(image screengrab and credit: Vimeo.com)

Advertising and Marketing Departments need to know good Stereoscopic 3D:

In a previous article we had addressed the subject of why an AD Agencie’s creative department needs to know a thing or two about stereoscopic 3D. After all they are responsible for approaching production houses and other freelance professionals (ad film makers and photographers) to produce stereoscopic 3D commercials or short films that directly or indirectly promote a brand.

Recently I came across another “branded” stereoscopic 3d video that is said to have been shot with Canon 5D cameras. Now this did ring alarm bells, because my first thought was about genlock sync. Even though it’s argued that shooting at 50fps or 60 fps (though at 720p) or even taking a snapshot to “sync” the cameras while in video mode would get the sensors to sync up… I still hold reservations on this.

(the side by side version can be seen here: http://vimeo.com/20030629 )

However, after I saw the video above, there were other anomalies that came to light, and in the pure interest of learning, I offer the below (hopefully Constructive) criticisms:

The numbers refer to the observations as they occur over time…

0:00:45 — Pseudo stereo.. watch the chrome railing warped (converted?)

0:00:55 — 3D Graphics at wrong depth – Pseudo stereo effect

0:01:14 — Large positive parallax on the fountains in background. Background semi-out of focus while foreground almost in silhouette. Where does the audience focus on? if foreground is in silhouette, normal tendancy would be for the audience to turn their attention to background elements (which are blurry and can cause headache, compounded with the large positive parallax)

0:01:39 — Non Tracked Graphics? Graphics go into wrong Z depth as camera dollys in.

0:01:49 — Excellent Stereoscopic 3D tracking on Graphic (180 degree Art Deco style)

0:02:13 —  How about using Hyper Stereo?: Where’s the depth? at the risk of choosing a Hyper Stereo base, it would be better than having no depth at all. Scene is rather flat.

0:02:17 — Definite Pseudo stereo, an ill-effect of 2D to 3D conversion (was this an automated conversion from 2D to 3D converter “boxes”) I ask due to the fake refraction through wine glass in foreground and also, is that a mirror reflection on the teacup? right image has a double teacup. Or is it an artifact of 2D to 3D conversion?

0:02:21 — Extemely huge neagtive Z parallax on the scene.

0:02:30 — Vertical misalignment of images (noticeable on man’s head and pillows, when viewed in anaglyph mode)

0:03:24– Effect of shooting non genlocked cameras? (pan at start shows side jitter on man). Also extremely large parallax on backgound (man)

0:03:38 – Pseudo stereoscopic 3D – Man’s face flattened with the sky. Buildings at wrong depth.

0:03:39 — Automated 2D to 3D conversion artifact?:  Double white shirt sleeve. (automated converter?)

0:03:41 — Inverted stereo on the scene. entire scene needs to be flipped (look at the gold pillars on the green building)

0:03:55 — 2D to 3D conversion artifacts (sofa has a “crease” where first bottom cushion ends, other image does not. When viewed in 3D an “extruded” effect appears. Secondly, mans hand has an artifact in right image.

Again… I emphasize that this quick critique is meant to be for educational purposes only, and in the hope that marketing departments, brand managers, AD agencies, and of course those professionals producing stereoscopic 3D content, keep best practices and the audiences well being in mind.

Watching pseudo stereo, (or reading text on scenes with conflicting depth cues) could or “may” for instance, cause headaches to audiences. This can only be detrimental to the brand or worse, give 3D a bad name in the long run.

  • Damiensteck

    Hi Clyde,
    I’m Damien STECK, the director of this video…
    I’ve no problem with the fact you use this video to show troubles and difficulties of stereoscopic work…
    But i can give you some important informations  to understand the context and the process based on this very very cheap approach of the 3D stereoscopic process (For the simple reason, the cost of a real 3D stereo process is not really compatible with the budget of Corporate movies today)

    So there is no conversion in this movie (2D to 3D) all is shot with 2 Canon 5D side by side, without rig… on the shoot there was no stereographer and no monitoring…  so you can imagine why i’ve choose to shot so wide.Most of 3D problems are fought in post production, with a very limited time… so some shots need to be distorted because the 2 lense wasn’t from the same brand…

