How is it trailers on Youtube are so sharp, but my videos are not

[Bencuri]

I have noticed something very odd regarding video quality on Youtube.

So, the laptop I cheked this thing on is: HP620 with Intel T3500 CPU.

My video:
https://www.youtube.com/watch?v=21mxVeRr3Iw

I have uploaded my video that was 24mbps in HD. The result: Even if i watch it in 360p on my laptop, there are occasional jumps, the playback is not continuous, and the sharpness is of course terrible, I have to manually select 720p playback, then the image becomes sharp.

Trailer:
https://www.youtube.com/watch?v=gtTfd6tISfw

Here is a trailer. This one plays continuosly on the same laptop, and even at 360p playback the image is sharp. Maybe not dead sharp, but as sharp as my video in 720p playback.

How is this possible? If I watch the original of the video that I uploaded, it has even more sharpness than the trailer. Yet, when I upload it to youtube, it becomes terrible. They use commercial paid accounts and Youtube gives them more bandwith or what's happening here?

[pandy]

I have noticed something very odd regarding video quality on Youtube.

Here is a trailer. This one plays continuosly on the same laptop, and even at 360p playback the image is sharp. Maybe not dead sharp, but as sharp as my video in 720p playback.

How is this possible? If I watch the original of the video that I uploaded, it has even more sharpness than the trailer. Yet, when I upload it to youtube, it becomes terrible. They use commercial paid accounts and Youtube gives them more bandwith or what's happening here?

YGWYPF

Commercial content is probably processed by YT in a different way than regular uploaders.
Your source is relatively LQ but still You can improve playback quality by upscaling video before uploading to Yt even to 4k - allocated bandwidth should be higher.
To sharp is no good as it may "steal" bits from video.
WebM (VP9) may give better results - check your player configuration.
Always verify video quality with good player (download uploaded file from YT locally to HDD).

[Baldrick]

No. The 360p trailer is not sharp. It looks like crap here. Nothing compared like your 720p video.

[KarMa]

YT does encode differently for highly viewed channels, compared to those who generate few views. Which makes sense, I just don't know where the threshold is.

[Bencuri]

Yeah, but something is just not fine about all this.

1. Maybe the trailer is crap in 360p, but it is definately better than my video in 360p. When, the original of my video is way much sharper. But okay, I will try to upconvert it to see.

2. But this is even more confusing: why is it that the trailer plays continuosly on my laptop? My laptop cannot play continuously any videos I record with my camcorder, though I have 4 different camcorders. And it cannot even play the 360p Youtube version continuously, that I uploaded. But, it plays the trailer continuously. How the hell this can be? The trailers are always continuous even at 720p, but my videos are only continuous, if I play the files from USB on HDTV or a Media Box, but not on the Laptop. I had an Intel i5 laptop before, that one couldn't play my videos continuously either, no matter it also had a dedicated nVidia GPU. This is confusing me too much.

[Cornucopia]

I was thinking, "are you serious?!", but I guess maybe you are, so I'll have to be serious and not too snarky:

That trailer was:
1. Shot with very high quality cameras & lenses -> Great optics & large sensors (incl. less self-noise & shallower DoF) = intrinsically sharp overall, but smoothed to leave only "focussed" area the sharpest
2. Shot with tripod/dolly/steadicam with VERY smooth pans, etc -> Predictable motion
3. Shot with very controlled lighting -> Predictable, controlled & constrained contrast
4. Shot at 24p FPS -> Less frames to encode = better compressability overall (though higher diff. per frame)
5. Recorded & edited & mastered on systems that maintained very high bitrates (probably > 400Mbps) -> no noticeable compression losses
6. Kept (at least until the upload to YT) at least at 2k (but probably 4k) -> higher density = sharper (and downsampling retains "perceivable" sharpness)
7. Trailer includes a number of smooth transitions to either black or credits -> simple FG/BG and predictable frame changes = less complex, less compressionally demanding
8. If that wasn't enough, I'm sure the trailer itself was pre-processed, noise-reduced to further improve compressability -> easier compressability
9. Pandy & KarMa are correct: commercial YT subscribers pay extra, but get correspondingly extra service, in the form of better compression quality and possibly even hand-tuning.

