Puzzling rendering results

Get answers to your questions about color grading, editing and finishing with DaVinci Resolve.
  • Author
  • Message
Offline

Mario69Rossi

  • Posts: 197
  • Joined: Fri Jan 29, 2021 4:11 am
  • Real Name: Mario Rossi

Puzzling rendering results

PostFri Jan 29, 2021 4:16 am

Admittedly I have a limited video editing knowledge, I'm trying to get educated. Said that, I was trying different settings in order to check what's the best result in term of quality/size. My videos are uploaded to YouTube.

I tried to render to UHD a 1080 timeline (about 10 minutes long) using different settings on the custom tab within the delivery tab (DaVinci Resolve 17.1) however the size is always the same (max 10MB difference):

1. res UHD, format QuickTime, codec H.264 - file size: 4,149,493,301 bytes
2. res UHD, format QuickTime, codec H.265 - file size: 4,158,619,908 bytes
3. res UHD, format mp4, codec H.264 - file size: 4,149,546,122 bytes
4. res UHD, format mp4, codec H.265 - file size: 4,158,617,799 bytes

Looks like format and codec has no impact on the results and actually H.265 results in slightly larger files than H.264. I had set the quality to 50,000KB. Then I tried to render the same timeline using the default YouTube tab with default setting (but they look horrible even on my laptop):

1. res UHD, format QuickTime, codec H.264 - file size: 864,332,591 bytes
2. res UHD, format QuickTime, codec H.265 - file size: 859,953,108 bytes
3. res UHD, format mp4, codec H.264 - file size: 864,323,974 bytes
4. res UHD, format mp4, codec H.265 - file size: 859,946,219 bytes

Why do format have no impact on the size of the rendered file? And codec very very little? Am I doing something wrong?
DaVinci Resolve Studio 19.1
MacBook Pro M1 (2020) | 16GB ram | 1TB SSD
macOS Squoia 15.1
Offline

Hendrik Proosa

  • Posts: 3312
  • Joined: Wed Aug 22, 2012 6:53 am
  • Location: Estonia

Re: Puzzling rendering results

PostFri Jan 29, 2021 8:09 am

If you set the bitrate to be the same, file size will be also very similar.
I do stuff
Offline

Mario Kalogjera

  • Posts: 1202
  • Joined: Sat Oct 31, 2015 8:44 pm

Re: Puzzling rendering results

PostFri Jan 29, 2021 9:03 am

QuickTime and mp4 are the same thing. What's your hardware, are you using native or GPU h264/h265 encoding?

Sent from my Mi 9T using Tapatalk
Asus Prime X370-Pro+R7 3700X@PBO+32 GB G.Skill AEGIS DDR-4@3200MHz
Sapphire RX6700 10GB
Adata A400 120GB System,A2000 500GB Scratch SSDs
Media storage:"Always in motion is it"
BMD Mini Monitor 4K
Windows 11 Pro+Resolve Studio 18+Fusion Studio 18
Offline
User avatar

Uli Plank

  • Posts: 25124
  • Joined: Fri Feb 08, 2013 2:48 am
  • Location: Germany and Indonesia

Re: Puzzling rendering results

PostFri Jan 29, 2021 9:13 am

QuickTime and MP4 are not the same thing. They are closely related, yes. If you are referring to .MOV by "QuickTime", it's just a container (like .MP4), but it can contain many more different codecs. The codec is highly relevant for quality, file sizes, playback speed etc. .
My disaster protection: export a .drp file to a physically separated storage regularly.

Studio 19.1.3
MacOS 13.7.4, 2017 iMac, 32 GB, Radeon Pro 580 + eGPU
MacBook M1 Pro, 16 GPU cores, 32 GB RAM, MacOS 14.7.2
SE, USM G3
Offline

Mario Kalogjera

  • Posts: 1202
  • Joined: Sat Oct 31, 2015 8:44 pm

Re: Puzzling rendering results

PostFri Jan 29, 2021 12:10 pm

Uli Plank wrote:QuickTime and MP4 are not the same thing. They are closely related, yes. If you are referring to .MOV by "QuickTime", it's just a container (like .MP4), but it can contain many more different codecs. The codec is highly relevant for quality, file sizes, playback speed etc. .


