best export settings for youtube and vimeo ?

Hello.
Please tell me what are the best export settings for youtube and vimeo?

Adjust resolution and frame rate accordingly.

Upload Settings.png (39.48 KiB) Viewed 15779 times

Adjust resolution and frame rate accordingly.

The attachment Upload Settings.png is no longer available

Are you kidding me? AVI in 2023? The OP asked for the best export settings not the worse possible settings. Terrible advice.

Hello.
Please tell me what are the best export settings for youtube and vimeo?

Google who actually owns YouTube has their recommended video settings posted here:
https://support.google.com/youtube/answer/1722171

Below is a screenshot of the settings that I use for YouTube and Vimeo uploads. Also, if you go to the Render tab and scroll to the right under render settings you can select the YouTube option and change the resolution to match your resolution (i.e. 4K). I don't use this method because that option hides too much information such as bitrate and what encoder will be used. I prefer to know all of my settings for YT.

YT_Vimeo_UploadSettings.JPG (57.76 KiB) Viewed 15760 times

Most of the folks I know of who want better Q recommend using Cineform or ProRes rather than H.264.

And those recommended bitrates are rather low it would seem. Might be what ends up after the inevitable re-encoding process. Not what I'd upload as.

Sent from my SM-S908U using Tapatalk

I just use NVENC H265 at "best" quality. Does the trick just fine.

If you're going to be using an HD timeline then, immediately before delivery, it's worth setting your timeline and rendered output to 2K resolution because that forces YT to encode with a bit more quality for its 1080p version (using AV1) so you get less in the way compression artifacts.

There's no benefit to using H264 or other codecs because H265 "best" is still much better than the resulting output from YouTube's rencoding process and H265 saves upload times.

Since you asked for "best" settings...

You will need to download the free voukoder software. This will give you access to the AV1 codec. Once you do that, your videos can have the best quality available for youtube streaming. The only other thing is, when you set up the voukoder and select it for the export, you have to click the 'configure voukoder' button, change the codec to AV1, then click apply. This has to be done at least once every time you open DR. But, if you are rendering multiple clips at once, voukoder will maintain the codec setting for all the renders. A very minor thing.

The video quality is superior to h.264 and h.265, in addition to the rendered file being smaller, sometimes by a very large amount. For example, a recent file I rendered is 11:32 in length, 4kUHD 60fps, and is only 2.79GB.

deliver settings

Screenshot 2023-03-14 195129.png (65.89 KiB) Viewed 15692 times

Oleg, as you can see we each have our own preferences. These are mine.

Hey guys, i did not know voukoder has connector (plugin) for DVR !

And i exported a short 5 min 1080p60 video in both best dvr studio h265 settings and in the best voukoder h265 settings. And then i compared them both with FFmetrics.

Voukoder DOES NOT even come close to the DVR export in terms of quality.

Are you kidding me? AVI in 2023?

Nope, not kidding.

The quality will be excellent.

Hey guys, i did not know voukoder has connector (plugin) for DVR !

And i exported a short 5 min video in both best dvr studio h265 settings and in the best voukoder h265 settings. And then i compared them both with FFmetrics.

Voukoder DOES NOT even come close to the DVR export in terms of quality.

Out of curiosity, what was the video resolution? I know for smaller resolutions AV1 does not do well at all. For higher resolutions, I have never seen AV1 not be as good if not better than h.265.

Most of the folks I know of who want better Q recommend using Cineform or ProRes rather than H.264.

And those recommended bitrates are rather low it would seem. Might be what ends up after the inevitable re-encoding process. Not what I'd upload as.

Sent from my SM-S908U using Tapatalk

Of course Cineform or ProRes are better than H.264 but remember the OP specifically asked about YT and Vimeo. I think it is safe to say that Google who owns YT will know how to get the best quality on their own platform and their own documentation says to use H.264. I like my bitrate at 16Mb/s because I do not shoot fast action or sports, so I don't have any macro blocking or artifacts on YT with that bitrate but I do save quite a bit on file sizes and upload times. If you are shooting fast action, sports, or something where every frame changes like a timelapse then a higher bitrate would make more sense.