OTOH, your clip was:
1. Probably shot using something like a DSLR or semi-decent prosumer video camera. So/so optics & small/medium sensors -> not as sharp overall, and wasting sharpness on all areas of the frame
2. Shot handheld -> unpredictable motion = harder to compress
3. Shot in natural lighting -> Unpredictable, uncontrolled and widely-varying contrast = harder to compress
4. Shot at 30p/30i/60p FPS -> More frames to encode = harder to compress
5. Don't know intermediate bitrates, but mentioned penultimate bitrate of 24Mbps is mediocre for pro/semi-pro video capture, BAD for editing, and very limiting for uploading -> Low bitrate = compression artifacts & worse quality
6. Probably kept at FullHD (1080) whole time -> Good, but not great density, with no downsampling of top-level YT master (only smaller versions) = not as sharp
7. No simple FG/BG frames or predictable dissolve frames -> more complex = less compressible
8. Pre-processing? I'm going to guess not... -> less compressible
9. Free/lower traffic YT account? -> not as good compression applied

As you can see, there are quite a bit of differences, that all ultimately work back to compressibility or not.

YT does, AFAIK, still do a traditional (2pass?) VBR-style encoding (as they are trying to maintain control over their outgoing bitrates, which is what 2passVBR was meant to address). With a 2passVBR encode of a given bitrate tier (average & peak) using the same codec family, a file that is much more easily compressible will give a higher quality result, even while possibly appearing "sharper", than a file that is much less easily compressible.

Pretty straightforward science & math, really.

There is a reason why those shots look so good. Learn why, adjust your practices, and you can improve your compression quality also (not even considering "content topic").

Scott

[manono]

When, the original of my video is way much sharper.

The original you made and then uploaded? You know that YouTube reencodes everything, right?

You're comparing apples and oranges here.

Yours is 30fps and theirs 24fps.
Yours was shot using handheld and shaky cameras and theirs using more professional methods.
Yours was shot using video cameras (right?) and poorly deinterlaced. Theirs was shot using cameras that filmed progressive.
Yours fills the screen and theirs has big black bars which use almost no bits.

why is it that the trailer plays continuosly on my laptop? My laptop cannot play continuously any videos I record with my camcorder,

One thing I found very odd was that, depending on the resolution, your video's bitrate is at least twice that of the trailer. Maybe YouTube lowered bitrates more recently? Anyway, a higher bitrate could mean more buffering of the video for you. Is it the same when you download the trailer to your laptop and then play it?

[Bencuri]

I was thinking, "are you serious?!", but I guess maybe you are, so I'll have to be serious and not too snarky:

That trailer was:
1. Shot with very high quality cameras & lenses -> Great optics & large sensors (incl. less self-noise & shallower DoF) = intrinsically sharp overall, but smoothed to leave only "focussed" area the sharpest
2. Shot with tripod/steadicam with VERY smooth pans, etc -> Predictable motion

......

Scott

Excuse me to write this, but I think you overreacted this thing a bit. I understand your answer clearly, and there is absolutely nothing new in this for me. I uploaded a video that is clear in my subjective terms. It plays in bad quality. There is no surprise about that knowing Youtube reconverts it, it is just fine. I watch the trailer, that is better at the same resolution, knowing what you wrote about, it is not a surprise either. But it becomes a surprise when I experience during the years, that in case I I downconvert the clip here on my laptop to the same resolution as the Youtube version, it is just as sharp as the trailer. Moreover if I follow recommendations of Youtube regarding preferences for uploaded videos, no matter if I upload 8 mbps or 24 mbps video in 1080p, the outcome is the same bad quality. Actually if I play the 8 mbps Full HD video on the computer that I converted myself, it plays better than the Youtube version, no matter if the Youtube version was uploaded from 24 mbps. Is plays as good as the trailer on the computer. Everything gets messed up regarding sharpness when it is uploaded to Youtube, and no matter what bitrate I use. I have to confess though I haven't tried anything over 24 mbps. But even so, if I do the conversion myself, I get something similar to the trailer. So in this case all this is odd. So this is not about that I desire to go for Hollywood quality with a household handycam. I just don't understand why it is so that when I convert myself, the quality is decent, but when I upload anything to youtube, it becomes way much worse, but at the same time I can see some other videos are just as good as the conversion I did. In this case I become confused.

[budwzr]

You're not color grading. That's the problem. Movies are about fantasy. Your video looks like a live reality show shot on a FamilyCam.

Look into Color Grading. It's the hardest discipline in editing, at least for me. But you can buy "looks".

[Bencuri]

When, the original of my video is way much sharper.

The original you made and then uploaded? You know that YouTube reencodes everything, right?

You're comparing apples and oranges here.

