@Antoine-BBTV

In addition to what Al (@AAProds) has written, this page shows YT's own recommendations on settings for videos uploaded to its server.

HTH

Jeff
Forum Moderator

Thanks @AAProds and @emmrecs. Both of these comments were very helpful. @AAProds I set the maximum bitrate to 50000 and it helps quality alot. At that setting what would you reccommend for average bitrate? Also what would you reccommend for CBP, Maximum GOP Length, GOP Structure, Profile, Level, Coding Quality, Smart copy or HRD? My video is shot at 60fps.

@Antoine-BBTV

I'm not an expert on the export settings but I use:

-the Max bit rate as 25% more than the average bitrate. If you look at a Mediainfo report of a file, the average bitrate is very close to the Average Bitrate you set in the export box.

-If you export as MPEG 4, you won't have access to CBR (Constant Bitrate). I think that VBR is better than CBR because you will use less data on areas that don't change much.

-Defaults for Maximum GOP Length, GOP Structure, Profile, Level

-Coding Quality set to Best. I have never understood a program that defaults to less than the best quality. Given the bane of every video editor's existence is quality, why would it default to "standard"?

-Smart copy or HRD? I can't smart copy out of MEP or VPX (would be great if I could!) and I've never understood HRD. The Help file has a comprehensive section on exporting.

@Antoine-BBTV, @AAProds

Hi

. . . . what would you reccommend for CBP, Maximum GOP Length, GOP Structure, Profile, Level, Coding Quality, Smart copy or HRD . . . .

Of these settings the only one you could change is Coding quality though this should not be necessary - all export presets are optimised for best quality vs file size.

YT will re-encode the video to various resolutions, down to as low as 360p, and bitrates such that viewers on lower speed Internet connections get a reasonably smooth playback - you have no control or influence over this.

The YT encoders may use VP9 encoding for the higher resolution uploads.

. . . . Smart copy or HRD? I can't smart copy out of MEP or VPX . . . .

Smart copy has very strict rules for it to be used when exporting - the source video and export settings must match, there must be no other edits other than trimming clips, no effects, no transitions, titles, or images.

HRD - Hypothetical Reference Decoder, is a 'virtual decoder' that checks and puts constraints on the data stream for compatibility with the video standard during the encoding process - this can slow down the encoding and, as I understand it, is likely to kick in when the data stream bitrate is set too high for the video standard.

HTH

John EB

@johnebaker @Antoine-BBTV

Of these settings the only one you could change is Coding quality though this should not be necessary - all export presets are optimised for best quality vs file size.

John, my point is that file size may irrelevant in this day and age. You wind up the bitrate as much as you can stand. And if "Best" gives a bigger file size than "Balanced" for better quality, then so be it. I have never had any trouble modifying the presets, in fact I never use the Magix export presets unmodified.

The YT encoders may use VP9 encoding for the higher resolution uploads.

Will, at 1440P or greater, as I mentioned above.

Smart copy has very strict rules for it to be used when exporting - the source video and export settings must match, there must be no other edits other than trimming clips, no effects, no transitions, titles, or images.

Magix's idea of smart rendering is fundamentally flawed in my view. The whole point/principle of smart copying/rendering is that the current file is copied using it's current characteristics (apart from edit points, obviously). Virtual Dub has been doing this for decades, and Magix itself used to do it with MPEG 2. And of course Videoredo will also do it on virtually any type of file.

@AAProds @Antoine-BBTV @emmrecs

Hi Antoine.

Al, (AAProds) makes some good points and almost answers his own question of why Coding Quality is set to Average rather then best

Only six out of ten people in the world currently use the internet either through budgetary constraints or technical.

Out of those 96% only access the internet via a mobile device as from 2021 last year. Those are the most current figures I can lay my hands on. Out of those the top ten with 5G are

Top 10 countries with Highest 5G Speeds in 2022 and also focus on India's 5G

China: United States: South Korea: United Kingdom: Spain: Canada:Australia:Saudi Arabia:

India is the next one on the list but connection is not even across such a vast country.

