120fps Vlog, Instructions in the Post
Added 2020-07-23 02:01:13 +0000 UTCBizarre experiment here, this is (in theory) a 120fps vlog that works within the limitations of YouTube to deliver some extremely rare ultra-high-framerate content and does so with a subject matter really capitalizing on the strengths of the format: a vlog sitting on the lawn rambling without script or purpose.
To watch the video "correctly" it will need to be played at 2x speed. If you have a high-frequency monitor this should (again, in theory) enable you to watch the video at an actual 120fps.
Production-wise the footage was recorded at 120fps off-speed with a timebase of 60fps. Audio was recorded in-camera and is attached, out-of-sync, to the raw video file. In post-production the footage was first re-interpreted into a 120fps timebase and dropped into a 120fps timeline for editing.
Here it's worth noting that Resolve interprets the video and audio framerates independently, thus changing the timebase of the video track re-syncs the audio and video. Despite the off-speed recording, the Pocket 4K is remarkably consistent, with minimal drift for shorter videos.
Once the HFR editing was done a 60fps timeline was created, the first timeline was nested inside, and then rate-stretched to 50% speed preparing the video and audio for the YouTube workaround. The camera footage was warp-corrected for the wide-angle lens, and a grade was applied.
This process did introduce a number of artefacts into the audio track that are apparent even when listening to the YouTube video at the rate-corrected 2x speed, as the audio was slowed to 50% speed, compressed, uploaded to YouTube, and then processed, effectively "baking in" the crackly nature of slowed audio.
Comments
Im quite sensative to frame rates and get nausea from games and films with lower framerate (the Hobbit was almost unwatchable for me). so for me watching on a 144hz monitor it felt much nicer to watch then when played at 60hz.
2021-09-09 06:02:25 +0000 UTCI barely noticed the difference... but it's so cool that there IS a difference!
2020-12-01 20:59:23 +0000 UTCI watched this on my High Frequency gaming monitor. I enjoyed your experiment! The pushback against high-frequency video is incredibly interesting to me. In my mind for consumers it's mostly against change, but I have to remind myself that on the creation side the hurdles are huge. I enjoyed the experiment, high frequency fascinates me!
2020-09-14 16:28:45 +0000 UTCI don't imagine I have the special tech you were talking about, but I can say that on completely vanilla setup your sound at 2X speed still had that sped up...I don't know how to describe it, that high-pitched electronic shadow that audio gets when played at slightly faster speeds but hasn't converted into full chipmunk.
2020-07-25 11:59:19 +0000 UTCThe sound on this one is really quiet for me :/
AndreGG
2020-07-23 23:18:38 +0000 UTC18 minutes of monetizable video with 9 minutes of filming? Now that's efficiency!
2020-07-23 14:24:06 +0000 UTCI have a 144Hz monitor, so I was interested to watch this, but sadly I can't tell a difference, even when watching the grass. I verified my monitor was working with https://www.testufo.com/ so I'm pretty sure the issue isn't with my computer. I'm just too much of a pleb to tell the difference outside of explicit side by side comparisons. I can see how super high framerate is useful in games and VR, but for standard video it doesn't seem worth the data cost. I still enjoyed listening to you discuss it though.
Mire
2020-07-23 14:22:29 +0000 UTCTragically my work monitor, for all its size and glory and like color accuracy, is only 60hz.
2020-07-23 13:44:37 +0000 UTCThis video is a real hoot! I like experimental videos (and generally any videos where artists and technicians share their process), though I feel like with the screens in my house, your video is like a filet mignon and my monitor is the Heinz ketchup that I experience it with. Nice work!
T. Delligatti
2020-07-23 12:06:42 +0000 UTCWatched this on a HFR Monitor running at 120Hz on Windows 10 in Firefox 78 and had stats for nerds open. Stats told me 65721 frames totel with only 4 dropped. Given a runtime of 1095 / 2 secs (18:15 shown runtime / 2 because of playback speed) I arrive at ~120.04 FPS. So unless some later part of the rendering system drops half the frames before This is committed to VRAM, it seems to check out. P.S.: I don't trust my eyes on this one, as my eyes are so accustomed to 30FPS video, I don't trust myself to judge whether I've actually watched 60 or 120 FPS.
Der Dackel
2020-07-23 08:59:06 +0000 UTCHDR processing, grading, and encoding is, put politely, a goddamn nightmare. There's a bunch of different standards, there's all kinds of hidden limitations to things like the total nit value of the screen in order to prevent the hardware from lighting on fire, and honestly the gains are really slim. So, yes, of course I am, eventually.
FoldableHuman
2020-07-23 04:14:08 +0000 UTC"I do kinda hope that at least one person who sees this has a monitor that is capable of actually showing it at 120 fps" - hey, that includes me! I knew my "gaming" monitor would be good for something. The motion in the background and your gesturing does look very smooth. Facial expressions at 120 fps seems to work well for this vlog style, but I am now fairly certain it would be completely distracting and immersion breaking for me in a movie. Oh, and are you thinking at all about going down the rabbit hole that is HDR recording?
Mike Carter
2020-07-23 03:26:37 +0000 UTCWatched this on an iPhone se. Though YouTube let’s you crank the speed to 2x I’m not sure if it was true 120fps or dropping frames.
2020-07-23 03:17:37 +0000 UTCI think recording the audio at a much higher bitrate using an external recorder would minimize the crackle effect by delivering the cleanest "slow motion" sound to YouTube, but the fact that this hack requires the audio to be sped up on viewing is always going to result in compromised quality.
FoldableHuman
2020-07-23 03:09:28 +0000 UTCI forgot I was watching this at 2x the speed and looked at the time passed like, "16 minutes? It's been 16 minutes already????" Legitimately thought my sense of time was screwed up
firelizard
2020-07-23 03:02:15 +0000 UTCIs there a process that would treat the audio a bit nicer? I've never heard anybody else mention this but I *swear* Chrome sounds better than Firefox running YouTube videos at 2x, and this sounds like a Firefox 2x with its robotic hum.
Joe L.
2020-07-23 02:45:09 +0000 UTCI tried to overclock my monitor to watch this, but I learned that it just goes black over 65 Hz. The audio is definitely super weird.
cal tal
2020-07-23 02:42:52 +0000 UTCI am not the most astute observer of framerates but I do have a 144hz monitor, running at 144hz according to my graphics card and Windows, and this video does seem to be actually playing back at 120fps. The wavy grass looked neat.
2020-07-23 02:30:00 +0000 UTCAlas, my monitor isn't fancy enough to see this in all its 120 fps glory. BUT watching the video at normal speed is fun because you seem very drunk.
2020-07-23 02:17:16 +0000 UTCYou mad man...
Larry Fried
2020-07-23 02:05:01 +0000 UTC