Yours is 30fps and theirs 24fps.
Yours was shot using handheld and shaky cameras and theirs using more professional methods.
Yours was shot using video cameras (right?) and poorly deinterlaced. Theirs was shot using cameras that filmed progressive.
Yours fills the screen and theirs has big black bars which use almost no bits.

why is it that the trailer plays continuosly on my laptop? My laptop cannot play continuously any videos I record with my camcorder,

One thing I found very odd was that, depending on the resolution, your video's bitrate is at least twice that of the trailer. Maybe YouTube lowered bitrates more recently? Anyway, a higher bitrate could mean more buffering of the video for you. Is it the same when you download the trailer to your laptop and then play it?

Well, actually this recording of mine was shot in progressive as well, in Full HD 50p. But maybe it is not interesting now at all.

Okay, let us skip this sharpness now, I am fully aware of that my knowledge regarding filming and my technical background is far from Hollywood ones, I never even intended to go for that quality, and I accept than that in this case this is the sharpness that Youtube is capable to produce of my videos, that I can see.

But why is it then that the recordings have jumps? This is still something I don't understand. If I download the trailer in 720p, it plays continuosly as well on this old laptop I have. My video: even the 360p version doesn't play continuously when downloaded. On the other hand, when I watch my video on iPad4, it is continuous. This is quite strange again, knowing when I had the i5 laptop with nVidia GPU, my video didn't play continuously on that either, no matter if it was downconverted to SD. But again: when I playback the video on our old HDTV or my AC Ryan Media Box, it plays flawlessly. This seems just too chaotic to me. It looks to me as if it was about codec problems and not the recording itself. Isn't it possible that the video should be reconverted with a specific codec, and then my recent laptop could play it flawlessly as well, and Youtube could make a better video from it. Actually what disturbs me very much about Youtube is not the sharpness, because if someone clicks the 720p option, the image of my videos will become acceptable for me as well, but the jumps in the videos are quite embarassing.

[Bencuri]

You're not color grading. That's the problem. Movies are about fantasy. Your video looks like a live reality show on a FamilyCam.

Okay, but which ones??? This would be very good to know finally. (This referred to your codec comment).


Color grading: skip that. As I described, I am not complaining about the difference my videos from that camcorder have compared to Hollywood trailers. My problem is that when I convert something on my computer, the result is fine, but no matter what I feed to Youtube, it becomes worse in quality. This was the initial problem. I am not interested in at all how to get a film look or a film quality.

[budwzr]

Try using mainstream codecs like Mpeg2. But like I mention below, GiGo.

Whick ones? I looked at your sample in the first post. The whole thing needs color work.

It's gonna be tough to do because your dynamic ranges are off the charts. You have to match exposures, white balance, bad lighting, etc. A lot to it.

[Bencuri]

Whick ones? I looked at your sample in the first post. The whole thing needs color work.

It's gonna be tough to do because your dynamic ranges are off the charts. You have to match exposures, white balance, bad lighting, etc. A lot to it.

You wrote Youtube prefers certain formats. I replied: but which ones?

And don't worry about the colors. It is totally not interesting now, moreover myself I am more or less satisfied with the colors in this video. Actually, with the video, as I have it on the computer I have no problem. I am fully aware of that this is an amateur grade video, compared to professional ones several things should be corrected. That is okay. Skip that, it is not interesting now for me.

The only thing that interests me is why Youtube is doing these things regarding sharpness and why there are jumps on playback in the video.

[Bencuri]

Try using mainstream codecs like Mpeg2. But like I mention below, GiGo.

Yes, that is something I considered. But as I remember mpeg2 will produce a much bigger file, won't it? It would be better to solve it in the H.264 world of file sizes, if possible. Handling big file sizes for average consumer is a problematic thing. I often hear this complaint about handycams that the recordings have jumps, when MiniDV camcorders didn't have this problem. It would be useful to be able to tell these people an advice that they can use in the everyday life and that doesn't involve storing hundereds of GB of data for a few minutes of video. But if everything fails, I will do that of course. I just wonder if this thing could be solved at a file size that h.264 produces.

[budwzr]

Try x264. The open source version. Or WebM.

[Bencuri]

Try x264. The open source version. Or WebM.

Okay, I will try.

On the other hand, what about this iPad thing? I have an iPad4. That plays this video flawlessly. Also it plays videos flawlessly that others reported to jump. What is it that that device does differently than my i5 laptop? That could never play this video of mine flawlessly, there were always some jumps. Maybe if we could answer it, we could get close to a solution that way, too.