For YT its very simple; follow Google's posted recommendations and you will get the best quality, upload anything else and there is a chance that YT's re-encoding process will trash your footage. Also, a lot of people don't realize that YT has two tiers of quality; for celebrities, Hollywood, etc. they provide higher quality streaming than for the rest of us so it is impossible to get some of the quality you see on YT from famous channels if your channel is not comparable. I don't use Vimeo much so I am not as familiar with it but for simplicity I just use the same export settings for both platforms.

I just use NVENC H265 at "best" quality. Does the trick just fine.

There's no benefit to using H264 or other codecs because H265 "best" is still much better than the resulting output from YouTube's rencoding process and H265 saves upload times.

The benefit is that when you use H.264 you are adhering to YT's posted video codec recommendations which means you are less likely to get artifacts from YT's re-encoding process.

As you can see from YT's recommendations, H.265 is not their recommended codec. Google is very clear in their recommendations and I match them exactly in my export settings except for the bitrate since I don't need that high of a bitrate for the content that I shoot.

https://support.google.com/youtube/answ ... 171#zippy=

Another benefit that I have seen is that by following their recommendations the YouTube video processing times are much faster; I have seen a 10min 4K video take less than 20min for the 4K version to be available on YT.

I think it is safe to say that Google who owns YT will know how to get the best quality on their own platform and their own documentation says to use H.264.

I think that's a poor argument. Those specs aren't really aimed at "best quality". They're aimed much more at "common usability".

You Tube also has this list of accepted codecs for upload. The recommended MP4 is maybe for the majority of less experienced users, and it is a very common format.

I now always upload h265 UHD video in a .mov container with Linear PCM audio and get much better results that h264 MP4 where the audio is restricted to AAC. H265 is also a much better compression algorithm with far fewer artifacts than h264.

I now always upload h265 UHD video in a .mov container with Linear PCM audio and get much better results that h264 MP4 where the audio is restricted to AAC. H265 is also a much better compression algorithm with far fewer artifacts than h264.

Agreed.

And I usually have a 20 minute UHD video rendered to 4K within about 10 minutes by YT when I upload as an H265. Even though I have gigabit fiber here, for some reason (from New Zealand) Youtube limits upload speeds to about 50Mb/S so the smaller filesize of H265 saves me time.

H.264 with the MP4 file format minimises the time it takes YouTube to encode. Pretty much creates the SD & HD on upload by the time you've done the description and studio stuff.

Like Bruce, I have no problem with the time taken by YT to render a 4k video. In fact I would say that YT now seems to create the various resolutions much faster than it used to.

I did VMAF testing of what settings give the best quality on YT. Results are here

https://www.reddit.com/r/videography/co ... &context=3

I did VMAF testing of what settings give the best quality on YT. Results are here

https://www.reddit.com/r/videography/co ... &context=3

That looks to be more of a review of Youtube's encoding of VP9 for 1440P+ and AVC for 1080P when we factor out the ridiculously low bitrate uploads

Like Bruce, I have no problem with the time taken by YT to render a 4k video. In fact I would say that YT now seems to create the various resolutions much faster than it used to.

Same experience here, also it seem to use their vp codec instead of old avc for h265 uploads.

Yeah...I've looked into the codecs a bit further, and I am still of the opinion that AV1 is THE option for HIGHEST quality on youtube. This does not factor in time; time for encoding or upload. It is a beast to encode...unless you have an Nvidia 40 series or Intel Arc series GPU. As cheap as the lower end Intel Arc is...it may be worth picking one up to add to your system just for the AV1 hardware encoding.

I did a quick check around on youtube to see if there were any direct comparisons (so I didn't need to do one again myself ) and found a few. They all show AV1 to be the best. Here are a couple (of the many) examples I saw.

