Manual controls for 5D II video, hack initiative

mattsand said:
you already got the answer. notice how the number "5" you mention is significantly lower than "24"? that's it, simple as that. it's definitely a sensor limitation, as well as a processor and bandwidth limitation. they've cut a lot of corners in every step to make this work at all, sampling only some of the pixel, processing with fast and memory efficient algorithms and so on, same with the 5d, and while these issues will probably be fixed in new models they're *not* something a firmware hack can necessarily take care of.

/matt
Click to expand...


Yeah but at 4.5fps, with resolution 4288x2848 = 12212224 pixels (100 JPEG at a time)
At 24fps with resolution 1280x720 = 921600 pixels (5 min clip limit)

Even though 24fps is roughly 5 times the amount of frames than 4.5fps, each frame of the 5fps is 13 times bigger!

Not only that, but each frame of the 720p movie mode isn't as good quality as a 640x480 photo.

I think it's highly possible with some firmware to overcome this. Maybe even a 24fps per sec continuous shoot mode (thus giving you full control of shutter and iso) and then use after effects to sync the jpgs into a film clip! Or would that knacker out the shutter?
 
Did you miss the complexity lecture when you took computer science? Very few things are linear in data processing. But yeah, the system is built for live view and was probably never meant to be recorded so they probably cut a few corners too many. I'm just trying to give some perspective and explain why there's no point comparing the stills to the video. /matt
 
Did you miss the complexity lecture when you took computer science? Very few things are linear in data processing. The frame rate might be more complex than the number of pixels for whatever reason. But yeah, the system is built for live view and was probably never meant to be recorded so they probably cut a few corners too many. I'm just trying to give some perspective and explain why there's no point comparing the stills to the video. /matt
 
mattsand said:
Did you miss the complexity lecture when you took computer science? Very few things are linear in data processing. The frame rate might be more complex than the number of pixels for whatever reason. But yeah, the system is built for live view and was probably never meant to be recorded so they probably cut a few corners too many. I'm just trying to give some perspective and explain why there's no point comparing the stills to the video. /matt
Click to expand...


Haha you're such an angry and sardonic man at times. No I didn't take computer science course.

But as to my original point, I would pledge £100, for manual controls hack iniative on a d90.

Humph!!
 
BandanaDan said:
Haha you're such an angry and sardonic man at times. No I didn't take computer science course.
Click to expand...

I'm always glad to please. :-) it's not mandatory to take cs in order to shoot video or even speculate about firmware, but it sure helps when you're trying to understand why things are never the way you'd like them.
 
Seattle in the snow -- nice example of what's possible with the 5DII, very convincing low light footage with a cinematic look. Could this have been shot with with a "normal" small sensor camcorder?
 
TrueIndigo said:
Seattle in the snow -- nice example of what's possible with the 5DII, very convincing low light footage with a cinematic look. Could this have been shot with with a "normal" small sensor camcorder?
Click to expand...

Most stuff I ever watched and liked couldn't have been shot with a "normal" small sensor camcorder... :D

I still think there is something odd with the 5D regarding sharpness. It feels a little over sharpened for my taste, but not in every movie. This was nice indeed, but the london night shots of the thames river and the wheel were the most impressive, IMHO, so far...
 
taubkin said:
Most stuff I ever watched and liked couldn't have been shot with a "normal" small sensor camcorder... :D

I still think there is something odd with the 5D regarding sharpness. I
Click to expand...


I think it's more accentuated and weird in that, for some reason, everyone is using the 5D to do a shallow depth of field with extreme bokeh on people... on what would normally be wide deep focus long shots.

So it just looks weird like people are too sharp and greenscreened/projected flat background. Haha. Bokeh mad filmmakers?
 
BandanaDan said:
I think it's more accentuated and weird in that, for some reason, everyone is using the 5D to do a shallow depth of field with extreme bokeh on people... on what would normally be wide deep focus long shots.

So it just looks weird like people are too sharp and greenscreened/projected flat background. Haha. Bokeh mad filmmakers?
Click to expand...

That has happened before, certainly. Reminds me of 35mm adapters when they first came out... :D
 
You must log in or register to reply here.
Back
Top