Yes, of course, I referred to the container part i.e. OP had made both .mov and .mp4 with h264 and h265 codec but in case of those codecs being in Mov or MP4 makes no difference to his case or size of video itself.

Sent from my Mi 9T using Tapatalk
Asus Prime X370-Pro+R7 3700X@PBO+32 GB G.Skill AEGIS DDR-4@3200MHz
Sapphire RX6700 10GB
Adata A400 120GB System,A2000 500GB Scratch SSDs
Media storage:"Always in motion is it"
BMD Mini Monitor 4K
Windows 11 Pro+Resolve Studio 18+Fusion Studio 18
Offline

Andrew Kolakowski

  • Posts: 9391
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Puzzling rendering results

PostFri Jan 29, 2021 1:45 pm

Mario69Rossi wrote:Admittedly I have a limited video editing knowledge, I'm trying to get educated. Said that, I was trying different settings in order to check what's the best result in term of quality/size. My videos are uploaded to YouTube.

I tried to render to UHD a 1080 timeline (about 10 minutes long) using different settings on the custom tab within the delivery tab (DaVinci Resolve 17.1) however the size is always the same (max 10MB difference):

1. res UHD, format QuickTime, codec H.264 - file size: 4,149,493,301 bytes
2. res UHD, format QuickTime, codec H.265 - file size: 4,158,619,908 bytes
3. res UHD, format mp4, codec H.264 - file size: 4,149,546,122 bytes
4. res UHD, format mp4, codec H.265 - file size: 4,158,617,799 bytes

Looks like format and codec has no impact on the results and actually H.265 results in slightly larger files than H.264. I had set the quality to 50,000KB. Then I tried to render the same timeline using the default YouTube tab with default setting (but they look horrible even on my laptop):

1. res UHD, format QuickTime, codec H.264 - file size: 864,332,591 bytes
2. res UHD, format QuickTime, codec H.265 - file size: 859,953,108 bytes
3. res UHD, format mp4, codec H.264 - file size: 864,323,974 bytes
4. res UHD, format mp4, codec H.265 - file size: 859,946,219 bytes

Why do format have no impact on the size of the rendered file? And codec very very little? Am I doing something wrong?


Size= bitrate*duration (+- small overshoot depending on the encoder). There is no puzzle at all. You used same bitrate for same source, so your end file sizes are the same.
It doesn't mean they are the same quality, which is probably what you had in mind (thought better codec will make file smaller). In bitrate mode it doesn't work this way. You have to set smaller bitrate.
If you tried to achieve same quality with h264 and h265 then 2nd one would be smaller (eg h265 with 35Mbit vs. h264 with 50Mbit). Visual quality measure is not that easy though. You have few metrics which could be used ( check here: https://www.streamingmedia.com/Articles ... _selection) or simply rely on your eyes and knowledge.
Offline

Mario69Rossi

  • Posts: 197
  • Joined: Fri Jan 29, 2021 4:11 am
  • Real Name: Mario Rossi

Re: Puzzling rendering results

PostSat Jan 30, 2021 1:06 am

Mario Kalogjera wrote:QuickTime and mp4 are the same thing. What's your hardware, are you using native or GPU h264/h265 encoding?

I'm using apple M1
DaVinci Resolve Studio 19.1
MacBook Pro M1 (2020) | 16GB ram | 1TB SSD
macOS Squoia 15.1
Offline

Mario69Rossi

  • Posts: 197
  • Joined: Fri Jan 29, 2021 4:11 am
  • Real Name: Mario Rossi

Re: Puzzling rendering results

PostSat Jan 30, 2021 1:22 am

Andrew Kolakowski wrote:Size= bitrate*duration (+- small overshoot depending on the encoder). There is no puzzle at all. You used same bitrate for same source, so your end file sizes are the same.
It doesn't mean they are the same quality, which is probably what you had in mind (thought better codec will make file smaller). In bitrate mode it doesn't work this way. You have to set smaller bitrate.
If you tried to achieve same quality with h264 and h265 then 2nd one would be smaller (eg h265 with 35Mbit vs. h264 with 50Mbit). Visual quality measure is not that easy though. You have few metrics which could be used ( check here: link removed as I cannot post links) or simply rely on your eyes and knowledge.


Okay I understand your equation size=bitrate*duration. Not sure why people say a format like mp4 results in smaller file sizes than QuickTime and the same for H.265 vs H.264. From what you say those formats and codecs gives you the ability to play around with the bitrate and find a balance where you end up with a same quality with a smaller file size because the format/codec combination is more efficient.