H.264 vs HEVC vs AV1- Recording + Support, Editing & Uploading

YouTube Video Codecs Comparison (AV1 vs h264 vs VP9)

As a final note, it may be worth considering the major platforms ARE moving toward the adoption of AV1 as the primary codec.

I don't think all over the place game footage, apart from hurting the eyes, is a great way to demonstrate the differences between codecs.

Also remember that "best" is the enemy of "good enough".

If you can't tell by looking it doesn't matter.

That looks to be more of a review of Youtube's encoding of VP9 for 1440P+ and AVC for 1080P when we factor out the ridiculously low bitrate uploads

You need to read the post to understand the bars. The bitrates show are low, those are what YT gives back to you, basically impossible to get more than 20mbit back.

I uploaded DNxHR HQ, 120/60/20 mbit h.265. So there was no lack of bitrate for the uploads.


Yes, clearly no one should ever upload 1080p to YT, it is butchered.

I don't think all over the place game footage, apart from hurting the eyes, is a great way to demonstrate the differences between codecs.

Actually, it is a surprisingly good method to demonstrate such. There is so much fast-paced activity happening, frequently with dramatic color contrasting happening in the scene, that it is more likely to induce glitching into the video. I don't particularly care for the subject material either...but it really is a great source to push the codecs into failure.

Also remember that "best" is the enemy of "good enough".

If you can't tell by looking it doesn't matter.

I agree this should be considered. I consider time to encode, time to upload, storage requirement, playback quality...basically everything. For me, AV1 just ends up being the most efficient. That's probably partly because I run on an i9-13900k. If I were to add an Intel Arc GPU to my system, then I could just do hardware encoding directly. Nvidia's 40 series are just laughably expensive, so I won't even consider a 4090 right now. But, a person could pick up an A390 (I think) for right around $140USD (last time I looked on newegg) and have AV1 hardware encoding.

That said...they asked for the best for youtube (and Vimeo). AV1 is objectively the best quality. It is also the smallest file size. So, really...the original question needs clarification. Are they looking for best quality or most efficient overall? But then, we need to know the hardware, etc. Since I didn't want to ask about hardware, I simply assumed they were asking for the best quality and ignored efficiency.

I had previously tested H.265 vs 264 and posted results above. Lots of discussion here about AV1. so I repeated methodology with AV1.



https://www.reddit.com/r/videography/co ... lity_vmaf/

AV1 is not a magic bullet to get better quality on YT. At 84 and 65 mbit, the VMAF scores of the videos from YT are practically identical 91.4 to 92.26. Uploading a 700 mbit DNxHR is really the only significant difference at 95.35


At low bitrates, 20 and 9mbit, AV1 does give you better quality. But if your goal is the best quality on YT, you should not be uploading a 20mbit file.


On my machine AV1 is 1 fps vs 90fps for h.265 Nvenc, so the answer is clear. If you can encode to AV1 at reasonable speed, there is no negative to it, just no significant upside either.

I did more testing with the type of content that I typically shoot (big events, promo videos, music videos, etc.) so I tested an H.265 video in an MOV container at a constant bitrate of 16Mb/s for a 4K30FPS project and Linear PCM for audio.

The video was 15min long and it took YT almost 2hrs to process it and I did not see a quality difference. I agree with @rcmodelreviews if you can't tell the difference then maybe it is not there. That is not to say that if your project includes a lot of fast motion or complex lighting that maybe different export settings would be better; so I think the best export settings are also dependent on the type of project.

I get no artifacts with my 4K30FPS bitrate set to only 16Mb/s for my typical project and a long time ago I did extensive testing to find the best tradeoff between size and quality. Even with complex transitions, whip pans, speed ramps, etc. I don't see a quality difference or the need to use a higher bitrate or different codec on YT.

The one exception is audio; I switch to Quicktime and Linear PCM for audio when shooting music videos or something with a complex audio track because DR has a known audio bug when exporting AAC audio for certain frequencies.