So if you have your own website and wish to host your own videos it may be prudent to make them watchable over poorer connections as high quality video needs quite high bandwidth if it is going to stream video real time.

I have taken the following information from this site.

Raw, uncompressed video carried over an HDMI, HD-SDI, or Ethernet cable requires a lot of bandwidth. The following chart provides approximate required bit rates to carry uncompressed digital video of different sizes (resolutions):

Resolution Uncompressed Bit rates

1280×720 (720p)~ 1.5 Gbps

1920×1080 (1080p)~ 3 Gbps

3180 x 2160 (2160p or 4K)~ 12 Gbps

Which if you are using a mobile device with or without data caps could quickly become expensive if you could in fact even view it over your connection.

However, for the remaining 4% of the population (or possibly much higher as there is always crossover users who may use their phones but link to a local wireless network from home or similar) They may still be using providers who cap their data usage.

The same can be said for the rest of the base settings within MEP when thinking about output quality.

However, for somewhere like YouTube that can host 8K UHD content and if your internet connection is good enough, as well as your PC with the ability to play it, it can be a good place to try to upload the best quality you can as it will be re-encoded again by YouTube and subsequently there will be a very small quality loss regardless of the quality of the uploaded file.

Again though. depending on the person at the end of the chain. the viewer, the experience may be even more limited or better than average.

The link I gave above I feel is s good layman's guide as to why certain settings may be important and a reasonable guide as to what settings could be used. There can be down sides to some of the settings as well as up sides when selecting what quality you wish the end user to have. It is a good idea to read the fairly short blog (for such a complex subject) To try to get some sort of handle on this.

I think Al is correct at saying adding 25% more than the average bit rate should be safe enough to avoid seeing additional artifacts creeping into the video stream but the GOP structure also plays a huge part. Read the blog for more detail. You may need more than 25% if you decide to encode the file with all I frames. More of that later.

Al is also correct in saying there is no constant bit rate setting for an MP4 export although selecting the correct GOP structure and quality settings can make the end file almost non variable in its overall bit rate.

Adding more bits to the bit rate can and often does one or more of a few things.

It can produce files with no additional data at all within some of the bits where it is not needed to encode the file, resulting in a much large file than is strictly needed.

Setting random numbers in the minimum / average / Maximum fields can result in files that are either unplayable or '0 bit files' containing no data, or files that stutter on playback.

Setting really high bit rates for files meant for higher resolutions may tax your system and cause crashes within the program.

Also what would you reccommend for CBP, Maximum GOP Length, GOP Structure, Profile, Level, Coding Quality, Smart copy or HRD? My video is shot at 60fps.

As Al correctly pointed out there is no option for a constant bit rate but it can be made minimal to the point of not being recorded in the media metadata.

I normally select a Magix preset that has either a higher frame rate but the same resolution as my project or one that has the next highest resolution to get a good balance between the Average Maximum and minimum bit rates and then adjust the file export to the correct frame rate and resolution. That for me has never resulted in an unusable file and helps keep the bit rate high.

The Maximum GOP length can be longer or shorter than the frame rate although the ones within the presets of Magix are a good starter in my opinion. I often don't vary them for YouTube uploads. Again there is an explanation and images to show why on the blog but using all I frames requires an overall higher bit rate to avoid excessive compression of the I frames.

GOP structure is another mine field. Personally I use all I frames most of the time or the default IBBP of the program sometimes. The advantage of an all I frame export for me personally is it helps retain the amounts of bits per frame for each frame with minimal compression(if the original file has enough bits and the bit rate is nearer 33% more rather than 25%. It allows the YouTube conversions to make less mistakes and retain quality when they re-encode the video to the various resolutions your project will eventually be able to be viewed at on their platform. Again there are pros and cons.

Profile set to 'High' as 'Main' is an older version of the same thing. There is no advantage to selecting Main as the compression is worse.

