p.3 #1 · potentially interesting 5D2/5D3 video news
skibum5 wrote:
Doesn't sharpen 60 or even 30 make it look all digital blocky looking and weird though? Which sharpen tool and what settings are you using though?
Not on the 5d3. I normally do about a 40 sharpen, and I use Premiere Pro CS 5.5. It handles the sharpen very nicely.
p.3 #2 · potentially interesting 5D2/5D3 video news
Shield wrote:
Not on the 5d3. I normally do about a 40 sharpen, and I use Premiere Pro CS 5.5. It handles the sharpen very nicely.
Hmm I find it goes weird if you go much above 36 and sometimes even at 24 it already looks a little odd. I usually do 18-36, usually around 22-26. You are talking the sharpen tool not USM tool right? USM tool I think can be used at higher settings. I don't think either tool works as well as the sharpen tools we have for stills sharpening though.
p.3 #3 · potentially interesting 5D2/5D3 video news
sqdstf wrote:
youtube link removed
Totally unrealistic expectations. We are never going to see raw video at 24fps out of the 5D3. This is the reason why a tiny bit of technical knowledge is a dangerous thing. Magic Lantern announces they can dump the live view buffer to 14-bit DNG. Next thing you know this guy comes along and starts a little internet frenzy for morons about how this means we will soon have raw video output from the the 5D2 and 5D3. Websites and forums all over pick this up and run with it. It's total BS. That video is one the biggest over-hyped, self-promotions I have ever seen. Somebody wanted their 15 minutes.
p.3 #4 · potentially interesting 5D2/5D3 video news
super35 wrote:
Totally unrealistic expectations. We are never going to see raw video at 24fps out of the 5D3. This is the reason why a tiny bit of technical knowledge is a dangerous thing. Magic Lantern announces they can dump the live view buffer to 14-bit DNG. Next thing you know this guy comes along and starts a little internet frenzy for morons about how this means we will soon have raw video output from the the 5D2 and 5D3. Websites and forums all over pick this up and run with it. It's total BS. That video is one the biggest over-hyped, self-promotions I have ever seen. Somebody wanted their 15 minutes. ...Show more →
True.
Yet, OTOH it does show that they are getting something really sharp off the sensor at 1920x1080 at 24fps with decent DR. So why do they then turn it to mush and hit the DR so badly before sending it to h.264 compression?? Are they doing horrible amounts of DR? Filtering it to make C-series look better? What?
p.3 #5 · potentially interesting 5D2/5D3 video news
Skibum, I hate to rub salt in the wound but have you seen the video made out of 5D3 sRaw files? " target="_blank" rel="nofollow">2877 Frames
This is what the 5D3 video would look like if it was raw and about 3k resolution. 2880x1920.
We should get together and make a video.
p.3 #6 · potentially interesting 5D2/5D3 video news
If you want to see how bad the 5D Mark III video is compared to the BMCC, check out this head-to-head comparison by One River Media. Be sure to download the footage, rather than just viewing it on Vimeo: the difference between the two cameras is even more stunning.
p.3 #7 · potentially interesting 5D2/5D3 video news
BluesWest wrote:
If you want to see how bad the 5D Mark III video is compared to the BMCC, check out this head-to-head comparison by One River Media. Be sure to download the footage, rather than just viewing it on Vimeo: the difference between the two cameras is even more stunning.
John
Stop complaining and do something... Don't just bitch about the 5D3... Go buy something made by Blackmagic for video. The 5D3 is a pro general purpose STILLs camera. The video on it is a bonus feature...
p.3 #8 · potentially interesting 5D2/5D3 video news
super35 wrote:
Skibum, I hate to rub salt in the wound but have you seen the video made out of 5D3 sRaw files? " target="_blank" rel="nofollow">2877 Frames
This is what the 5D3 video would look like if it was raw and about 3k resolution. 2880x1920.
We should get together and make a video.
Yeah I've seen that sort of movie. Some amazing natural world smooth time-lapse ones were posted up the other year.
p.3 #9 · potentially interesting 5D2/5D3 video news
skibum5 wrote:
Hmm I find it goes weird if you go much above 36 and sometimes even at 24 it already looks a little odd. I usually do 18-36, usually around 22-26. You are talking the sharpen tool not USM tool right? USM tool I think can be used at higher settings. I don't think either tool works as well as the sharpen tools we have for stills sharpening though.
Yes the sharpen tool. Now keep in mind I start with sharpness all the way to the left (0) and contrast -4 (all the way to the left). When you say a little odd what do you mean? What is your in-camera sharpness/contrast setting(s)?
p.3 #10 · potentially interesting 5D2/5D3 video news
BluesWest wrote:
If you want to see how bad the 5D Mark III video is compared to the BMCC, check out this head-to-head comparison by One River Media. Be sure to download the footage, rather than just viewing it on Vimeo: the difference between the two cameras is even more stunning.
