is a rational response.

    On the other hand, since rumpled images are harder to substantiate, it would make sense to me. It seems to me that being skeptical about rumpled images would make sense given edits are harder to discover, but then again they often look more like what we see or expect to see. If there is an image of a car exploding in the middle east, I assume the photographer didn’t have time to set up the camera on a tripod and do a nice long exposure. I expect jitter, blur, and shakiness.

    Anyhow, I just came here to say that I think our expectations are rational. Maybe that leaves us more vulnerable to deception, but at some point you have to take people’s word for it – you really can’t be totally skeptical of everything without rejecting all information presented that you didn’t see, which really paralyzes you on making decisions.

    Reply  |  Quote
  2. James Heaver UNITED KINGDOM Mozilla Firefox Windows says:

    There are two great examples of this – the first are HDR images.

    These ‘High Dynamic Range’ images are produced by combining three (or more) images of the same scene, one taken very dark, one neutral and one very light – combined cleverly and they can produce a hyper-real image of the scene.

    Now allot of the images you now see touted as HDR are coloured to look unrealistic and paranormal, but a true HDR image is just a /more accurate/ representation of the scene.

    I couldn’t find any online quickly, but the best examples are ariel photography (one professional photographer did a set over Turin IIRC) and they look like scale models. If you didn’t know better you would /swear/ that they weren’t real.

    The other example that sprung to mind was a band called The Dragons
    http://www.last.fm/music/The+Dragons/+wiki

    These were a bunch of session musicians (who became very well known) in the 60s/70s. When they were young they recorded an album which was never released.

    30 years later the master tapes were discovered by DJ Food of ninjatune. He remastered them and the album was finally released. Allot of people simply didn’t believe it was real, and were convinced it was a marketing hoax – the production was too slick, it didn’t sound liek a 30 year old recording, it didn’t sound like it had been recorded on late 60s kit.

    It was such a good quality recording, with such good production by DJ Food people actually did their won digging and investigating to convince themselves that it was real.

    Reply  |  Quote

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>