Welcome to our community

Be a part of something great, join today!

Register Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

YouTube Video Codecs, Resolutions & Data Rates

Les Dittert said:
Brian,
I forgot to say before: You might want to add some grain in you test images, say on the right half. Grain is the dilemma for all codecs to deal with. It's all about preserving the grain.
If you want to get 'fancy' you can digitally compare the original with the compressed, and look at how things are preserved.
In the industry we use a term : PSNR , to quantify the quality.

Regards,
-Les Dittert
Click to expand...
All three of my examples, just about each scene already have grain added.
 
Les Dittert said:
Brian,
I forgot to say before: You might want to add some grain in you test images, say on the right half. Grain is the dilemma for all codecs to deal with. It's all about preserving the grain.
If you want to get 'fancy' you can digitally compare the original with the compressed, and look at how things are preserved.
In the industry we use a term : PSNR , to quantify the quality.

BTW, late next week I may be organizing a get together at my work, to review and rerun my DSLR vs Scarlet finding I have mentioned before. I'd like at least 4 users to attend.

Regards,
-Les Dittert
Click to expand...

Those are two really excellent suggestions. Thank you Les. =)

I'll add grain to the 1080p version of the file and upload that to do some more tests on. Also, I just obtained a utility for PSNR measurement, so I'll post those results also. These comparisons between the peak signal-to-noise ratio of the different files are really interesting. Excited to share them. :)
 
PSNR Tests (Continued)

Test 2
Same as Test 1, but with a different frame.


Loss in quality from original and YouTube-encoded QuickTime Uncompressed 10-bit 4:2:2 file:

2_original_youtube_qt_uncompressed_10bit.png



Loss in quality from original and YouTube-encoded MainConcept H.264 (10 Mbps, Level 4.1, High Profile) file:

2_original_youtube_mc_h264_10mbps.png



Difference in compression artifacts between YouTube-encoded Uncompressed 10-bit and MainConcept H.264 file:

2_difference_between_qt_uncompressed_10bit_mc_h264_10mbps.png


Again, the loss of quality is substantial. The additional compression artifacts from uploading an H.264 file rather than an Uncompressed file are even more visible than in Test 1.
 
Last edited:
Speaking of bit rates and youtube videos, take a look at this one I stumbled on :

http://youtu.be/y4JYe_TIVRg

( It's shot on a Panasonic GH2 .... however it is hacked to 176 Megabits/sec GOP1 here .... amazing how far they have pushed the little thing. )

-Les Dittert
 
Brian Iannone said:
Those are two really excellent suggestions. Thank you Les. =)

I'll add grain to the 1080p version of the file and upload that to do some more tests on. Also, I just obtained a utility for PSNR measurement, so I'll post those results also. These comparisons between the peak signal-to-noise ratio of the different files are really interesting. Excited to share them. :)
Click to expand...

Brian,
if you have the chance. You can test Cineform Neo 5.5. codec.
would be interesting to see how it behaves.
 
Les Dittert said:
Speaking of bit rates and youtube videos, take a look at this one I stumbled on :

http://youtu.be/y4JYe_TIVRg

( It's shot on a Panasonic GH2 .... however it is hacked to 176 Megabits/sec GOP1 here .... amazing how far they have pushed the little thing. )

-Les Dittert
Click to expand...

Very interesting video. It's amazing what $1,000 cameras can do now.

By the way, is there a PSNR testing utility that you'd recommend? I started using MSU's Video Quality Measurement Tool, but I'm just wondering if there's anything else you think may be better or more accurate.


luigivaltulini said:
Brian,
if you have the chance. You can test Cineform Neo 5.5. codec.
would be interesting to see how it behaves.
Click to expand...

Will do. That would be interesting to see.


I uploaded 70.56GB of 4K video to YouTube last night, so I'll (finally) be able to do the 4K tests today.
 
Hi Brian, another test you might want to do is measure the colour (space) shift that happens when vimeo and youtube re-encode. Maybe upload a few seconds of video of a colour chart, let vimeo and youtube do their thing, grab a screen shot and compare the differences in colour values. Anyone who's uploaded an anaglyph to vimeo should understand this.
 
Mark Pavy said:
Hi Brian, another test you might want to do is measure the colour (space) shift that happens when vimeo and youtube re-encode. Maybe upload a few seconds of video of a colour chart, let vimeo and youtube do their thing, grab a screen shot and compare the differences in colour values. Anyone who's uploaded an anaglyph to vimeo should understand this.
Click to expand...

Great idea. Would you happen to know where I can find a sample color chart shot? While I would shoot it myself, I don't yet have a camera capable of capturing something that I would consider to be accurate.
 
