General
Complete name : GOPR5496.MP4
Format : MPEG-4
Format profile : Base Media / Version 1
Codec ID : mp41 (mp41)
File size : 1.12 GiB
Duration : 5 min 18 s
Overall bit rate mode : Variable
Overall bit rate : 30.2 Mb/s
Encoded date : UTC 2019-01-13 10:12:35
Tagged date : UTC 2019-01-13 10:12:35
AMBA : x
Video
ID : 1
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.2
Format settings : CABAC / 1 Ref Frames
Format settings, CABAC : Yes
Format settings, RefFrames : 1 frame
Format settings, GOP : M=1, N=30
Codec ID : avc1
Codec ID/Info : Advanced Video Coding
Duration : 5 min 18 s
Bit rate mode : Variable
Bit rate : 30.0 Mb/s
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate mode : Constant
Frame rate : 59.940 (60000/1001) FPS
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Progressive
Bits/(Pixel*Frame) : 0.241
Stream size : 1.11 GiB (99%)
Title : GoPro AVC
Language : English
Encoded date : UTC 2019-01-13 10:12:35
Tagged date : UTC 2019-01-13 10:12:35
Color range : Full
Color primaries : BT.709
Transfer characteristics : BT.709
Matrix coefficients : BT.709
Codec configuration box : avcC
Audio
ID : 2
Format : AAC LC
Format/Info : Advanced Audio Codec Low Complexity
Codec ID : mp4a-40-2
Duration : 5 min 18 s
Bit rate mode : Constant
Bit rate : 128 kb/s
Channel(s) : 2 channels
Channel layout : L R
Sampling rate : 48.0 kHz
Frame rate : 46.875 FPS (1024 SPF)
Compression mode : Lossy
Stream size : 4.86 MiB (0%)
Title : GoPro AAC
Language : English
Encoded date : UTC 2019-01-13 10:12:35
Tagged date : UTC 2019-01-13 10:12:35
Other #1
ID : 3
Type : Time code
Format : QuickTime TC
Duration : 5 min 18 s
Time code of first frame : 10:22:21:49
Time code, striped : Yes
Title : GoPro TCD
Language : English
Encoded date : UTC 2019-01-13 10:12:35
Tagged date : UTC 2019-01-13 10:12:35
Bit rate mode : CBR
Other #2
Type : meta
Duration : 5 min 7 s
Bit rate mode : VBR
Other #3
Type : meta
mdhd_Duration : 318752
Bit rate mode : VBR
I’ve read so many people ask about Youtube. Now I’m far from being anything near an expert, maybe that’s why I took the easy way.
I use the default settings that come up. I change 1920x1080 to 1280x720, and I change the codex from 60% to 70%. The frame rate is 25, I shoot in 30, so I don’t mess with that. Basically I use this because I don’t know enough to try something else.
That’s all I do. I can’t tell you if I’m doing it right, but I have put over 300 videos on Youtube over the last 3 years, and no one has ever said anything to me about quality.
I’ve DL’ed the primary footage and I’ve watched it on YouTube. I’m no expert on settings - I just use the preset at 30fps and had no problem, but there’s much less movement in what I produce which I think might be contributing to the problem.
I’d appreciate someone else’s opinion on this observation, but to me it looks like the pixellation becomes more severe when there is a flash of light through the trees, i.e when a significant part of the picture becomes predominantly white for a few frames, it’s noticeably more pixellated for a few seconds after.
I’ve noticed similar artefacts appearing when the view goes from light to dark on other videos (even on broadcast TV) - where a dark space shows as a patch of heavily blocky blackness with minor details obscured until it corrects itself.
Do you have another piece of similar speedy footage filmed on a cloudy day that could be uploaded to compare to (assuming my observations are on point).
Sorry for the lack of response but since i’m new to the forum the system blocked me from posting more replies during 13 hours (such a nice feature)
@Chris319 I don’t think the video it’s just fine, comparing it to the one exported on shotcut, the quality loss it’s just big. Keep in mind that one is original and the other one is already encoded with much lower bitrates.
@QDSOV i already notice that yesterday, maybe it’s the auto low light setting on the camera that it’s causing it. You have more videos on my channel feel free to check. The last year ones are mostly all encoded with the settings that i posted above in the picture.
There is a lot of people complaining about the quality loss that youtube gets on the videos after upload.
Oh I’ve seen them complain and it does vary wildly depending on the kind of video you make. Vimeo is doing very well out of it! I’ll take a look at some other videos then
EDIT: I’ve looked at a few of them and I do think it’s something to do with light levels. You’ve been lucky to be blessed with great weather on most of your rides, but there does look to me like a relationship between sudden increases in white on screen and drops in quality.