Aynway, when you are evaluating this jumping problem I am talking about, consider that the raw uncut videos from the camcorder doesn't have problems, at least when played back from the camcorder. The problems start occuring only after I edit and render the videos in editing softwares, or when the raw videos are played back on computers. I just mention this because I am not sure how cleary it reflected from my descriptions.

[Cornucopia]

You've got too many unknowns there that are conflating the issue.

To compare playback on various devices, use the EXACT same clip COPIED to RAM/FLASH/SDCARD/SSD memory (so HDD subsystem's slowness don't confuse) and played back natively (aka not via a browser), but not cached (so it's a simple block raw read). So, use an original pre-Youtube master clip. Then try multiple devices with a YT clip of the H'wood trailer (after DL'd into multiple target profiles), and then try multiple devices with a YT clip of YOUR clip (after DL'd into multiple target profiles). Also, as a reference, look at each DL'd clip profile with a bitrate viewer and with MediaInfo. But take the DL'ing out of the equation.

Compare only the devices and only WRT a single, specific clip. Then see how the different clips change that comparison.

THEN,
Compare the clips but keeping ONLY the same single device. Do a Before and an After/Round-trip (obviously, you can't do that completely for the H'wood clip, because you don't have access to the Before version).

Not sure what you'll find, but my guess is:
Since your clip exhibits wilder ranges of motion & contrast, it ought to exhibit wider/wilder swings in bitrate. Those swings may be mild enough that your device(s) and/or streaming DL speeds can handle it, yet not be able to handle the less mild swings in bitrate exhibited by your clip(s).

Encoding choice could be part of the story (preference for VP9 vs. H.264, or vice-versa, etc). But encoding choice is only a partial answer also, since you have Levels & Profiles and specific tweaks & tunings. Those L@P, tweaks & tunings might also conspire properly with the H'wood clip, yet somewhat conflict when involving your clip.

Bitrate: I was serious what I said earlier...you may consider 25Mbps reasonable, but the truth is, it doesn't hold up well with re-encoding (especially to the low levels that YT resorts to).

But, you ask: why does my re-encoding of my master look better than YT's re-encoding? First you must compare apples-to-apples and use the same encoding format (either H.264, or VP9, as YT DLs will show), but like I said above, it's still not the SAME comparison. You have to have equal Profiles & Levels and equal tunings and equal bitrate ranges.
But finally, where the could be some difference that may explain it on your end: even h.264 encodings are not all created equal. It depends on the encoder implementation. x264 is commonly used in freeware and is probably your encoder of choice, but it may or may not be YTs (they may be required to use a commercial, licensed encoder, either hard or soft). It is speculated that YT uses ffmpeg's implementation of x264, but this is only assumed based on reverse engineering clues. Even then, you don't know which implementation/build it is, and that can affect quality optimization also.
However, my hunch is that you're re-encoding using x264 libraries and YT's re-encoding is actually a VP9 for certain bitrates and h.264 for others, and that would easily explain the difference you're seeing.

It's all about using standard scientific troubleshooting methods: stepwise/componentwise iteration, cause-effect in-line replacement, deduction/process-of-elimination, apples-to-apples comparisons/comparisons to known quantities...

Scott

[Cornucopia]

<duplicate>

Scott

[_Al_]

Are you sure here about special treat for those videos with more viewing? Is it a sure thing?

That trailer has almost 3x less average bitrate comparing to that home video. Seems like no special treatment there.

[Bencuri]

Well, I will try to check these things again based on what you wrote.

But back to those jumps in the video: I had several camcorders from simply every brand through the years, and apart from having this problem with converted videos and with those that youtube converted from them, there was one camcorder that was exceptional in this regard: JVC-GZHM550. That produced videos that could played back even on my recent slow laptop almost flawlessy. I wondered what if I tried to copy its settings for conversion, and use that when I make videos, maybe it would give acceptable results. I will try to upload its videos to Youtube as well to see how they perform.