Brian Iannone said:
YouTube Encoder Data Rates

[TABLE="class: grid, width: 500"]
[TR]
[TD][/TD]
[TD]23.98 fps[/TD]
[TD]29.97 fps[/TD]
[/TR]
[TR]
[TD]4K (4096x2304)[/TD]
[TD]Results on 1/22[/TD]
[TD]Results on 1/22[/TD]
[/TR]
[TR]
[TD]1080p (1920x1080)[/TD]
[TD]3.75 Mbps[/TD]
[TD]5 Mbps[/TD]
[/TR]
[TR]
[TD]720p (1280x720)[/TD]
[TD]2 Mbps[/TD]
[TD]2.5 Mbps[/TD]
[/TR]
[TR]
[TD]480p (854x480)[/TD]
[TD]825 Kbps[/TD]
[TD]1 Mbps[/TD]
[/TR]
[TR]
[TD]360p (640x360)[/TD]
[TD]550 Kbps[/TD]
[TD]725 Kbps[/TD]
[/TR]
[TR]
[TD]240p (400x226)[/TD]
[TD]300 Kbps[/TD]
[TD]325 Kbps[/TD]
[/TR]
[/TABLE]

So, watching a video on YouTube at 1080p will, of course, show you a higher quality file than if you select 480p. But... If the original file that was uploaded to YouTube had a frame rate of 30p, it will be of "higher quality" than if it were a 24p file. Basically, the higher the frame rate, the larger the data rate that YouTube encodes your video in.

YouTube encoder quality: 1080p30 > 1080p24

Keep in mind however that YouTube encodes using VBR. So, there will always be a fluctuation in the data rate. The table above is the most accurate data I've been able to put together.
Click to expand...

Brian, this is a really interesting thread. Thank you for the effort you've put in to doing this.

One note though: The difference in data rates between 30fps and 24fps doesn't indicate a consistent difference in quality. At 720p, it would be completely accounted for by the increased data required by additional frames. At 1080, the 24fps file has a lower-than-expected data rate, but as the frame sizes get smaller, the 24 fps files seem to have an increasing data rate (relative to the 30fps files) that isn't explained by the difference in frame rate. So as frame size decreases, quality of 24fps files seems to increase relative to 30fps files.
 
D Fuller said:
Brian, this is a really interesting thread. Thank you for the effort you've put in to doing this.

One note though: The difference in data rates between 30fps and 24fps doesn't indicate a consistent difference in quality. At 720p, it would be completely accounted for by the increased data required by additional frames. At 1080, the 24fps file has a lower-than-expected data rate, but as the frame sizes get smaller, the 24 fps files seem to have an increasing data rate (relative to the 30fps files) that isn't explained by the difference in frame rate. So as frame size decreases, quality of 24fps files seems to increase relative to 30fps files.
Click to expand...

That's exactly what I would normally think. And, I'm nearly certain that in almost every scenario, that would occur. I'm not sure why YouTube's encoder is allocating higher quality to 30p files rather than 24p files, but all my tests are showing that. Still trying to understand why...

What I find even more interesting is that YouTube is encoding 60p files at an even higher quality than 30p files. (By "higher quality," I mean less compression artifacts and noise.) It doesn't appear to be an increase in data rate since YouTube is keeping it the same across 24p to 60p files; it's just assigning it based on the resolution.

I'm thinking about trying some 10-15p files. Based on my results so far, I'd expect the quality to decrease.
 
With these types of compression, "optimal" data rate does not scale proportionally to pixel count nor it does to fps.

Youtube is not responsible for perceivable higher quality at 60p, again it's the nature of this type of compression.
Also, some artifacts are less perceivable because they exchange 2.5 times faster.
 
Youtube recomend 50,000 kbps, that gives quiet big files to upload. Would vbr instead of cbr make it much less suitable for youtube and vimeo to encode from? I was thinking of putting maximum to 50,000 kbps but change the target to 20,000 kbps.
Also why is youtube so much faster to upload to compared to vimeo on a fast internetconection? We uploaded 1.6g on 6 minutes but the same file took perhaps 50 minutes on vimeo?
 
What was the final conclusion for the best settings to upload a 4K file to youtube then?
Additionally for people with poor upload speeds, what was the best upload settings in the H264 compression?

Thanks for this thread!
 
jake blackstone said:
Interesting discussion.
I wasn't aware, that youtube supports Prores uploads. Obviously this is much better, than uploading H264 and then let youtube transcode it again to H264. So, to test it I did two uploads-one Prores and the same material in H264 at same resolution from the same source. Unsurprisingly, Prores upload ended up looking better, than H264. Here are both examples:
Prores http://www.youtube.com/watch?v=ynrGtPmgJ90
H264 http://www.youtube.com/watch?v=htCyvK3bW1k
Herde is the same file on Vimeo at the same resolution
http://vimeo.com/29028959
Click to expand...

Do you think YouTube applies as high Bitrate settings as you can apply on your own H264?

-Michael
 
michael zaletel said:
Do you think YouTube applies as high Bitrate settings as you can apply on your own H264?
-Michael
Click to expand...

Good question.

H.264 encodes (at anything but VERY high bit rates) toss an awful lot of image data. It's hard to imagine that even low quality ProRes to H.264 encodes would be worse than what would happen to H.264 source material subjected to a second pass. I suppose the other key variable is the level of optimization for their player built into their encoding profile. Anyone know how sophisticated they are in that respect?

Cheers - #19
 
You must log in or register to reply here.
Back
Top