Level. Use the lowest of the options where numbers are available. Quite often there is only the choice of Auto or either 5.2 or 5.1. I tend to use the lowest numbered option so in that instance 5.1. If there is a lower number, use that.

Coding quality. Use Best.

That combination I find gives the largest file size output with hardly any variation if any on the average bit rate providing the source file also has next to no variation in its bit rate.

Normally for me it keeps the file close to the files maximum bit rate. Hardware encoding enabled if available and

working.

Audio should be set to the best quality settings the program allows which is normal 48kHz at a 192 kBits per second rate.

At this point you may find that Smart Copy is greyed out. Not a problem. There can be no Smart Copy as the bit rates no longer match.

HRD. HRD, as John Baker pointed out to me a long time ago, double checks the written encoding to make sure what is written to disc is the same as the file being encoded. It will slow down the export as will all the alterations I've previously mentioned. It is the main downside of trying to export the best quality you can.

Personally I have never used it. I think it would be useful if rendering the file to an old SATA drive or a slow external USB drive where there may be some data loss during rendering. Most SSDs would be looking out for errors anyway but a dodgy USB connection or an old worn mechanical drive may not pick up on the errors.

If for any reason you export to a lower frame rate or resolution than the source files, then the only setting that may be needed to be altered to keep file integrity would be the GOP Length for a frame rate change. Just rendering to a lower resolution nearly never makes any difference to quality. It may make the resulting video look sharper viewed on the same screen as was rendered on.

Altering frame rates can result in visible tearing of frames (Jagged lines on sharply defined objects or visible juddering) but not always.

Lowering the resolution almost never upsets anything other than when viewed on larger screens may appear slightly softer.

MEP and VPX can upscale quite well with the correct settings but may strain the system or cause crashes. A lot will depend on the overall strength of the system and experimenting with the settings both for export and with some subtle use of some of the effects within the editor. There will be more ways of getting a bad looking render than a good looking one, so strictly for experimentation at first.

Things to look out for in the final rendered file.

Blockiness in the shadow areas compared with the master file.

Blockiness in large ares of constant colour. Both tend to move around while viewing and can be a distraction.

(Well I find it a distraction) Some people never seem to notice.

Visible lined graduation of large areas of the same colour.

Tearing of frames. (Jagged edges on sharply defined objects within the frame or uneven wobble on panned shots)

Over sharpening of softening of objects / subjects / background. Lack of detail compared to the original footage or 'Haloing', Bright specs appearing at random from over-sharpening or a thickening of lines at the edges of sharply defined objects / People, etc.

Those are my personal basic choices and things to look for but nothing much helps bad footage to begin with.

It is very difficult indeed to make any footage look better than the original but artistic alterations can give footage a more pleasing look.

Ray.

@AAProds

Hi Alwyn

. . . . Magix's idea of smart rendering is fundamentally flawed in my view. The whole point/principle of smart copying/rendering is that the current file is copied using it's current characteristics  . . . .

. . . thus the export setting must have the same 'characteristics' so the video it does get re-rendered - this is where things get tricky when using video from cameras, drones etc and mixing source video types.

Personally I think Smart copy is of no use for most users as they will be adding titles, effects, audio etc to their projects all requiring re-rendering.

I have used Smart copy successfully for creating BD discs which have a large number of individual short movies, that VPX balked at when rendering. The last one, a trip down your way, had 32 videos plus 6 videos for menu backgrounds with a total run time of 2hrs 37mins, these were created over a long period, all exported in BD compliant format. For burning to disc these were imported back into VPX, the menus modified, and using the same preset for burning as for exporting the compliant videos and Smart copy - total disc creation to ISO ready for burning took 40 mins in VPX 12. The final disc was created using the same BD compliant videos in DVD Architect, as the VPX disc did not have some functionality I wanted in the menu system ie it does not support Play lists.