    For the graphism troubles, it’s just a question of time… with more time i could be able to fix most of them… but budget and deadline exists…Finaly, most of troubles you observe are only in this poor anaglyph version… it’s really not the best way to judge a work made for an active diffusion…

    The result of the active diffusion is really good, and is made for only 4 minutes… i agree with the fact i won’t be supportable for a long movie… In conclusion, every film have their own story and reason… You say,  why Ad adgency need to know 3D…    In fact we know, and we just try to break a budget limit to made this movie, and i think it’s not so bad for the price…

    The trouble with the 3D, it’s too much technologique and limit the creativity or cost too much to be realistic…So thanks for your analyse, and i hope we’ll have the budget next time…Damien

    • Anonymous

      Hi Damien.
       Nice of you to reply. I appreciate your taking the time to clarify and provide some insights into the video.

      However in all fairness, as mentioned by you there was a lot of post-processing work involved.

      This is where most of the Pseudo stereo work seems to arise from. Take a look at the time codes mentioned for the errors listed.

      >>.Finaly, most of troubles you observe are only in this poor anaglyph version..

      This is simply not true unfortunately. Arguments such as this is mostly given only to clients or audiences who do not know much about stereoscopic 3D. The only “downside” of anaglyph is that it does not represent the full color spectrum. Depth and the “3D ness” of a scene is represented equally well in anaglyph as any other presentation tech such as passive/active polarized,or infitec etc.

      I just thought I should clear that point :-)

      Meanwhile, the article is for AD agency creative depts to know what makes good and bad 3D, so that hopefully they dont create adverts that harm audiences (and thereby again make 3D go down the drain yet again as in the past).

      Also on the bright side, maybe now AD agencies will sanction budgets for proper 3D shoots.
      The premise that 3D is much much more expensive does not hold true as it used to be just 2 years ago. Equipment, at least for Ad film making in 3D is today very affordable and compact. (Panasonic, Sony etc).

      Other professional equipment is also come down in rental and post production for 3D can be done on a Laptop. Some articles on this site describe these workflows and equipment.

      Wishing you well on your upcoming projects,

      Thanks and Best Regards
      Clyde

      • Damiensteck

        “This is simply not true unfortunately. Arguments such as this is mostly given to people who do not know much about stereoscopic 3D. The only “downside” of anaglyph is that it does not represent the full color spectrum. Depth and the “3D ness” of a scene is represented equally well in anaglyph as any other presentation tech such as passive/active polarized,or infitec etc.I can’t completly agree with you on this point, the result of the anaglyph is clearly badest than with the active glasses…  with the anaglyph you can see both picture in same time, it create some double images some time… It’s completly correct by the active process…
        I mean, yes the two pictures are the same for the active process and anaglyph, but it’s really different for people who are watching anaglyph or active process…And you probably know that the size of the watching change everything…It’s the same old question about the color settings of a picture… when everyone got a different TV and different contrast at home… so where is the big deal?In fact it’s always depend of wath is you goal, or your sensibility… and personally i really don’t care about technical perfection… the result of the active diffusion of “Eyes Wide Open” was clearly amazing… and too short to create any headache…Finaly i find quite curious this thechnical wondering, when you know that every people will see something different, specially with the 3d stereo, and 10% can’t see the 3d anyway…   So we can probably open our mind to different way to create 3D…It’s my opinion…

        • Anonymous

          Damien.
          Your missing the point. Your advocating the bad errors in the video to the presentation format of the video on vimeo (in anaglyph).

          The truth is, even if you supplied the original side by side or full left/right versions as well the errors would still show. I’m referring to the errors pointed out in the article in the time codes mentioned.

          These are errors in stereoscopic 3D production. We can go on this argument for days, but I would suggest you show the video to stereographers or even a production studio that is experienced in S3D and get their opinion.

          My intention is to critique stereo3D for the sake of the medium and for people who want to learn how to better produce s3D. It has nothing to do with technicalities versus art.

          A 2D video shot on a cell phone and projected on a cinema would just be a technically bad camera but could tell a great story.

          A 3D video shot on unsynced cameras, with pseudo stereo can cause PHYSICAL harm to audiences. so its a different comparison.

          If 4 wrongly done 3 minute commercials play before a 2 hour movie in a cinema that adds up to 12 minutes of headache before a movie starts… you see where i’m getting at with this…
          All the best,
          Clyde