Here is the Mediainfo analysis of the mts files that this camcorder produces:

General
ID : 0 (0x0)
Complete name : C:\Users\Hallgato\Desktop\00045.MTS
Format : BDAV
Format/Info : Blu-ray Video
File size : 523 MiB
Duration : 3mn 2s
Overall bit rate mode : Variable
Overall bit rate : 24.0 Mbps
Maximum Overall bit rate : 24.0 Mbps

Video
ID : 4113 (0x1011)
Menu ID : 1 (0x1)
Format : AVC
Format/Info : Advanced Video Codec
Format profile : High@L4.0
Format settings, CABAC : No
Format settings, ReFrames : 2 frames
Format settings, GOP : M=2, N=14
Codec ID : 27
Duration : 3mn 2s
Bit rate mode : Variable
Bit rate : 22.6 Mbps
Width : 1 920 pixels
Height : 1 080 pixels
Display aspect ratio : 16:9
Frame rate : 29.970 fps
Color space : YUV
Chroma subsampling : 4:2:0
Bit depth : 8 bits
Scan type : Interlaced
Scan order : Top Field First
Bits/(Pixel*Frame) : 0.364
Stream size : 496 MiB (95%)

Audio
ID : 4352 (0x1100)
Menu ID : 1 (0x1)
Format : AC-3
Format/Info : Audio Coding 3
Mode extension : CM (complete main)
Format settings, Endianness : Big
Codec ID : 129
Duration : 3mn 2s
Bit rate mode : Constant
Bit rate : 256 Kbps
Channel(s) : 2 channels
Channel positions : Front: L R
Sampling rate : 48.0 KHz
Bit depth : 16 bits
Compression mode : Lossy
Delay relative to video : -34ms
Stream size : 5.57 MiB (1%)

Text
ID : 4608 (0x1200)
Menu ID : 1 (0x1)
Format : PGS
Codec ID : 144
Duration : 3mn 2s
Delay relative to video : -34ms



If you check this list, what are the essential points in it that might play a major role in that these videos play back better on my laptop than videos from other camcorders like Canon HF M41 or Panasonic HDC-SD80 and Samsung HMX S10? My other camcorders are similar in bitrate. So that cannot be the answer, and also in fps. Most of them are 50i, this JVC was also set for 60i, I don't really understand why it was detected 30p. I have a camcorder that records in 50p, but I ignore that now, I am sure flawless playback of 50p is far way beyond most of the playback devices I have here at home, even if converted.

[Bencuri]

A follow up: I uploaded a video of that JVC Gz-HM550 here:
https://www.youtube.com/watch?v=XJYnCUyQkfs&feature=youtu.be

The result: The video seems dead sharp even on 360p playback (okay, not that sharp, but way much better even as the trailer). But the playback is awful, it jumps frequently all through the video. Acutally I am quite surprised now, because this is the first time I managed to get this result on Youtube regarding sharpness since 8 years. I regret now I sold this camcorder... :/

When watched on my laptop, the playback is way much better. It almost lack jumps, maybe there are some minor ones, like 1-2 durint the whole video, but even those are not as drastical as those you can see in the video that I linked in the first post.

So all in all, regarding to my subjective expectations, 24mbps is definately enough to get a sharpness that those "top class" Hollywood trailers have. So I think we can ignore now what you insisted on Cornucopia. You see here is the live example that that kind of sharpness is not a matter of whether you use a handycam or a million dollar camcorder. It is the crazyness of Youtube.

[Cornucopia]

You're deceiving yourself.

Sharpness may be better than your previous personal clip, but motion jerkiness is just as bad and contrast is just as wild. And by your own admission the source is ~25Mbps, so you will certainly have some compression artifacts (certainly on the re-encode).

NO COMPARISON with the H'wood trailer clip.

Scott

[Bencuri]

Yeah, that is true these clips will never be as good as those Hollywood clips in all aspect. But you misunderstood something here seriously.

This jumping thing and sharpness issue with Youtube uploads is something that has causing me lots of headaches for years, and recently others started to complain about it. And to be annoyed is totally understandable, because I think it is weird that 20 years after the VHS format, it is still an issue to get a continuous video on the average consumer level.

So did you get the point now? I think most of the people who post here are aware of that Hollywood technique is way beyond the capabilities of most of us, who are on consumer level. Yet, with all the advertisements around, and remembering what VHS was capable of, or the MiniDV, it is odd that we have to complain about not being able to get continuous video, even from semi-pro labelled camcorders. This is just something ridiculous.

And I didn't post that trailer for arguing about why my cheap camcorder doesn't produce million dollar film quality. I wanted to know more about this sharpness issue, why Youtube is behaving so crazily, and that trailer was the easiest to find as a counter example. But it is not only trailers that look better than my video. There are dozens of other videos shot by cheap camcorders that look better. I just didn't note their names, but you can find a trailer easily, that always have good quality. So I used that as an example, because it was the easiest to find.

After what you replied, one might have thought just to get a decent sharpness for the video at youtube, you have to forget about handycams. Well, we can see now it is not so.

