“Apollo 13” IMAX

[Michael Coate]

Ron Howard’s “Apollo 13” is said to have been the first 35mm live-action/DMR conversion to IMAX back in 2002. It doesn’t appear to have played that many IMAX theaters, though, and so I’m curious who saw it (and where)?

[Brad Miller]

Wasn't that IMAX conversion a "cut down" where they tossed a number of scenes because the running time was too long for the platters to handle at the time?

[Marcel Birgelen]

Yes, the first few IMAX DMR releases had some scenes cut, some also converted the credits to slides.

The first feature-length IMAX release was Fantasia 2000, but it didn't use the "DMR method" as a transfer method.

Now, DMR is just a sticker IMAX puts on their blow-up technique. They claim they use the "hidden resolution between the grains" of a 35mm film... Whatever that may be. I guess, they use the hidden resolution between the pixels when the source is a Digital Intermediate...

[Kenneth Wuepper]

I think they are using the grain of film as if it were samples. Then it is possible to interpolate between grains with a very high sample quantity.

Analog to digital conversion involves sampling at some level.

[Marcel Birgelen]

The word "sample" doesn't really imply anything particular. Sampling is the process of digitizing an analog signal, it doesn't imply how this is done. I think the story of DMR with their information between the grains is just b.s. marketing. I doubt they even scan films at a resolution where you can see "in between the grains". And since most films are delivered as a digital master nowadays, the whole "in between grains" thing is nonsense anyway.

The problem with all kinds of upscaling is that you need to fill in the gaps with something. The most common way those days is by using some kind of scaling algorithm that uses some kind of smoothed interpolation. I've written some bicubic scaling routines in the past and the results of bicubic scaling are pretty stellar, but the end result will get pretty blurry if you're really pushing the scaling.

You can use edge-detecting algorithms like the well-known Canny edge-detector to identify edges in your picture. This information can then be used to sharpen the image around the edges. Nowadays, with modern graphic processors, we can do this even real-time. Some TV sets for example, come with SD to HD upscale algorithms that mostly work this way.

Then, there are some recent developments in AI, where neural networks are being used to interpolate upscaled images, which show quite promising results. Still, upscaling is making up information out of thin air and will never beat the real thing.

[Martin Brooks]

I saw it at the Loews Lincoln Square IMAX in a free screening as part of either a SMPTE meeting or an AES meeting. I think it was SMPTE.

[Frank Angel]

Hey Martin, how did I miss you?! Yes, I was there too -- it was SMPTE -- and I wasn't much impressed either with the cropsy-turvy presentation (hmmm, this doesn't look like scope to me!) and the dropped scenes or with the preshow, or post show talk, can't remember which, about the wonders of this new DMR process that was the new AI for film. And that's when the lies started. DMR could magically figure out what was image and what was film grain and how that was going to be able to take a film shot in 35mm look as good on the IMAX screen as a film shot in IMAX 70mm. Most of it sounded like BS to me. I don't remember much of that explanation of how DMR worked, mostly because IMAX was not about giving away any of their trade secrets, except for the story one of the IMAX guys told of that shot in the film where the astronaut picks up a hand full of moon sand and it falls thru his fingers. Turns out the DMR processor determined that the sand was film grain and totally eliminated it (laughter from the crowd; me, I am thinking, "Yah, well that's what you get.") They were quite sure this new process was going to save the company. Then came digital and LieMAX and the magic of IMAX 70mm was gone for good.

[Leo Enticknap]

I saw it at the National Museum of Film, Photography and Television in Bradford, England, and my memory of that screening is that the image was grainy to the point of being a serious distraction.