Personally, YT processing speeds do matter to me because I frequently upload previews for clients and time is money; the longer it takes YT to process the preview the longer I have to spend waiting for the processing to complete before I can send the link to the client. So for me, Google's recommended settings are in fact the best export settings and they represent the best tradeoff between time, file size, and YT quality unless DR's AAC bug rears its ugly head; something which DR's team is hopefully going to fix soon.

When I did my testing years ago the biggest quality difference that I saw was you pretty much never want to upload anything below 2K, even if the source project is 1080P, upscale it to 2K just to get YT to transcode it at higher quality.

Are you kidding me? AVI in 2023?

Nope, not kidding.

The quality will be excellent.

No it won't, YT will still transcode it using their VP9 codec and after waiting hours for that massive file to upload and hours more for it to transcode it will look exactly like an MP4 or MOV upload that took a fraction as long.

...
I get no artifacts with my 4K30FPS bitrate set to only 16Mb/s for my typical project and a long time ago I did extensive testing to find the best tradeoff between size and quality. Even with complex transitions, whip pans, speed ramps, etc. I don't see a quality difference or the need to use a higher bitrate or different codec on YT.
...
When I did my testing years ago the biggest quality difference that I saw was you pretty much never want to upload anything below 2K, even if the source project is 1080P, upscale it to 2K just to get YT to transcode it at higher quality.

Those are two very useful bits of info. I may do some tests restricting my bitrate to see what I get. May end up saving some storage space. Those 4k60 files are quite large even when using AV1.

It's also good to know it wasn't just my imagination that my YT shorts actually did look better when uploading above 1080p...even when playing back at 1080p.

I did more testing with the type of content that I typically shoot (big events, promo videos, music videos, etc.) so I tested an H.265 video in an MOV container at a constant bitrate of 16Mb/s for a 4K30FPS project and Linear PCM for audio.

The video was 15min long and it took YT almost 2hrs to process it

That's odd.

I create a lot of videos which are around the 20 minute mark and render out at UHD/50FPS. These tend to come to around 5-6GB in size using H265 and YT normally renders them to 4K in less than 15 minutes.

Are you setting the Network flag on the encoder so that YT can start encoding while the file is still uploading instead of having to wait until it has the entire file?

I did more testing with the type of content that I typically shoot (big events, promo videos, music videos, etc.) so I tested an H.265 video in an MOV container at a constant bitrate of 16Mb/s for a 4K30FPS project and Linear PCM for audio.

The video was 15min long and it took YT almost 2hrs to process it and I did not see a quality difference. I agree with @rcmodelreviews if you can't tell the difference then maybe it is not there.

Did you do it more than once, maybe on a different day, YT can do those slow encodes on MP4 too, when I notice it I"ll upload again, and often the new upload will process normally and be encoded before the original . I have not tried .mov, but have tried H.265 as mp4 and mkv. The H.265 MP4 saw no processing penalty as compared to H.264 version, the H.265 MKV processed faster than both H.264 and H.265 MP4. It's possible the MKV was encoding with fast start while the MP4's weren't, but also maybe it wasn't and MKV does process faster. I'd have to test again.

I did more testing with the type of content that I typically shoot (big events, promo videos, music videos, etc.) so I tested an H.265 video in an MOV container at a constant bitrate of 16Mb/s for a 4K30FPS project and Linear PCM for audio.

The video was 15min long and it took YT almost 2hrs to process it

That's odd.

I create a lot of videos which are around the 20 minute mark and render out at UHD/50FPS. These tend to come to around 5-6GB in size using H265 and YT normally renders them to 4K in less than 15 minutes.

Are you setting the Network flag on the encoder so that YT can start encoding while the file is still uploading instead of having to wait until it has the entire file?

I have not tried setting the network encoding flag but the file was only 2.4GB and it took about 2hrs. The same type of project and file size using H.264 takes around 30-40min. I am not sure if that flag would help, it might save a few min since the process starts sooner, but I don't see it going from 2hrs to under 15min.

In fact I don't think I've ever seen YT render a 20min file even with H.264 in under 30-40min.