Here, the thing is not about that dumbs try to behave like Hollywood filmmakers, we are just trying to improve on our video in the possible limits. That's all. And you see it is not impossible, this example shows it. Also the iPad. You are repeating here that low quality video this way, Hollywood quality that way, but the thing is that my shit video is played flawlessly by the ipad. Interestingly, that is not a problem for you.

So I wonder when you realize that here I am not going for proving you that I am just as pro with my handycam as Hollywood filmmakers, I just want to know why is it that usually my videos looks bad and jump, when some of them from similar level camcorders with the same bitrate are clear and don't jump.

That's all. But if this trailer thing confuses you too much, I can search for a general video with similar sharpness, can use that as a counter example,and thus we can clarify this.

Maybe then you will understand what I am interested in is how to achieve with these equipments and knowledge to have a better result on youtube.

Now after you know this, maybe I can get answers. I accept even if you reply that it is not possible to get any better result. But I wouldn't mind if we could move away from this Hollywood comparison now. You see, that trailer was just an example for a general problem, not my final aim to achieve. And also I mentioned though my videos are just 24 mbps, when I convert them, they look as satisfactory as those Youtube trailers, for me. Not the same, but they pass the acceptable level. What Youtube produces doesn't. And I would like to know why Youtube is doing such a bad job, and whether I can have it do a better one.

[pandy]

I would advise you to watch this movie https://vimeo.com/49875510 - albeit not directly related to your question it shows how different video can be when significantly better quality equipment is used (DSLR vs Video Camera) - after this you should be at least able to name problems in your footage - use this as reference and don't expect miracles.

[_Al_]

Here, the thing is not about that dumbs try to behave like Hollywood filmmakers, we are just trying to improve on our video in the possible limits. That's all. And you see it is not impossible, this example shows it. Also the iPad. You are repeating here that low quality video this way, Hollywood quality that way, but the thing is that my shit video is played flawlessly by the ipad.

You mean downscaled 1280x720 with bitrate 1000 kbps ? That is basically bitrate of that trailer. Downscale your video to that resolution and encode it 2pass with average 1000kbps. Then compare the quality to that YouTube trailer.

That would tell you that your video is impossible to encode in any decent way with low bitrate. And YouTube gives you low bitrate alwyas, for any resolution. Youtube even distributed 3x more bitrate to your video as oppose to that trailer, still not enough. You cannot even remotely compare that trailer to your video.

Fluidity of video playback, ..., you cannot compare some YouTube flash or HTML5 browser playback fluidity to computer rendering capability. Also check 50i or 50p to 25p problem. Go into shutter speeds theories, what the shutter speed is. You cannot expect objects to be fluid going that way. You'd need to shoot it directly 25p to have better smoothness in video, comparing it to 50p to 25p conversion, or 50i to 25p conversion, actually this last one would be smoother than conversion from 50p. Crappy video, frames blended, showing ghost artifacts, depending on deinterlace method, etc. but smoother.

[poisondeathray]

For the playback issue - there are so many variables, such as gpu vs cpu, flash version, up and down stream bandwidth, os and browser version (huge differences here, it also partially determines if you get a flash h264/aac version vs. a VP9 version - you or other people might not actually be "seeing" the same thing)

Start by right clicking whne the video is playing and check "stats for nerds" . This will tell you how many frames are dropped

[Bencuri]

Well, personally I don't think that in 2015 expecting playback without jumps from clips recorded by Full HD handycams would be expecting miracles, considering in the world of VHS and MiniDV, it was possible.

The things that are analized in the video you pasted, they are actually not crucial things. If you are just a beginner in making videos, like when you buy a handycam for shooting family occasions, all these things in this video becomes actually marginal. if the colors are not that good, or the white balance, etc., you can still live together with that. But continuous playback is a crucial point. Bad colors don't neccessarily kill the video at the low end, but issues with jumps does. And so does sharpness, when the camcorder is labelled FullHD. I understand that handycams lack the sharpness of pro camcorders, but the raw video that my ones produce are fine. They are satisfying me. The problem is when I start editing them, and feed the exported videos to youtube, or even the raw ones, the outcome is terrible. This is again something that is not acceptable even at the low end. You cannot expect from a general user to have the knowledge to convert like a pro. But basic editing is defiantely neccessary at the low end, too. And if the user experiences that from the point the video goes through an editor, there will be sharpness or playback issues, that is unacceptable. Because then it is simply not HD equipment that he purchased. Regrding the start it might be, but considering the end, it is not.

