Separate names with a comma.
Discussion in 'Digital Video' started by grooveattack, Nov 6, 2008.
Well? i was thinking about it and i dont know how it would be done. Can it?
Theoretically yes, but AFAIK the technology doesn't exist yet to be able to do it (at least not outside of lab).
Forgive my optimism but I think it can be done right now. But you'd have to do a bit of batch processing (with an app like Graphic Converter).
What you would do is extract frames from the clip you want processed, run them through a batch process in an image editor to bring up the shadows, then recombine them.
Of course you can do better than this by having the camera on a tripod and shooting two exposure values, extract frames from each clip, then combine the two sets of images via batch processing.
It would help if you captured RAW - you can do this from some consumer HD cameras via HDMI to a computer (you'd need a PCI card to do that though). Results won't be as good from compressed video.
But you'd have to shoot multiple exposure values at the same time at least 24 times a second for, what, 5min? 10min? What's a reasonable, minimum amount of time to be useful for video recording. Yeah you could find something perfectly still to record and bracket exposure manually (hopefully w/o moving the camera too much), but if you are shooting something perfectly still you might as well use a still camera.
When RED camera out w/their camera lots of people were kicking out ideas for motion picture HDR and nothing, given todays current tech, ever came close to panning out.
Ah of course. I wasn't thinking! But you could do some sweet time lapse that way - and it wouldn't matter if there was ghosting (which there will be). It would just be part of the effect.
The better direction to go is larger dynamic range on the sensors. That gives you all the detail you need in one shot, this improves photo or video images, but currently in still cameras anyway, they are pushing megapixels, not dynamic range. The highest dynamic range I have heard of is 11 stops in a line scan camera. I could be wrong but I think most DSLRS are in the range of 3 stops. The 11 stop camera produces something equivalent to a 3 shot HDR each time you open the shutter.
2 Cameras get put together through one lens to create a 3d image so why cant the same system be used with the camera aligned together to create a 2 shot HDR video.
Me thinks I could be on to something....
For shooting 3D two cameras are placed side by side (or a specialized single camera w/two lenses is used). They do not share a common lens because it's the space between the two lenses, just like the space between our eyes, that is used to create the sense of depth.
Theres the new one which has one of the cameras vertical looking down to a mirror in front of the horazontal cameras sensor. Works somewhat similar to the DSLR viewinder+sensor
Interesting. Do you have a link w/more info about it?
These are precision enough to align the two images perfectly; but there would still be potential issues with differing depth of field I would assume.
The required frame-by-frame process, regardless of it is part of a batch process, will take an insane amount of time for most purposes.
Unless you have some serious computing power or a great deal of time on your hands, HDR video with your typical video camera just doesn't offer enough benefit to outweigh the inconvenience.
Someone posted a video about a new (Microsoft?) technology maybe a year ago that would do it. You shoot your video, then take photographs of the set/environment. The technology then maps your photos to the elements in the video to enhance the quality and detail.
So you could just do a lot of HDR photography and then feed them into that, whenever it becomes mainstream.
But really though, HDR sometimes can be an easy way out for people who aren't good photographers (though not always, of course). Concentrating on lighting and optical filters is the way to go, in my opinion.
Thanks KeithPratt for posting the link. I didnt actually have one but I saw it featured on the Wired News site and I saw the real thing at the ski show in london. They had 2 EX3s mounted looked totally crazy. I spoke to the guys about it but I didn't really understand what they said.
Lag1090....I can see it being done either way for big films. I would assume it would be a faster process than making an animated film.
I've seen films and commercials with scenes where the dynamic range is definitely far higher than normal. Of course they're pouring a lot of resources into that though.
I think that would be pretty interesting, I have actually been working on a small hdr stop animation video.. for the last couple of hours, processing take sa loooong time tho.. I know htis is completely different to what you guys are talking about but woh well post it anyways haha
Im pretty sure ILM or Pixar would be willing to put the hours in
I just got my RED back after a 4-month rental. While I was in Brazil picking up my camera someone emailed me and reminded me of this crazy HDR idea I had a couple years ago.
Process in REDCINE and save 16bit TIFF image sequences from it three times. One for example -1(or-2)/stop underexposed, one 0/stop and one +1(or+2)/stop overexposed. In Photomatix (http://www.hdrsoft.com/) select Generate HDR selecting the three different TIFF images. Verify the bracketed exposures are as you adjusted and combine the images. Next in the Tool Palette select Tone Mapping and from there make any necessary edits...
I'm waiting for a nice sunny day so I can go shoot some contrasty scenes and test out this HDR video.
I thought of ideas for video HDR when I first heard of still image HDR.
Then I realized you would either need a still shot, to film 3 times over with 3 different exposures, or for the actors to re-act the same part EXACTLY the same 3 times over.
Then I forgot about it because it's impossibly (impractical) for now.
Not sure if it's HDR-video, but at the end of Gladiator, when Max is dying and his subconscious mind is floating, there's a scene where the clouds are moving rapidly and it looks like HDR to me. Not sure how they did it, but I'm sure it cost a lot of money.
I haven't seen Gladiator in a long time but it was probably just a sky replacement VFX shot. Gladiator used a ton of VFX and CG.