mirror of
https://github.com/HandBrake/HandBrake-docs.git
synced 2025-12-10 00:44:52 -06:00
127 lines
5.5 KiB
Markdown
127 lines
5.5 KiB
Markdown
---
|
|
Type: article
|
|
State: [ draft ]
|
|
Title: Constant Quality vs Average Bit Rate
|
|
Project: HandBrake
|
|
Project_URL: https://handbrake.fr/
|
|
Project_Version: 1.7.0
|
|
Language: English
|
|
Language_Code: en
|
|
Authors: [ Scott (s55) ]
|
|
Copyright: 2023 HandBrake Team
|
|
License: Creative Commons Attribution-ShareAlike 4.0 International
|
|
License_Abbr: CC BY-SA 4.0
|
|
License_URL: https://handbrake.fr/docs/license.html
|
|
---
|
|
|
|
Constant Quality vs Average Bit Rate
|
|
====================================
|
|
|
|
HandBrake supports two modes of encoding video.
|
|
|
|
- With the average bitrate, you control the size of the output file but give
|
|
up control over the video's quality.
|
|
- Constant quality mode does the opposite; you specify a quality level and
|
|
HandBrake adjusts the bitrate (that is, the size) to meet it.
|
|
|
|
## Constant Quality
|
|
|
|
### Benefits over Average Bitrate
|
|
|
|
- Faster Encoding - No need to do 2-Pass encoding (thus almost halving the
|
|
encode time.) Constant Quality can produce the same quality as a 2 pass
|
|
encode.
|
|
|
|
- Reduced file sizes Videos usually have a mix of complex and less complex
|
|
frames. (The latter requiring less bitrate to achieve a set quality level)
|
|
If less complex frames do not have bits wasted on them, your overall output
|
|
filesize is reduced without any loss of quality.
|
|
|
|
### Output File Sizes
|
|
|
|
Because it takes a different bitrate to reach a given quality level for any
|
|
given part of any movie, the output size is unpredictable and varies from source
|
|
to source.
|
|
|
|
Example: Given 2 sources of the same length, one could turn out to be 1.2GB, the
|
|
other 1.9GB. The difference can be more or less depending on the source. If your
|
|
source material is consistent, you will get roughly the same file sizes out for
|
|
each.
|
|
|
|
In particular, grainy sources tend to come out larger as they require a higher
|
|
bit-rate to maintain all that extra detail in the video. In this case, you may
|
|
try turning on the "Denoise" filter or reduce the RF value a few points.
|
|
|
|
You should also note that even if a file size is significantly smaller than a
|
|
counterpart encoded with target filesize, it does not mean it is any lower
|
|
quality. It's simply encoded more efficiently.
|
|
|
|
### x264 Recommended Quality RF Values
|
|
|
|
The Quality slider ranges from (Low Quality) 51 to 0 (High Quality) For good
|
|
quality output at reasonable file sizes, the following values are recommended:
|
|
|
|
Recommended settings for x264 are:
|
|
|
|
- Standard Definition (e.g DVD's) Use an RF value of around 20 +/- 1 As an
|
|
example using the AppleTV2 preset at RF20, with 20 different sources, the
|
|
average size was 925MB per hour of video. (Min: 625MB/h Max:1,503MB/hr)
|
|
|
|
- High Definition (e.g Blurays 720/1080) Use an RF value of 22 +/- 1 Since HD
|
|
sources are typically quality, you can get away with a slightly higher RF
|
|
value than SD content without any perceived difference in quality.
|
|
|
|
You should note that the quality slider is not linear for x264. It is
|
|
logarithmic (like the Richter or pH scales.) This means small movements in the
|
|
numbers can have large results. You can control the granularity of the slider in
|
|
the preferences / options.
|
|
|
|
You can encode a few chapters of a source to experiment with different quality
|
|
values if the above recommendations are not suitable. Encoding a 10minute sample
|
|
in the middle of your source should give you a good idea of what to expect.
|
|
|
|
### x264 and RF 0
|
|
|
|
RF 0 applies no compression. It is lossless: it compresses the source without
|
|
throwing away practically any detail.
|
|
|
|
So should you use RF 0 to perfectly preserve the source? Nope. Not at all. In
|
|
fact, you'll end up with video that's way larger than the DVD, but doesn't look
|
|
any better.
|
|
|
|
See, DVDs use lossy compression to squeeze down the raw video the studios use to
|
|
make them -- sort of like a quality level of RF 20. It throws away detail. When
|
|
HandBrake uncompresses the video prior to conversion, the quality lost when the
|
|
DVD was made is still gone. When you use RF 0 quality with x264, you're telling
|
|
it to losslessly preserve the decoded, uncompressed video feed, not to
|
|
losslessly preserve the DVD. Both have the same picture quality, but the
|
|
uncompressed feed takes up a lot more space.
|
|
|
|
To sum up: when converting from a DVD source, there is no reason to go above an
|
|
RF of \~19, which is roughly equivalent to how heavily the DVD is compressed. If
|
|
you do go higher, your output will be larger than your input!
|
|
|
|
### Recommended values for other encoders
|
|
|
|
HandBrake will default to a sensible default when the encoder is changed. As with x264
|
|
small adjustments can be made to achieve the desired result.
|
|
|
|
## Average Bitrate
|
|
|
|
Unless you really need to aim for a target filesize (which we recommend
|
|
against), it is highly recommended that you use Constant Quality.
|
|
|
|
Bitrate is the number of bits dedicated to the video in a second. And remember,
|
|
that gets split between \~24-30 frames. When you set an average bitrate, the
|
|
encoder will vary the number of bits given to any one portion of the video, but
|
|
try to keep everything at the average you set.
|
|
|
|
- The video bitrate can be set on the "Video Tab"
|
|
- It is recommended that you use 2-Pass encoding.
|
|
- This will provide better quality and better bit-rate distribution. A single
|
|
pass will lead to sub-optimal results.
|
|
- The "Turbo" checkbox will make the first pass quicker at the expense of
|
|
quality. Most of the time this loss is minimal and not appreciable by
|
|
the viewer.
|
|
- The downside is it can take up to twice as long to encode a file.
|