John EB

Thanks you guys so much for your input on this. I've been trying to find these answers online everywhere and could not find a clear understanding. This info helps me a ton! What I've been finding is my quality is amazing on youtube when the bandwidth is good. But in certain situations where the bandwidth is lower my video seems to lose more quality then others on youtube under the same low bandwidth. I will play with some of these configurations you guys mentioned and see if it can help me get a better average quality under low bandwith playback so general users will have a better viewing under low bandwidth situations. Thanks again

@Antoine-BBTV

Hi Antoine.

Let us know how you get on.

I'm also not quite sure what you are saying. Are you saying there are times when you can't see your videos at their full resolution and have to view them at say 480p or are you saying you watch them at full resolution but the quality is worse because of the limited bandwidth?

Unlike AAProds (Al) I can get YouTube to use the vp09 codec using a maximum resolution of video of 1920 x 1080p. which is good as I don't always shoot at higher resolutions if I need to mix slow motion with normal footage as my cameras can't capture slow motion at 4K. However, that requires more tinkering with the settings and can still fail if I can't keep the bit rate of the original recordings high.

How my HD videos look when I can get YouTube to use vp09

How much softer that looks when viewed at 480p full screen

Sections of the two side by side enlarged to see how the compression works

My own experiments have shown me that the source files play a vital part in retaining detail and it is not down purely to getting YouTube to encode to vp09 although that helps.

Some blockiness still occurs at the lower viewing resolutions and sometime banding of graduated tonal regions can also happen.

Shown at 4K

Shown full screen at 480p Notice the shadow area and overall softening.

Although I do edit in 4K I currently only have 1920 x 1080 monitors.

That can happen regardless of which encoder gets used by YouTube but it does tend to be less if you can invoke the vp09 codec.

But why do these differences happen in the first place?

I really don't know. All I can do is observe what I manage to produce as an end resulting rendered export and what effects that render.

What I see is that the higher the bit rate I can record on camera the higher a bit rate I can get on the final render regardless of how much I increase the bit rate settings. Beyond a certain point the extra bits become padding with no additional information in them as far as I can tell. I expect at the YouTube end, if their encoder spots that lack of data it will reduce the amount of bits the encoder uses at their end.

If I add clips some clips with lower bit rates than other clips to a project it will lower the bit rate of the rendered project file. Even going well above adding the extra 25-35% recommended earlier to say 60% to the bit rates it makes very little difference.

Here I have the data from two clips used in the same project on the left of each image. On the right is the final render data.

This clip has a much wider gap in bit rates

This one shows no variation in the bit rate although overall lower

Combined they have reduced the final renders bit rate. You cannot avoid some compression regardless of settings. All you can do is try to make the difference as little as you can.

Recording at 4K can produce files with as little as 36Mb/s using the MP4 format depending on the camera and how much compression the camera actually adds to the encoding. Over and under exposed areas can also contribute to lower bits being recorded.

It is my personal experience that I have to take care when shooting video and get as much information into each file as I can. Different cameras may also compress the output more or less than others for a given resolution / frame rate / codec.

So when you compare your output on YouTube to another persons output, you may be trying to compare apples to oranges.

It is very difficult to know.

All you can do is the best you can and tinker at the edges from time to time to see if you can improve the results.

Ray.

@CubeAce

Unlike AAProds (Al) I can get YouTube to use the vp09 codec using a maximum resolution of video of 1920 x 1080p.

Ray, I'd be interested in the details of your 1080 files that invoke VP9; Mediainfo report and screenshots of your main export screen and the Advanced screen.

@AAProds

Hi Al.

It's been a while since I did one. I'll have to try to recreate one as I didn't save the settings as a preset.

I do remember the rendered file sizes were huge.

I tend to work with 4K files only now. I'll see what I can do.

In the mean time I found another one of mine I did the same way.

You can check to see if the lower resolution files are worth bothering about. They often don't look any better than YouTube using their avc codec.

Also if you can't supply source files with enough bits per second it tends not to work.

This may take some time.

Ray.

@CubeAce

Ray, don't go too much effort. Just a Mediainfo on one of the originals you uploaded would be good.

