The present and future of post production business and technology | Philip Hodgetts



16:9 from 4:3 – you can’t get there from here!

I’ve been prompted this week to think about 16:9. 16:9 is great but the problem is that clients (oh, them!) buy or rent these lovely big plasma screens and suddenly we have to miraculously convert 4:3 source into 16:9 and make it look good. How do we communicate that you can’t get there from here! How do we communicate to clients that 16:9 display is not a decision made somewhere between commissioning the project and turning up at the trade show? How do we make it clear that, unless the decision to shoot 16:9 for 16:9 display is made at the start of the project, then everything from there on is compromised?

There are three ways to get 4:3 source to “fit” a 16:9 display. Two of them are quality compromises and the third is a compositional compromise. Some choice! First option is to blow the 4:3 source up 133% so the 4:3 fills the full width of the 16:9 but crop at the top and bottom of the image (oh, right, that’s both a quality and a compositional compromise). Choice two is to simply keep the 4:3 at full height and stretch the width out to fill the 16:9 space. That keeps the top and bottom composition correct, but compromises the quality (it’s about 120% width stretch) and makes everything look wide and fat. (Now that’s the way to keep a client – make them look fat! Maybe it is a way to make the point that this is not an appropriate way to get 4:3 to 16:9?). Still, there’s a good chance the client won’t even notice that the image is stretched. Seriously, every sports bar and restaurant takes this approach for their 16:9 displays and typically no-one notices. There is one difference here though… typically a 16:9 set will apply a non-linear stretch so the effect is more exaggerated at the edges than in the center. That’s not an option in NLE or compositing tools right now.

The final way to convert 4:3 to 16:9 is to consider that there’s a 16:9 canvas into which we place a 4:3 element and surround it with “something” relevant. Design elements or additional information. Place the 4:3 element in the center or off to one side to create a more balanced display. Useful for trade show type displays where the information can be useful.

But regardless of how well we work around the problem, ultimately it comes down to an unreasonable client request. How do we handle them? Same way as always… More than a small part of the job of a post production specialist is to educate clients and it seems it has been for a while now. Ever since the world stopped being “BetaSP=Professional.” That’s the problem with a diverse set of choices: it’s no longer a simple message and then suddenly it’s an “educational opportunity” with the specialist as the educator.

So, how do we deal with this educational role? Depends a lot on the client. If you have a long standing relationship with the client who’s happy with your work, then taking the “mmm, this isn’t such a great idea, here are the (unsatisfactory) alternatives” will probably work. If it’s a one-off or new client, then it’s more difficult. Then you have to feel out the client to find out what their level of discomfort is. If they’re budget focused (a very nice euphemism for cheap) then the stretched 4:3 into 16:9 solution is probably going to meet their needs. A client who cares about their public image will either realize that a 16:9 display isn’t the right solution, or allow the time and budget to find a creative solution for the extra real estate not used by the 4:3 image.

No tags


  • Josh · February 14, 2005 at 3:18 pm

    Our plasma has a “justified” option which keeps the center quarter of the picture unaffected then progressively (and logarithmically) stretches the picture so that the edges are very stretched.

    In shots where the subject is near the center of the screen and elements in the periphery of the frame are out of focus or otherwise non-descript, this approach works surprising well. When someone’s head nears the edge of the frame, it makes them look like they are approaching light speed.

  • Jason Stelzel · February 14, 2005 at 3:47 pm

    Has anyone done a blow-up of 4:3 (with top/bottom cropping) that was intended for theatrical delivery (i.e. converted to film) and how satisfied was the client? I’m faced with having to deal with this and I’m not sure how to proceed. I’m wondering if a hybrid approach might be best where I blow it up halfway on both axes and then stretch it horizontally for the remainder so I’m not cropping as much and not distorting as much either (given that so few people seem to notice horizontal stretch, reducing the need for it should be even better plus not having to crop so much makes recomposition easier).

  • Nick Meyers · February 14, 2005 at 3:56 pm

    adding to josh’s comment,
    my sony widescreet (crt) cals it “smart” mode.
    but as well as doing the edge stretch, it blows it up a fraction, too. (it’s the same blow as 14/9.)
    TV shows, with their “looser” or dare i say “blander” composition look pretty good with this setting.
    i’d be surprised if most TVs didnt have this mode, and cant figure out why the bar owners havent figured it out yet
    movies that have already been pan&scanned suffer, and one cant help but feel a sense of claustophibia watching them in this mode. i feel more cimfortable taking them back down to 4/3, and i can kind of “sense” the rest of the composition in the black pillar-box areas

    Graeme Nattress was developing a plugin for FCP that did the edge stretch thing, too, and i guess it’ll come out with his “Set3” when it’s ready.

  • Nick Meyers · February 14, 2005 at 4:01 pm

    i’d say that realistically, each shot will have it’s own requirements, and to get the best results, you’ll have to work on a case-by-case basis.
    this is what happens (or should happen) when film makers go the other way, and reduce their widescreen format for 4/3 television.

  • Jason Stelzel · February 14, 2005 at 5:38 pm

    Any word on an expected delivery date for “Set 3”?

  • George Dyke · February 21, 2005 at 9:02 pm

    Given the compromises in HDTV consumers displays and signal paths, I’m wondering if decent image enhancers can’t make YUV NTSC Betacam acceptable for the 16:9 HD? Years ago, with NTSC 3 tube Betacams, we took pains to crank down detail and filter lenses to get a “film look”. HD resolution seems to be the buzz, but in this marketplace and HDV compression, clean regular resolution video, enhanced by Farouja type devices, could be “acceptable”. 20 years ago we used to say that if folks just saw a 1st generation Betacam component signal (rather than some crappy zenith home TV) they would think they were watching HDTV. And for the most part we were right. I shot HD in the 80’s. Technically beautiful but grossly unappeciated then, and rapidly becoming indifferent now. As always, 35mm film qulaity is the best protectant. 70mm to be sure!



February 2005
    Mar »