Red One, 5D Mk II, and GH1 Footage Comparison

Nektonic said:
I'm just not seeing the ghosting you talk about.
Click to expand...
I watched just once but if my memory serves, there's something when the girls 'rides' down the stairs. It can be used as a new aesthetics anyway. I'm not joking, it doesn't bother me so much, 'cause it is new and different, becomes interesting at times. Sure as art form, not so much or even at all for commercial gigs, I believe.
 
Emanuel said:
Kholi,

Have you been testing Shake or Twixtor?

E :-)
Click to expand...


I can't even figure out how to make Twixtor work, but I have done Pixel Motion with After Effects and I'm okay with that. Issue is that it takes freaking forever.

If I could figure out how to use Twixtor in Adobe Premiere I'd be good.
 
Okay, so I watched it again, and now that I was looking out for the ghosting on the 5D mkII stuff I noticed it in a few spots. There was only like one shot that it stood out on. I guess my worry would be that it would be amplified on a big 1080p HD set or if projected theatrically. Still though, it was not too bad.

The GH1 looked great up until the end. The back and forth shots between the two women at the crosswalk, with the Red One and the GH1, the GH1 just didn't look good. Granted this is before post color correction and grading, but still it looked kinda blah. The weird thing is, the cyan tinged stuff of the woman in the parking garage looked better on the GH1 to me. It had a more filmic and cinematic quality.

Kholi said:
I can't even figure out how to make Twixtor work, but I have done Pixel Motion with After Effects and I'm okay with that. Issue is that it takes freaking forever.

If I could figure out how to use Twixtor in Adobe Premiere I'd be good.
Click to expand...

Kholi, how is conversion from 30-24p in Compressor time wise? That would be a major issue for me if I had to spend a week or some ridiculous amount of time converting.
 
Last edited:
Nektonic said:
Okay, so I watched it again, and now that I was looking out for the ghosting on the 5D mkII stuff I noticed it in a few spots. There was only like one shot that it stood out on. I guess my worry would be that it would be amplified on a big 1080p HD set or if projected theatrically. Still though, it was not too bad.
Click to expand...

You probably wouldn't have to worry about this in a Horror or Drama piece. As long as it was all consistent or tamed.

The GH1 looked great up until the end. The back and forth shots between the two women at the crosswalk, with the Red One and the GH1, the GH1 just didn't look good. Granted this is before post color correction and grading, but still it looked kinda blah. The weird thing is, the cyan tinged stuff of the woman in the parking garage looked better on the GH1 to me. It had a more filmic and cinematic quality.
Click to expand...

The Nikon 300/2.8 was probably at like F8 judging by the image characteristics. Chances are, they may not have had ND filters for that setup, hence the smaller aperture. Not to mention it wasn't stabilized versus the 85mm and 50mm shots.

The 24p, however, looks exactly like RED's where the conversion and all of the MKii footage sticks out in this regard. Jello is another thing.
Kholi, how is conversion from 30-24p in Compressor time wise? That would be a major issue for me if I had to spend a week or some ridiculous amount of time converting.
Click to expand...

I can't remember how long it took but it's a fair amount of time for the best settings. And certainly much more time than just shooting 24p to begin with. LoL
 
Kholi said:
I can't even figure out how to make Twixtor work, but I have done Pixel Motion with After Effects and I'm okay with that. Issue is that it takes freaking forever.

If I could figure out how to use Twixtor in Adobe Premiere I'd be good.
Click to expand...
Not Shake? A few say it is better than Twixtor. Aren't you @ Mac side?
 
Emanuel said:
Not Shake? A few say it is better than Twixtor. Aren't you @ Mac side?
Click to expand...

It's probably the "best" method for it, however it's also the longest. I just have no time or patience to wait for a 24p conversion. Take that 12seconds of 720p in After Effects using Pixel Blend looked great to my eyes but took an Hour and a half or more to render.

Twixtor is faster, but with worse results.
 
Caesar said:
Sorry but now it's me who doesn't understand. You say 30p is an issue (ghosting, etc) and now the issue is the rendering after all?
Click to expand...


The footage in this test isn't Shake. It's Compressor. Shake's issue is rendering time. So the basic issue is that you're not shooting what you intend to exhibit at: 24p.
 
Kholi said:
Twixtor is faster, but with worse results.
Click to expand...
Worse than what? Is Twixtor worse than Shake? Or is Twixtor worse than After Effects? Or yet, is After Effects really worse than Shake? What have you been (real) testing?
 
I haven't used Shake for a conversion yet. I have had some footage run through After effects (pixel Motion) and Twixtor (someone else did the 30 to 24).

Shake is going to be a long render no matter what. I'm not up for that personally. After Effects > Pixel = 1hr 56minutes for a 12s 720/30p to 720/24p conversion.

Twixtor was like fifteen minutes but the example looked worse (frame stpeping) than After Effects > Pixel Blend.
 
Kholi said:
I haven't used Shake for a conversion yet. I have had some footage run through After effects (pixel Motion) and Twixtor (someone else did the 30 to 24).

Shake is going to be a long render no matter what. I'm not up for that personally. After Effects > Pixel = 1hr 56minutes for a 12s 720/30p to 720/24p conversion.

Twixtor was like fifteen minutes but the example looked worse (frame stpeping) than After Effects > Pixel Blend.
Click to expand...
I'm sorry, I was referring quality not rendering time. :)
How would you rate each method you've had experience with?
 
I have come to accept that I may not ever see 24p on my 5d. I have also accepted that if I want it the conversion takes forever.

Though Shake can take hours for a five minute clip. The conversion is 100% perfect with limited motion and 99% perfect for action stuff. I am willing to deal with this because I'm in love with the image the camera creates. At least - I'm in love with a calibrated image from the camera. After making my own settings I look back at the standard settings and wonder how anyone can deal with the yellow contrasty weirdness that comes stock with the camera.
 
I must say that i was surprised how bad red one can be in low light with this closeup shot with car.I like lumix on the roof.Dont quite like lumix on the streets,red looks quite nice there even if its so bright,it might be overexposed but it doesnt look like this bad overexposed video footage from video cams,it looks quite filmic.Real nice test,thanks.
 
Last edited:
Kholi said:
The Nikon 300/2.8 was probably at like F8 judging by the image characteristics. Chances are, they may not have had ND filters for that setup, hence the smaller aperture. Not to mention it wasn't stabilized versus the 85mm and 50mm shots.
Click to expand...


I just stumbled on this thread with the same title as the one in the other section of the DVXuser GH1 forum.

The 300mm Canon was actually at T16, with the shutter speed upped a bit either (1/60 or 1/90) too (to protect whites). I believe there was 1-stop of ND involved.

It was far from the ideal shooting situation- Sunset Blvd, during rush hour, guerrilla style, without walkies. There's definitely some on the fly decisions being made. If the team with the Lumix and the Red could have communicated we could have figured out a better way to match. I don't know why the images from the 5D weren't included on that particular shot.
 
You must log in or register to reply here.
Back
Top