I'll have a play around with bitrates but I can't find anything on the internet that indicates that, at the present time, you'll get VP9 with less than a 1440P original. I have seen a couple of posts which say you'll get it if your channel is big enough, subscribers/views-wise.

@AAProds

Hi Al.

I think I have six subscribers 🤣.

The right hand file in both the MediaInfo images is the one from the exported file file that I used to upload the War and Peace video to YouTube.

Individual files for the project did not vary by more than 2Mb/s as all were taken on the same day which was bright and cloudless.

Here is the MediaInfo for one of those files.

General
Complete name                            : H:\2019 War and Peace Revival\DJI_0901_C1660759.MP4
Format                                   : MPEG-4
Format profile                           : JVT
Codec ID                                 : avc1 (avc1/isom)
File size                                : 224 MiB
Duration                                 : 23 s 424 ms
Overall bit rate mode                    : Variable
Overall bit rate                         : 80.4 Mb/s
Encoded date                             : UTC 2019-07-24 10:11:35
Tagged date                              : UTC 2019-07-24 10:11:35
Comment                                  : DE=None, Type=Normal, HQ=Normal, Mode=P
gpt                                      : +27.60
gyw                                      : +0.70
grl                                      : -3.50

Video
ID                                       : 1
Format                                   : AVC
Format/Info                              : Advanced Video Codec
Format profile                           : High@L5
Format settings                          : CABAC / 1 Ref Frames
Format settings, CABAC                   : Yes
Format settings, Reference frames        : 1 frame
Format settings, GOP                     : M=1, N=30
Codec ID                                 : avc1
Codec ID/Info                            : Advanced Video Coding
Duration                                 : 23 s 423 ms
Bit rate mode                            : Variable
Bit rate                                 : 80.0 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)                       : 0.644
Stream size                              : 224 MiB (100%)
Title                                    : DJI.AVC
Language                                 : English
Encoded date                             : UTC 2019-07-24 10:11:35
Tagged date                              : UTC 2019-07-24 10:11:35
Color range                              : Limited
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                                 : 23 s 424 ms
Bit rate mode                            : Constant
Bit rate                                 : 192 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                              : 541 KiB (0%)
Title                                    : DJI.AAC
Language                                 : English
Encoded date                             : UTC 2019-07-24 10:11:35
Tagged date                              : UTC 2019-07-24 10:11:35

Other
Type                                     : meta
Duration                                 : 1 s 0 ms
Bit rate mode                            : Constant
Default                                  : No

You can see there is no higher or lower bit rate recorded even though it states it is variable.

I have no idea how that is possible.

I have now tried twice with another project with no success. I will try to recreate it with the War and peace project again if it is still intact. For my own curiosity if nothing else.

Ray.

@Antoine-BBTV @CubeAce @johnebaker @emmrecs

I've done some more test uploads. Every case of 1080P50 (ie 1920x1080P at 50fps), even at 100MB/sec, was encoded by YT in AVC1. I uploaded the files 12 hours ago and they are still AVC1.

I then tried 1440P (2560x1440P) at both 50fps and 25fps at a lowly 6mbps and YT encoded in VP9 both times, immediately.

For all exports, I used default settings apart from the resolution and bitrate.

Unless there are some esoteric Magix export settings that will invoke VP9, I stand by my hypothesis that anything less than 1440P won't give you YT VP9 encoding. The 1440P can be 16:9 or 4;3.

VP9 encoding is better than AVC1.

@AAProds, @CubeAce

Hi Alwyn, Ray

MediaInfo data does not always show the Maximum Bitrate, because this has not been written into the header data.

Experimenting with export as 4K UHD with an Average bitrate of 60 Mb/s and Maximum Bitrate of 80 Mb/s from a project consisting of a 720p 4Mb/s average BR video clip and 4K UHD clip 56 Mb/s average BR and 60 Mb/s maximum BR, resulted in an export with an average BR of 60 Mb/s.