The " SBV - Dolcevita" video (for example) looks like a more overcast day, but as you approach the trees and look up, the pale sky suddenly comes into view and the quality drops, then recovers a little while after. I see this a lot on YT even from people editing with Premiere Pro or any of the other “top end” editors. I don’t know how much is due to the edited output or the compression YT uses. I suspect the latter.
I don’t know enough to advise on codecs, GOPs and formatting etc. for the best, but I see a link and I’m glad I’m not alone in that.
I will try to switch off autolight to see how it gets. In some case is a fact that YT compression sucks
Auto Low Light allows you to shoot in low-light environments or when
quickly transitioning in and out of low-light conditions. When possible, the
camera automatically adjusts fps to achieve the optimal exposure and best
results.
Auto Low Light is available for frame rates higher than 30fps.
When Auto Low Light and Video Stabilization are both on and Auto Low
Light activates because of dark lighting conditions, Video Stabilization is
temporarily turned off to deliver optimal image quality in the dark scene.
In some case if someone could give me a better setup for youtube i apreciated
You’ve already been given Shotcut’s YouTube preset. Did you even try that? Did you post a link to it? Or do you think you know what you’re doing and have tried everything but?
Dude i already told you so, just read. “that default setting problably doesn’t give the best quality settings. Besides that the default quality/speed encoding is poor”. The link to youtube above was encoded with youtube preset…
I tried a lot of variations and basicaly didn’t see any improvements after upload to youtube… Or do you think that the sugestions you made were the first time that i tried them
There is a very good article on encoding videos for Youtube here: https://www.videomaker.com/article/c05/17034-encoding-youtube-videos-at-the-highest-quality
Apparently Youtube trancodes whatever video you upload, so the author of the article recommends uploading the highest quality video you can feasibly do, since compressing an already compressed video is likely to increase video compression artifacts (pixelation and blurring), which is what you appear to be experiencing.
So try increasing the quality of the H.264 encoding from 59% to a higher value, upload that to youtube and see how that fares. My guess is that 59% is probably a reasonable value for a video with a moderate amount of motion in it, but yours is highly dynamic and will have a lot of differences between frames, so probably needs a lower compression ratio (higher quality) to compensate for this.
Er, FYI, it was all shot at 60 fps, edited at 60 fps, rendered at 60 fps, and uploaded to YT at 60 fps. What YT did with it after that is anyone’s guess.
Why 60 fps? I don’t like smearing seen with 25/30 fps. Some people want smearing and “sense of motion”. It just doesn’t make my day.
AFAIK no frame rates were harmed in the process from raw to finished video.
Sorry for the confusion, I did not mean to suggest you were changing or increasing frame rate. Rather, frame rate is right next to the fields for resolution in Export, and there may be a temptation to change frame rate as well, which is what I was discouraging. Again, the most important thing to note from your post is to use 4K/UHD. Your experience and tests suggest to me there may something different about YouTube’s processing pipeline with that class of input.
I had the same issues and posted a similar question a few months ago. I’ve tried a multitude of settings, and frankly it seems to make almost no difference what settings are used in Shotcut. (I also tried Handbrake).
This is raw video file direct from the camera. (From about 10:20 on the first video is the same as above). To my eye, in Cinema or full screen there is not a whole lot of difference, so it’s not Shotcut munging the video quality.
I’m not real sure why this guy’s videos seem much sharper:
YouTube is pretty much a black box. Stuff goes in and comes out in some form that may or may not match what went in. I’m not surprised by the notion they’re trying to get the best content into the least space. Why transcoding from 1060@60 to 4K@60 and sending that to YouTube seems to give better results than sending up the same material as 1080@60 is… it’s a black box. By happy mostly coincidence, I have my video of a pass and the same trip shot by someone shooting 1080@30. Lighting is different as are the cameras. NTL it’s not too hard to see how Alexander’s original video compares with my video.
Original video by Alexander Thiessen:
My video, set to start at 0:50 to get past the titles:
(As a thanks for getting me into this videos and bikes thing, I tried to duplicate Alexander’s video. The open, and transition near the toll booth, start at the same spots, +/- a frame. We didn’t travel at the same speed, making a side by side comparison somewhat iffy. Past the top off the pass, forget it. I got stuck behind a bus. Once clear of the bus, though, it’s the same route and even the same last frame - +/-)
If you want to see how Youtube does its transcoding (in parallel, split over multiple servers, each one processing a small chunk of your uploaded video) have a look at the link below:
