CODEC questions

If you would like to post, you'll need to register. Note that if you have a BCG store account, you'll need a new, separate account here (we keep the two sites separate for security purposes).

jeffnles1

jeffnles1

Well-known member
Supporting Member
Marketplace
Member's gallery
OK,
I think I understand the container thing etc. My question is around the difference between different formats.
ProRes, H265, H264, FFV1, V9, QuickTime V210, etc.
When would I export to one vs. another?

I've been shooting video for a couple years but mostly just bringing them into iMovie, splicing them together in the sequence I want, adding some voice over and then exporting in whatever format they were broth in (4K usually).

I'm rather new to editing video. I do have Topaz Video AI that I've been looking at mainly for adding some stabilization (shoot a lot hand held not on tripod). It does a decent job but exporting is excruciatingly slow (2 hours for a 5 minute video). I'm using a MacBook Pro, M1 chip, 16gb ram and working from internal drive. Still slow.

so, back to the original question, when would a person use one of these codecs vs. another? What impact on export (rendering?) speed would it have?

Thanks in advance,

Jeff
 
Member's gallery
Jeff,

You may already know this, but to establish a baseline, codec usage can be broken down into 4 main categories...

Capture codecs - Used in cameras as appropriate for quality and storage efficiency - H.265, H.264, ProRes, camera specific RAW's

Intermediate codecs - High quality, low compression for easy handling by NLE. - ProRes in various flavors, DNxXX in various flavors...often used for "Proxy" files because even though they're large files, the edit machines can handle them easily because they're minimally compressed.

Distribution - High compression, quality appropriate for usage - H.264 H.265

Archival - FFV, DNx in various flavors, ProRes in various flavors, etc.

For file sharing and uploading to social media sites, you can pretty much stick to H.265 or H.264. When you upload to any site that is going to stream it for you, H.264 and H.265 are accepted pretty much everywhere and I can't tell you the last time I exported to H.264. Well, actually just did an export to compare file sizes on a 35 second clip, but more about that in a moment. Now there are some options within both those H.XXX codecs...bit depth (8, 10, 12 bit), color encoding (4-2-0, 4-2-2, 4-4-4) but for the most part, 4-2-0 is fine for non-critical viewing.

The only advantage of H.264 over H.265 is that some lower end laptops and PC's might struggle with playback of 4K H.265 because of it's increased decoding demands on the playback device. Newer systems, both Mac and PC's have hardware decoding for H.265, so will be fine...no issues with late model cell phones or TV's either.

As for that 4K NRaw, 60fps, HDR, 35 sec clip I rendered...rendered at a "Best" quality preset...so high bitrate setting for H.264 and H.265...

H.264 - 531MB (120Mb/s video data rate)
H.265 - 144MB (32.2Mb/s video data rate)
DNxHR - 14.6GB (3494Mb/s video data rate)
FFV RGBA - 18,4GB (4403Mb/s video data rate)

What export codec looks good to you? :)

Cheers!
 
Member's gallery
jeffnles1 said:
OK,
I think I understand the container thing etc. My question is around the difference between different formats.
ProRes, H265, H264, FFV1, V9, QuickTime V210, etc.
When would I export to one vs. another?

I've been shooting video for a couple years but mostly just bringing them into iMovie, splicing them together in the sequence I want, adding some voice over and then exporting in whatever format they were broth in (4K usually).

I'm rather new to editing video. I do have Topaz Video AI that I've been looking at mainly for adding some stabilization (shoot a lot hand held not on tripod). It does a decent job but exporting is excruciatingly slow (2 hours for a 5 minute video). I'm using a MacBook Pro, M1 chip, 16gb ram and working from internal drive. Still slow.

so, back to the original question, when would a person use one of these codecs vs. another? What impact on export (rendering?) speed would it have?

Thanks in advance,

Jeff
Click to expand...

The issue is Topaz, not your editor, computer or output format. Try using the same process and format without it, see what happens. Topaz is very very GPU intensive. I tried it, didn't love the results (visible artifacts for me) and if I need to stabilize in post I use the built-in stabilizer in Premiere Pro of DaVinci.

The output format should be driven on how you plan to display it (YouTube, Insta, TikTok, Facebook, etc.) Each of those will have a specific recommendation for format. There are also a bunch of YouTubers that suggest other format to trick the algorithm for more views.
 
Member's gallery
MotoPixel said:
Jeff,

You may already know this, but to establish a baseline, codec usage can be broken down into 4 main categories...

Capture codecs - Used in cameras as appropriate for quality and storage efficiency - H.265, H.264, ProRes, camera specific RAW's

Intermediate codecs - High quality, low compression for easy handling by NLE. - ProRes in various flavors, DNxXX in various flavors...often used for "Proxy" files because even though they're large files, the edit machines can handle them easily because they're minimally compressed.

Distribution - High compression, quality appropriate for usage - H.264 H.265

Archival - FFV, DNx in various flavors, ProRes in various flavors, etc.

For file sharing and uploading to social media sites, you can pretty much stick to H.265 or H.264. When you upload to any site that is going to stream it for you, H.264 and H.265 are accepted pretty much everywhere and I can't tell you the last time I exported to H.264. Well, actually just did an export to compare file sizes on a 35 second clip, but more about that in a moment. Now there are some options within both those H.XXX codecs...bit depth (8, 10, 12 bit), color encoding (4-2-0, 4-2-2, 4-4-4) but for the most part, 4-2-0 is fine for non-critical viewing.