However, the problem is comparing bitrates and ranges does not show the whole picture of what the BR is doing in the exported file, eg the above exported file looks like this:

The 720p 4Mb/s upscaled has an average BR of 25 MB/s, as can be seen above - it was the first clip on the timeline, when the 4K UHD was encoded it has an average BR of 60 Mb/s, however the peak BR is a whopping 240 Mb/s and this happens to occur with the I frames - the original video clip use also peaks to 250 Mb/s.

Note:

• I have rounded the Mb/s to the nearest whole number
• The 720p clip has a GOP length of 25 and is I and P frames only, the 4K source video file is I, B and P frames with a GOP length of 12.
• The peak bitrate is correct, meets the HDR check when exported with the option turned on and is allowed - see below.

As I understand it, the Maximum setting is, an 'average' maximum the exported clip can be, ie when an I frame goes above the maximum setting and the B and P frames are below, and the average is less than the maximum set, the high peak values are allowed. The HRD option compensates if the B and P frames are of higher peak value, by lowering the I frame peak value. I do stand to be corrected on this.

John EB

@AAProds @Antoine-BBTV @emmrecs @johnebaker

Hi Al, Antoine.

I'm glad this question cropped up. It has made me question the whole topic all over again as most of us I'm sure want to present our videos in the best quality possible even if the public in general don't always seem to mind.

I have so far, like Al, completely failed to upload one more video at 1920 x 1080 so I am at a loss as to why my earlier attempts succeeded. No matter. As Al has correctly pointed out, just upping the resolution to 2560 x 1440 seems to invoke vp09 every time.

I did upload the War and Peace project again although I couldn't get the end rendered file to exactly match the previously uploaded file that worked to show Al the settings. I don't know why that is yet but I'm still working on it in my spare time. I'm wondering if Magix has slightly altered some of the base settings or the codecs themselves have been slightly altered (Newer versions?) or YouTube has changed how it decides to encode a given upload.

I would put my money on the last idea. YouTube constantly updates how it does things. We have all seen it.

I do have a few questions though now I've had a chance to do some comparisons.

As I said, I uploaded one of my projects I had uploaded before and successfully acquired a vp09 encoding for it. So all the source files were the same, the editing was the same. The only difference was not having the settings I had earlier for export.

So my second attempt at uploading my War and Peace video ended with it being encoded to avc1.

I looked at both videos online at once getting them as close in sync as I could. Possibly within 10 frames or less.

Which is vp09 or avc1?

Second try anyone?

Personally I could not tell the difference. So the two videos to check out are below.

You can check both were only encoded to 1920 x 1080 for yourselves. You can even try running them both together to check the buffer reading you get were similar to mine. You can tell from the time code that I had both running at the same time. It doesn't seem to matter which you load first or play first as to which video gets the better buffer depth.

So what I found interesting was the question is vp09 doing a better encode than avc1 or just a more efficient one?

If it uses less bits to transmit the image back to the user then the usual wisdom as I understand it is it take more power at the receiving end to decode it. I will ask those wiser than me on the forum if they can answer that.

I also ask that because the buffer health on my machine was better to the avc1 encoding than for the vp09 version. I assume that to mean I am storing more data locally before I see it on the monitor and there is less chance of buffering. Please correct me if I'm wrong.

So I'm now wondering, if the problem of playback quality it one of internet download speed, which one is better to watch?

Ray.

@CubeAce

Hi Ray

. . . . I have so far, like Al, completely failed to upload one more video at 1920 x 1080 so I am at a loss as to why my earlier attempts succeeded.  . . .

Probably the same reason why a lot of the earlier 'methods' posted on the Internet no longer work - they have cottoned on to this and are eliminating hacks to get VP9 encoding .

John EB

@johnebaker

Hi John.

I wonder if YouTube are trying to save power? I think that's doubtful but if the files take up less space you would think they would encode all uploads to vp09 unless it is so that the 5% of the worlds population still running Windows Explorer still have content to watch as it can't decode vp09.

Looking at the information on the Wikki page for VP9 (I can't find anything better) it states there is no benefit to video resolutions at 1920 x 1080 or below which I'm not completely convinced of although at 1080p I confess I can't see a difference between the two uploads.