I guess this is going to require lots of tests. However I'm surprised how the default rendering settings for youtube results in a great file size but the quality is awful. In the video mentioned I ended up using handbreak to further compress the file at an acceptable file size. I have not the fastest internet connection and having to upload 8GB files for 20 min video is just insane. I usually have to upload in three languages so it's over 20GB that should be uploaded at once. Handbreak cut that in half.

When I was using Adobe Premiere using the default rendering settings for YouTube resulted in a pretty good quality, I'm wondering if I messed it up somewhere or you guy experience similar results using default YouTube setting in Davinci Resolve.
DaVinci Resolve Studio 19.1
MacBook Pro M1 (2020) | 16GB ram | 1TB SSD
macOS Squoia 15.1
Offline

Jim Simon

  • Posts: 34455
  • Joined: Fri Dec 23, 2016 1:47 am

Re: Puzzling rendering results

PostMon Feb 01, 2021 3:09 pm

I prefer to upload Cineform files. Those things are huge!

But...I'm fortunate to have a Synchronous 1GB Fiber connection. :)
My Biases:

You NEED training.
You NEED a desktop.
You NEED a calibrated (non-computer) display.
Offline

Andrew Kolakowski

  • Posts: 9391
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Puzzling rendering results

PostMon Feb 01, 2021 3:21 pm

Mario69Rossi wrote:
I guess this is going to require lots of tests. However I'm surprised how the default rendering settings for youtube results in a great file size but the quality is awful. In the video mentioned I ended up using handbreak to further compress the file at an acceptable file size. I have not the fastest internet connection and having to upload 8GB files for 20 min video is just insane. I usually have to upload in three languages so it's over 20GB that should be uploaded at once. Handbreak cut that in half.

When I was using Adobe Premiere using the default rendering settings for YouTube resulted in a pretty good quality, I'm wondering if I messed it up somewhere or you guy experience similar results using default YouTube setting in Davinci Resolve.


Premiere has better h264 encoder (based on Mainconcept SDK). Resolve uses GPU one (fast but never really good) or OS provided one. If you want optimal final h264/5 files don't use Resolve, but as you did eg. Handbrake (or whatever x264/5 powered). This applies to you even more due to need of smaller files.
Offline

Andrew Kolakowski

  • Posts: 9391
  • Joined: Tue Sep 11, 2012 10:20 am
  • Location: Poland

Re: Puzzling rendering results

PostMon Feb 01, 2021 3:26 pm

Mario69Rossi wrote:Okay I understand your equation size=bitrate*duration. Not sure why people say a format like mp4 results in smaller file sizes than QuickTime and the same for H.265 vs H.264. From what you say those formats and codecs gives you the ability to play around with the bitrate and find a balance where you end up with a same quality with a smaller file size because the format/codec combination is more efficient.


Because they have no idea what they are talking about (like 95% of youtube 'experts')?
Containers are rather meaningless when it comes to quality, specially when we talk about ones which can work with same codecs.
Last edited by Andrew Kolakowski on Mon Feb 01, 2021 5:19 pm, edited 1 time in total.
Offline

Jim Simon

  • Posts: 34455
  • Joined: Fri Dec 23, 2016 1:47 am

Re: Puzzling rendering results

PostMon Feb 01, 2021 4:18 pm

Andrew Kolakowski wrote:Premiere has better h264 encoder (based on Mainconcept SDK).
:lol:

Thanks, Andrew. I needed that.
My Biases:

You NEED training.
You NEED a desktop.
You NEED a calibrated (non-computer) display.
Offline

Mario69Rossi

  • Posts: 197
  • Joined: Fri Jan 29, 2021 4:11 am
  • Real Name: Mario Rossi

Re: Puzzling rendering results

PostThu Feb 04, 2021 4:33 am

Thank you everybody for the explanation. I'm trying to find the right balance through trial and error.
DaVinci Resolve Studio 19.1
MacBook Pro M1 (2020) | 16GB ram | 1TB SSD
macOS Squoia 15.1

Return to DaVinci Resolve

Who is online

Users browsing this forum: Bing [Bot], Fully_Loaded34, Mads Johansen, qsipher, Robert John and 261 guests