I've watched that before - it's pretty much shot in a way that gives every advantage to the Blackmagic. Meaning it's shot at like F/8-F/11, so the shallow DOF of the 5d3 isn't shown. The 5d3 footage isn't sharpened either, and they gloss over the 2.4x crop factor for the most part. My 5d3 footage doesn't look like what's in that video - does yours?
Yes it has more resolution and dynamic range. Doesn't do stills and is a 4 pound "brick" that requires thousands more in disk storage and SSD drives to record to, and battery/rigging solutions. Not to mention you need to get really, really wide lenses for a normal view. I'll pass.
p.3 #11 · potentially interesting 5D2/5D3 video news
super35 wrote:
Stop complaining and do something... Don't just bitch about the 5D3... Go buy something made by Blackmagic for video. The 5D3 is a pro general purpose STILLs camera. The video on it is a bonus feature...
...and to add on to your point, the 5d3 takes great video as well. Still waiting to hear a better stills/video combination camera for the money.
p.3 #12 · potentially interesting 5D2/5D3 video news
Shield wrote:
...and to add on to your point, the 5d3 takes great video as well. Still waiting to hear a better stills/video combination camera for the money.
p.3 #13 · potentially interesting 5D2/5D3 video news
Shield wrote:
Yes the sharpen tool. Now keep in mind I start with sharpness all the way to the left (0) and contrast -4 (all the way to the left). When you say a little odd what do you mean? What is your in-camera sharpness/contrast setting(s)?
sharpness 0 in cam, contrast anywhere from -4 to -1 (sometimes if it is -4 it's hard to bring things back to get enough pop without grading issues though, OTOH at default contrast the shadows can really get clipped). I think I have it on neutral profile.
It's hard to describe but sort of it starts looking a bit painterly and digital blocky in a way if I bump the PP sharpen tool way up. So I was usually down more at low 20s rather than 40. Sometimes I try adding a little bit of USM at really small radius first to try to grab a hint of extra fine detail.
The processing tools in PP, even with extensions, seem weaker than in photoshop.
p.3 #14 · potentially interesting 5D2/5D3 video news
I haven't noticed that if I'm honest until about sharpness 40, and only with certain moire-prone objects in the scene.
What I do struggle with is having the contrast so low it's really hard to pull focus - I'm thinking of turning it back up a bit.
p.3 #15 · potentially interesting 5D2/5D3 video news
I wonder if they did change something about the video with the new firmware after all. Some fine details seem to dance with the tiniest movement, maybe hints of more AA or mosquito noise (I did some one very thing where a dark line popped in and out across a white letter almost like a line skip which was odd). Maybe they put out a bit crisper output now than before?
I need to shoot something and then revert to the old firmware and reshoot and compare. HDMI vs internal seems about the exact same crispness but I'm not quite sure how internal new vs internal old compares, maybe it is crisper (hopefully not just more aliased)?
p.3 #16 · potentially interesting 5D2/5D3 video news
I shot some stuff last night finally with the new firmware - you're right; it'd a tad crisper. I'm going to back off of 40 sharpness and regrade some footage.
p.3 #17 · potentially interesting 5D2/5D3 video news
Shield wrote:
I shot some stuff last night finally with the new firmware - you're right; it'd a tad crisper. I'm going to back off of 40 sharpness and regrade some footage.
I still didn't carefully compare, but now I'm doubting it again. I will have to carefully compare.
p.3 #18 · potentially interesting 5D2/5D3 video news
OK so that weird mosquito noise and shimmer on fine details is NOT in the files the 5D3 puts out either internally or recorded on a NINJA2. Premiere Pro compression or sharpning or something is making it get all weird. I figured h.264 high 5.1 at constant 40Mbps should be plenty, not sure what is going on but is NOT the 5D3 or the new firmware.
And it seems the new firmware does give 4:2:2 over HDMI, EOS HD guy was wrong.
EDIT: well other than in one file where there is some weird black band that pops on and off across a few letters of white text in a weird way on one file, that is on the original
FURTHER EDIT: hmm they do seem to be in the original files, just harder to see before sharpening
p.3 #19 · potentially interesting 5D2/5D3 video news
The out of camera video is slightly sharper now with the 1.2.1 firmware. Before a scene that I could sharpen @ + 50 with PPro now I see some moire @ + 35 sharpen.
Will need to shoot some more tests; I'm hoping that it's still moire-free if you leave in-camera sharpness @ 0. Might just have to only go to +20 sharpen now in post.
p.3 #20 · potentially interesting 5D2/5D3 video news
Shield wrote:
The out of camera video is slightly sharper now with the 1.2.1 firmware. Before a scene that I could sharpen @ + 50 with PPro now I see some moire @ + 35 sharpen.
Will need to shoot some more tests; I'm hoping that it's still moire-free if you leave in-camera sharpness @ 0. Might just have to only go to +20 sharpen now in post.
Are you noticing any weird sort of dancing aliasing on fine details if the camera gets juggled even a smidgeon?
Really weird is that one shot I took a black diagonal 1 pixel high line danced in and out going across some white text (oriented at the same angle) as the camera was barely nudged.