And on the other hand, in spite of that you mention these pro and low end differences all the time, the case is that even in sharpness and playback, you can find positive exceptions. Like I mentioned, I have several camcorders, none of them produces videos that my computer can play flawlessly. Then I manage to find a camcorder at the same level with same bitrate, that produces a video that even the old laptop plays near flawlessly. After this I don't understand why insist on more on comparing to pro levels. It is crystal clear that in this level where my hardver is, it is possible to get a decent outcome. I would be more interested in getting an answer to how it is possible that in spite of that the videos from the JVC are same in bitrate and resolution as the videos from my other camcorders, the laptop can handle it. If I knew it, maybe I wouldn't have problem with future conversions. And also, how is it possible, that when I upload the 24 bit video of my Riverside Journey, it is not that fine on Youtube, however a same 24 bit video from the JVC uploaded to Youtube produces satisfactory results (so a quality that I call sharp). When the previous one was shot in 1080p 50fps, and the latter in 1080i.

It would be totally enough for me to answer these two things. Because I think to get an answer for this is still far from waiting for miracles.

[poisondeathray]

\

That would tell you that your video is impossible to encode in any decent way with low bitrate. And YouTube gives you low bitrate alwyas, for any resolution. Youtube even distributed 3x more bitrate to your video as oppose to that trailer, still not enough. You cannot even remotely compare that trailer to your video.

But youtube is especially bad for re-encoding practices, even with a given bitrate - for example when I download a h.264 version of his video locally, it has zero b-frames.

[pandy]

Well, personally I don't think that in 2015 expecting playback without jumps from clips recorded by Full HD handycams would be expecting miracles, considering in the world of VHS and MiniDV, it was possible.

Well... it depends from your software and hardware - we don't know so many details so it is difficult to advise you something except general advises how to verify if problem is related to browser or something else...
FYI for online streaming additionally problem can be related to link quality (WiFi? ISP issues?).
Download file and verify it locally, use HW acceleration in player (i use MPC-HC + dxva) to remove CPU limitations...

But youtube is especially bad for re-encoding practices, even with a given bitrate - for example when I download a h.264 version of his video locally, it has zero b-frames.

No longer valid for H.264 but valid for VP9... from some reason unknown to me...

VP9:

{
"frames": [
{
"media_type": "video",
"stream_index": 0,
"key_frame": 1,
"pkt_pts": 0,
"pkt_pts_time": "0:00:00.000000",
"pkt_dts": 0,
"pkt_dts_time": "0:00:00.000000",
"best_effort_timestamp": 0,
"best_effort_timestamp_time": "0:00:00.000000",
"pkt_duration": 16,
"pkt_duration_time": "0:00:00.016000",
"pkt_pos": "4731",
"pkt_size": "242",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "I",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 17,
"pkt_pts_time": "0:00:00.017000",
"pkt_dts": 17,
"pkt_dts_time": "0:00:00.017000",
"best_effort_timestamp": 17,
"best_effort_timestamp_time": "0:00:00.017000",
"pkt_duration": 16,
"pkt_duration_time": "0:00:00.016000",
"pkt_pos": "9352",
"pkt_size": "56",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 33,
"pkt_pts_time": "0:00:00.033000",
"pkt_dts": 33,
"pkt_dts_time": "0:00:00.033000",
"best_effort_timestamp": 33,
"best_effort_timestamp_time": "0:00:00.033000",
"pkt_duration": 16,
"pkt_duration_time": "0:00:00.016000",
"pkt_pos": "9414",
"pkt_size": "62",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 50,
"pkt_pts_time": "0:00:00.050000",
"pkt_dts": 50,
"pkt_dts_time": "0:00:00.050000",
"best_effort_timestamp": 50,
"best_effort_timestamp_time": "0:00:00.050000",
"pkt_duration": 16,
"pkt_duration_time": "0:00:00.016000",
"pkt_pos": "9489",
"pkt_size": "46",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 67,
"pkt_pts_time": "0:00:00.067000",
"pkt_dts": 67,
"pkt_dts_time": "0:00:00.067000",
"best_effort_timestamp": 67,
"best_effort_timestamp_time": "0:00:00.067000",
"pkt_duration": 16,
"pkt_duration_time": "0:00:00.016000",
"pkt_pos": "9549",
"pkt_size": "227",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 83,
"pkt_pts_time": "0:00:00.083000",
"pkt_dts": 83,
"pkt_dts_time": "0:00:00.083000",
"best_effort_timestamp": 83,
"best_effort_timestamp_time": "0:00:00.083000",
"pkt_duration": 16,
"pkt_duration_time": "0:00:00.016000",
"pkt_pos": "9783",
"pkt_size": "329",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 100,
"pkt_pts_time": "0:00:00.100000",
"pkt_dts": 100,
"pkt_dts_time": "0:00:00.100000",
"best_effort_timestamp": 100,
"best_effort_timestamp_time": "0:00:00.100000",
"pkt_duration": 16,
"pkt_duration_time": "0:00:00.016000",
"pkt_pos": "10126",
"pkt_size": "418",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},