The only advantage of H.264 over H.265 is that some lower end laptops and PC's might struggle with playback of 4K H.265 because of it's increased decoding demands on the playback device. Newer systems, both Mac and PC's have hardware decoding for H.265, so will be fine...no issues with late model cell phones or TV's either.

As for that 4K NRaw, 60fps, HDR, 35 sec clip I rendered...rendered at a "Best" quality preset...so high bitrate setting for H.264 and H.265...

H.264 - 531MB (120Mb/s video data rate)
H.265 - 144MB (32.2Mb/s video data rate)
DNxHR - 14.6GB (3494Mb/s video data rate)
FFV RGBA - 18,4GB (4403Mb/s video data rate)

What export codec looks good to you? :)

Cheers!
Click to expand...
Thank you. I have seen 4-2-0, and 4-2-1, etc. before. What do the 3 numbers mean? I assume they are some parameters but not sure what each means.

I apologize for asking such basic questions. I’ve been editing photos since the Dawn of digital but up until the past 8 of 9 months video was a curiosity and something I grabbed on my iPhone of cute dogs or silly stunts at family get together events
 
Member's gallery
Nimi said:
The issue is Topaz, not your editor, computer or output format. Try using the same process and format without it, see what happens. Topaz is very very GPU intensive. I tried it, didn't love the results (visible artifacts for me) and if I need to stabilize in post I use the built-in stabilizer in Premiere Pro of DaVinci.

The output format should be driven on how you plan to display it (YouTube, Insta, TikTok, Facebook, etc.) Each of those will have a specific recommendation for format. There are also a bunch of YouTubers that suggest other format to trick the algorithm for more views.
Click to expand...
Hmm that’s interesting. I will have to check that out.
 
Member's gallery
MotoPixel said:
+1 on using built-in stabilizing in Resolve or Premier...though that may be one the items lacking in the Resolve free version.
Click to expand...
I do not have DaVinci or FinalCut yet. Have been using iMovie. As I get slide deeper into the video rabbit hole I will have to get one of the more feature rich editors.
 
Member's gallery
jeffnles1 said:
Thank you. I have seen 4-2-0, and 4-2-1, etc. before. What do the 3 numbers mean? I assume they are some parameters but not sure what each means.

I apologize for asking such basic questions. I’ve been editing photos since the Dawn of digital but up until the past 8 of 9 months video was a curiosity and something I grabbed on my iPhone of cute dogs or silly stunts at family get together events
Click to expand...
You're welcome, Jeff...and no apologies necessary...I'm more than happy to share what I've learned and if I don't know something, I probably know where to find it or can point you to one of the resources I've made use of over the last 7-10 years that I've been getting more serious about video. The capabilities of hybrid cameras today are nothing short of incredible.

Rather than me type it all out, here is a good well-presented overview of what those chroma subsampling numbers mean...essentially it's a form of compression of the color information in a digital image. Jump past the "Test Results" section on the capabilities of various TV's and the explanation and some graphics follow.

https://www.rtings.com/tv/learn/chroma-subsampling

It's a deep rabbit hole, but very rewarding! :)
 
Member's gallery
jeffnles1 said:
I do not have DaVinci or FinalCut yet. Have been using iMovie. As I get slide deeper into the video rabbit hole I will have to get one of the more feature rich editors.
Click to expand...

Per Apple, iMovie has stabe. Never used it, so no idea how well it works.

20240405_072338.jpg
You can only see EXIF info for this image if you are logged in.
 
Member's gallery
MotoPixel said:
You're welcome, Jeff...and no apologies necessary...I'm more than happy to share what I've learned and if I don't know something, I probably know where to find it or can point you to one of the resources I've made use of over the last 7-10 years that I've been getting more serious about video. The capabilities of hybrid cameras today are nothing short of incredible.

Rather than me type it all out, here is a good well-presented overview of what those chroma subsampling numbers mean...essentially it's a form of compression of the color information in a digital image. Jump past the "Test Results" section on the capabilities of various TV's and the explanation and some graphics follow.

https://www.rtings.com/tv/learn/chroma-subsampling

It's a deep rabbit hole, but very rewarding! :)
Click to expand...
Thank you. I will check this out when I get home
 
Member's gallery
uday592 said:
I am planning on getting Topaz video AI. I currently use Adobe p[remiere pro for video editing. Like feedback
Click to expand...
sorry, just now seeing this.
Topaz Video works fine. My biggest problem with it is it takes forever to render videos. For example, I had a 5 minute video 4K. All I had Topaz do was apply stabilization. I wasn't doing any upscale, create slow motion, etc, just stab. It took 2 hours to render a 5 minute video. My laptop is an Apple MacBook Pro, M1 chip, 16gb RAM. Not a powerhouse computer but faster than most of the mid price Windows based machines. Topaz must really hammer the CPU and GPU since after about 20 minutes, the fan on my laptop was running off and on for the next hour and a half. Even with short 30 second type video clips it takes15-20 minutes to render.

Again, the product works and does what it says it will, but be sure your computer has enough processing power (GPU and CPU) to allow the software to perform fast enough to make it usable for you.

Hope this helps.
Jeff
 
Member's gallery
uday592 said:
I am planning on getting Topaz video AI. I currently use Adobe p[remiere pro for video editing. Like feedback
Click to expand...

Topaz video AI works as advertised but it requires lots of processing horsepower or lots of patience. If you are processing 1 second of video at 30fps that is equivalent to processing 30 still images. I minute of video at 30fps contains 1800 images.
 
You must log in or register to reply here.
Back
Top