I did more testing with the type of content that I typically shoot (big events, promo videos, music videos, etc.) so I tested an H.265 video in an MOV container at a constant bitrate of 16Mb/s for a 4K30FPS project and Linear PCM for audio.

The video was 15min long and it took YT almost 2hrs to process it and I did not see a quality difference. I agree with @rcmodelreviews if you can't tell the difference then maybe it is not there.

Did you do it more than once, maybe on a different day, YT can do those slow encodes on MP4 too, when I notice it I"ll upload again, and often the new upload will process normally and be encoded before the original . I have not tried .mov, but have tried H.265 as mp4 and mkv. The H.265 MP4 saw no processing penalty as compared to H.264 version, the H.265 MKV processed faster than both H.264 and H.265 MP4. It's possible the MKV was encoding with fast start while the MP4's weren't, but also maybe it wasn't and MKV does process faster. I'd have to test again.

I only tried it once, I was uploading it to a client's YT channel so not really the best time to perform a lot of testing. I will do more testing on my own channel. I will probably just stick with MP4 and H.264, sometimes I have to send the footage directly to clients, other times I have to mail them a USB drive, etc. and H.265 can be very problematic for playback on your typical home PC. For me its best to just have a single export preset for all of my projects.

If I saw a visible quality increase on YT when comparing H.264 vs H.265 then I would probably add an additional YT render preset, but I just don't see it. AV1 looks interesting but I don't have a video card that supports it.

Absolutely right to export to DNx / prores, not only for the less compressed video, but also audio. I found a large audio increase when exporting 24bit over 16. I'm sure youtube compresses it to 16 bit, but the difference was massive. Some guys have more knowledge about the OPUS audio codec youtube uses, and found the audio quality to also have better quality at -16 LUFS (?) not sure if I remember correctly.

Why the big need for quality audio and video when uploading to YT/Vimeo.

Virtually all the people who watch this content do so on their phone or a desktop PC with a crappy non-calibrated monitor and $65 speakers!

Because we make our movies for the 0.5% with better equipment

Why the big need for quality audio and video when uploading to YT/Vimeo.

Virtually all the people who watch this content do so on their phone or a desktop PC with a crappy non-calibrated monitor and $65 speakers!

Yeah, but consumer electronics are advancing rapidly. A simple look at videos people uploaded 10 years ago and you will understand. You don't want to be the guy with a channel consisting of mostly crap-quality videos. Heh...this is why I started doing everything in 4k. Most people don't even have access to 4k right now...but they will.

This is very valuable information, thank you to everyone that has posted

Why the big need for quality audio and video when uploading to YT/Vimeo.

Virtually all the people who watch this content do so on their phone or a desktop PC with a crappy non-calibrated monitor and $65 speakers!

In recent years, more and more brands have started comissioning films for their brands. Take a look at some of the documentaries that Patagonia has posted for example. The quality is actually really high, and it's your job as an editor to provide the highest quality sample. I've seen work on youtube that is much higher quality than some of the stuff on Netflix.

Why the big need for quality audio and video when uploading to YT/Vimeo.

Virtually all the people who watch this content do so on their phone or a desktop PC with a crappy non-calibrated monitor and $65 speakers!

Because I take pride in making the video and audio as good as I can for my YT channel. For those watching on a phone or crappy PC it's their loss.
Like Uli I cater for that 0.5% as well.

Why the big need for quality audio and video when uploading to YT/Vimeo.

Virtually all the people who watch this content do so on their phone or a desktop PC with a crappy non-calibrated monitor and $65 speakers!

I would change my export settings if I saw a noticeable quality difference, but for my content I simply do not. I do believe in getting the highest YT quality that makes sense though, but I use YT only to show future clients past work and most of my future clients are going to use desktop computers to see the footage while making a hiring decision so in my particular use case, quality is important. With that being said, some of the export recommendations in this thread make no sense, but that's just me.

I think H.264 and H.265 are reasonable export options, but anything beyond that just results in massive files, upload times, and processing times with no discernable quality increase.