And H.264 - same video:

{
"frames": [
{
"media_type": "video",
"stream_index": 0,
"key_frame": 1,
"pkt_pts": 0,
"pkt_pts_time": "0:00:00.000000",
"pkt_dts": 0,
"pkt_dts_time": "0:00:00.000000",
"best_effort_timestamp": 0,
"best_effort_timestamp_time": "0:00:00.000000",
"pkt_duration": 1500,
"pkt_duration_time": "0:00:00.016667",
"pkt_pos": "48",
"pkt_size": "428",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "I",
"coded_picture_number": 0,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 1500,
"pkt_pts_time": "0:00:00.016667",
"pkt_dts": 1500,
"pkt_dts_time": "0:00:00.016667",
"best_effort_timestamp": 1500,
"best_effort_timestamp_time": "0:00:00.016667",
"pkt_duration": 1500,
"pkt_duration_time": "0:00:00.016667",
"pkt_pos": "909",
"pkt_size": "67",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "B",
"coded_picture_number": 2,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 3000,
"pkt_pts_time": "0:00:00.033333",
"pkt_dts": 3000,
"pkt_dts_time": "0:00:00.033333",
"best_effort_timestamp": 3000,
"best_effort_timestamp_time": "0:00:00.033333",
"pkt_duration": 1500,
"pkt_duration_time": "0:00:00.016667",
"pkt_pos": "1341",
"pkt_size": "67",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "B",
"coded_picture_number": 3,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 4500,
"pkt_pts_time": "0:00:00.050000",
"pkt_dts": 4500,
"pkt_dts_time": "0:00:00.050000",
"best_effort_timestamp": 4500,
"best_effort_timestamp_time": "0:00:00.050000",
"pkt_duration": 1500,
"pkt_duration_time": "0:00:00.016667",
"pkt_pos": "476",
"pkt_size": "69",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 1,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 6000,
"pkt_pts_time": "0:00:00.066667",
"pkt_dts": 6000,
"pkt_dts_time": "0:00:00.066667",
"best_effort_timestamp": 6000,
"best_effort_timestamp_time": "0:00:00.066667",
"pkt_duration": 1500,
"pkt_duration_time": "0:00:00.016667",
"pkt_pos": "2403",
"pkt_size": "243",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "B",
"coded_picture_number": 5,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 7500,
"pkt_pts_time": "0:00:00.083333",
"pkt_dts": 7500,
"pkt_dts_time": "0:00:00.083333",
"best_effort_timestamp": 7500,
"best_effort_timestamp_time": "0:00:00.083333",
"pkt_duration": 1500,
"pkt_duration_time": "0:00:00.016667",
"pkt_pos": "3011",
"pkt_size": "308",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "B",
"coded_picture_number": 6,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},
{
"media_type": "video",
"stream_index": 0,
"key_frame": 0,
"pkt_pts": 9000,
"pkt_pts_time": "0:00:00.100000",
"pkt_dts": 9000,
"pkt_dts_time": "0:00:00.100000",
"best_effort_timestamp": 9000,
"best_effort_timestamp_time": "0:00:00.100000",
"pkt_duration": 1500,
"pkt_duration_time": "0:00:00.016667",
"pkt_pos": "1772",
"pkt_size": "631",
"width": 1920,
"height": 1080,
"pix_fmt": "yuv420p",
"sample_aspect_ratio": "1:1",
"pict_type": "P",
"coded_picture_number": 4,
"display_picture_number": 0,
"interlaced_frame": 0,
"top_field_first": 0,
"repeat_pict": 0
},

[poisondeathray]

His video was uploaded in 2012 and doesn't have a 1080p version . The 720p h264 version doesn't have b-frames

Deadpool trailer was uploaded Aug 2015, the 720p h264 version doesn't have b-frames either, but the 1080p h264 version does

Youtube also uses distributed encoding - the more joins, the lower the quality