As the files seem easier to stream with only a slight hit on CPU usage I'm now wondering why there are no available encoders for public use. Google have said it is open source or do they mean only the decoder is open source? Think what it could mean for camcorders and editing.

[Edit]

Then there is that buffer difference to consider. Assuming you have more data in your buffer then at a lower connection speed I assume there is less chance of actual buffering occurring if you have more on your computer or phone at any one time and that the avc1 codec would perform better.

I'm also assuming others will have the same experience in buffer differences.

Ray.

@CubeAce

Hi Ray

. . . . Then there is that buffer difference to consider . . . .

In essence when you view a video stream you do not get a directed stream of video data packets taking a single route to your PC, the video data which is 'thrown out' in data packets onto the Internet tagged with your PC as the destination. As the data packets can take many different routes it can arrive at your PC at different times and out of order, once enough data is in your PC it is re-assembled into the video stream and played - this is where the buffers come in to play.

The browser requires a minimum amount of data arriving to play the video smoothly, if this does not occur, the sender 'gets notified' and will drop down to a lower resolution, there is usually a 'negotiation' between the browser at the initial request for the video and when the buffers are not getting enough data.

Manually overriding the resolution you receive to a higher one can introduce pauses during playback while sufficient data has been received for playback to resume.

The codec used to compress the video can help reduce buffering issues, eg h.265 and VP-9, are great compressors for higher resolution video, ie > 1920 x 1080, which means the data rate out of the servers is lower, however more processing is required at the receiving end.

HTH

John EB

@johnebaker

Hi John.

Is there also a limit to what an individual can view based on their system specs?

I ask because this video is put up as a 12K example on YouTube

But the top resolution I get offered is 8K and that doesn't play smoothly but has a constant jerkiness to it which is a consistent micro jerk to describe it as best I can. I can see my internet speed is fast enough and the buffer seems enough to cope but my graphics capabilities are maxed out.

The picture looks amazing even though my monitor is limited in resolution and nit brightness. Just not smooth.

I'm wondering, do those of you that have better graphics cards get offered the 12K version or is the video itself 8K but recorded originally at 12 K?

Can you get the video to stream smoothly at the maximum resolution you are offered?

Ray.

@CubeAce

Hi Ray

. . . . I ask because this video  is put up as a 12K example on YouTube . . . .

The descriptions say they are 8K videos and they play smoothly in Chrome and Edge, does not play smoothly at 8K resolution in Opera or Firefox with them dropping up to 50%+ frames.

12K - clickbait !

John EB

@johnebaker

Hi John.

For me, the playback issue on YouTube is getting curiouser and curiouser.

I have found some videos always but always play back with roughly the same low buffer readings coming back on each visit. (not my videos but other YouTubers) but not all their videos suffer, just specific videos. Time after time, regardless of resolution watched or their uploaded resolutions.

I do use Firefox. I didn't get dropped frames watching the video I posted above, just the jerkiness of playback.

I did visit the main channel and they record at 12K but YouTube doesn't support 12 K.

I played the same video again this morning but on my Intel powered monitor and it played back smoothly with no issues. Weirdly, and I have no idea why this happens but look at my PC GPU usage.

The video is playing on the Intel GPU monitor and yet there is no discernible Intel GPU usage whereas if I play the video on the nvidia powered monitor the Intel GPU is also pressed into service and I get the juddering.

I can't see any discernible difference in the amount of work the nvidia card has to do either.

The more I delve into this, the more perplexed I become although I'm now accepting that some of the problems of quality on playback may be beyond an individual viewers or content creators ability to rectify.

I always knew that no matter what I do or see at my end that another persons experience may differ due to internet connection speeds as well as web browser choice and equipment at the viewing end may not be calibrated at all for colour. That another person may even get a better experience that I can have if their monitor is better and set up well.

I never considered or gave a thought to how a graphics card monitor combination could also have an effect if anyone chose to view web content from one graphics card output over another.

Ray.