Why the big need for quality audio and video when uploading to YT/Vimeo.

Virtually all the people who watch this content do so on their phone or a desktop PC with a crappy non-calibrated monitor and $65 speakers!

Because I take pride in making the video and audio as good as I can for my YT channel. For those watching on a phone or crappy PC it's their loss.
Like Uli I cater for that 0.5% as well.

I cater to future clients so for me quality is particularly important but I just don't see a difference between H.264 and H.265. The only real differences I see is when uploading 4K vs 1080P and Linear PCM vs AAC.

Why the big need for quality audio and video when uploading to YT/Vimeo.

Virtually all the people who watch this content do so on their phone or a desktop PC with a crappy non-calibrated monitor and $65 speakers!

The break down depends greatly on the content of your channel.

by watch time viewership on my channel is.

tv - 35.6%
pc - 34.3%
phone - 20.5%
tablet - 9.6%

The only real differences I see is when uploading 4K vs 1080P and Linear PCM vs AAC.

It seems that 2K is enough to trigger better encodes. Maybe even a tad over HD, but in 16:9, would suffice. Might be worth a try.

The only real differences I see is when uploading 4K vs 1080P and Linear PCM vs AAC.

It seems that 2K is enough to trigger better encodes.

Yes, when I upload 1440p footage YT encodes it with VP9.

The only real differences I see is when uploading 4K vs 1080P and Linear PCM vs AAC.

It seems that 2K is enough to trigger better encodes. Maybe even a tad over HD, but in 16:9, would suffice. Might be worth a try.

I agree, I used to shoot everything in 1080P and upscale it to 2K to get the resolution bump. But now all of my cameras shoot between 4K and as high as 8K so I just upload everything in 4K now. Another little known fact about YT is they will also rank your content higher in search results if it is 4K, not sure about 2K but definitely for 4K they will rank it higher.

I worked with a content producer with close to 1M followers and on one of the calls with a YT account manager they mentioned that Google does look at resolution as well when determining rank for search results. Comments also increase a video's ranking so if growing your YT channel is your thing, you should reply to every single comment whenever possible.

Echoing Dan Sherman, my YT viewing by device is strong in the PC and TV sections as well.

I just tested a few 4K renders for youtube to compare quality pre-upload for an animation. I did;

Resolves stock Youtube setting
Youtube recommended H264
Voukouder AV1
High Bitrate H265

The resolve stock YT setting is terrible. Huge, blocky artifacts and massive loss in detail. Best was the recommended Youtube and the AV1. Basically the same. The H265 was fine too, but did vary from the main edit more than the others.

So i've just made a youtube preset as close to their recommendations as possible.

I just tested a few 4K renders for youtube to compare quality pre-upload for an animation. I did;

Resolves stock Youtube setting
Youtube recommended H264
Voukouder AV1
High Bitrate H265

The resolve stock YT setting is terrible. Huge, blocky artifacts and massive loss in detail. Best was the recommended Youtube and the AV1. Basically the same. The H265 was fine too, but did vary from the main edit more than the others.

So i've just made a youtube preset as close to their recommendations as possible.

Is your assessment before or after the uploading and re-encoding by youtube? Thanks!

Most likely a very dumb question, but....

looking through this thread I noticed, nobody seems to use the presets for YT or Vimeo to render the projects.

So question #1 would be: Why not?
Question #2 would be: If the presets are crap (please excuse my french here), why dosn`t BM build the presets to better delivery settings or at least offer more option within the presets to customize them?

I noticed, in example, that the subtitle options in the YT preset have been removed, why?

Yes it seems ridiculous that the subtitles option has been removed from the You Tube delivery preset, but the new "Audio Normalisation" option is there!

So it seems if you want got take advantage of the new YT audio normalisation option AND choose the option for sub title upload...You can't!

Why Not???

I don't use the YT presets because of their low bit rates much preferring to render to my computer. The Audio Normalization works fine for this too.