I have played around (both using Videostudio and TmpGenc) with various bit rate settings and constant and variable bit rate setting and while it is easy to determine that changing the bit rate affects quality but I cannot determine if quality is affected between constant and variable.
Could someone please clear this up for me. It is obvious that using variable results in a (rendered) smaller file file than when using constant so why wouldn't one use variable all the time. Is there a preferred method?
Also Two Pass encode (when using variable) takes longer to render but presumably results in higher quality - is that correct? Should two pass encode always be used when using variable? Can the difference be seen on a DVD player (connected to a TV)?
Thanks
John
Rendering - constant or variable bit rate - which is best?
Moderator: Ken Berry
- jparnold
- Advisor
- Posts: 991
- Joined: Sat Dec 17, 2005 10:45 am
- System_Drive: C
- 32bit or 64bit: 64 Bit
- motherboard: Gigabyte Z390 UD
- processor: Intel Pentium i7-9700 3.6Ghz
- ram: 16GB DDR3
- Video Card: Gigabyte RTX2060 OC 6GB
- sound_card: Onboard Realtec ALC887
- Hard_Drive_Capacity: 2048Gb mix
- Monitor/Display Make & Model: Samsung S27C450B
- Corel programs: Videostudio X10, Paint Shop Pro X8
- Location: Blue Mountains, NSW, Australia
Rendering - constant or variable bit rate - which is best?
John a
VS X10 Ultimate, Paint Shop Pro X8, Audacity, Panasonic HC-X920M, Nikon Coolpix S8100
VS X10 Ultimate, Paint Shop Pro X8, Audacity, Panasonic HC-X920M, Nikon Coolpix S8100
-
- Moderator
- Posts: 1126
- Joined: Sat Dec 11, 2004 6:34 am
- System_Drive: C
- 32bit or 64bit: 64 Bit
- motherboard: eMachines ET1861
- processor: 3.20 gigahertz Intel Core i5 650
- ram: 12GB
- Video Card: ATI Radeon HD 5400 Series
- sound_card: ATI High Definition Audio Device
- Hard_Drive_Capacity: 700GB
- Location: Taipei, Taiwan
- Contact:
"using variable results in a (rendered) smaller file file than when using constant"
You have sort of answered your own question there. If you have a project you need to get on to a disc and the project is going to come close to the maximum size of the disc then two-pass variable is a good way to allow the codec to apportion the maximum amount of data to high motion scenes and save a bit on low motion areas.
Using a single pass variable always seemed a bit pointless to me as the codec is being asked to estimate the motion on the fly.
With two-pass the program scans through the file once to mark the high and low motion areas and on the second pass the encoding is actually done.
If, however, your project is relatively small and disc real estate is not a problem then a constant bitrate up around 6000 to 8000kps is going to give you a larger file (but you don't care cos' you have acres of space!) at top quality.
You have sort of answered your own question there. If you have a project you need to get on to a disc and the project is going to come close to the maximum size of the disc then two-pass variable is a good way to allow the codec to apportion the maximum amount of data to high motion scenes and save a bit on low motion areas.
Using a single pass variable always seemed a bit pointless to me as the codec is being asked to estimate the motion on the fly.
With two-pass the program scans through the file once to mark the high and low motion areas and on the second pass the encoding is actually done.
If, however, your project is relatively small and disc real estate is not a problem then a constant bitrate up around 6000 to 8000kps is going to give you a larger file (but you don't care cos' you have acres of space!) at top quality.
The main difference between VBR and CBR is that CBR is variable quality at a given fixed bitrate, whereas VBR gives constant quality at a variable bitrate.
Basically, VBR uses more bits to represent complex scenes with a
lot of motion and saves space on static scenes which don't change much.
Therefore, conventional wisdom is that it's generally preferable to use VBR regardless of the file size.
VBR often requires two passes, but with the speed of modern computers this is not a problem. I use VBR for everything.
Basically, VBR uses more bits to represent complex scenes with a
lot of motion and saves space on static scenes which don't change much.
Therefore, conventional wisdom is that it's generally preferable to use VBR regardless of the file size.
VBR often requires two passes, but with the speed of modern computers this is not a problem. I use VBR for everything.
Terry
Hi
I use constant bit rate for all my video.
As a guide if I select 6000kbps for example, at Constant all the video will be coded at 6000.
If I select Variable 6000 this is denoted as Max 6000. Some but not all the video will be coded at 6000, the majority will be reduced to a lesser bit rate, even further for the motionless frames. The average bit rate being about 5000 + at a guess.
I like the idea and concept of variable, It creates a smaller file size allowing you to produce a longer video. But that’s the only advantage I can see. I cannot see that quality is better.
See Bit Rates & File Sizes from the link below
Trevor
I use constant bit rate for all my video.
As a guide if I select 6000kbps for example, at Constant all the video will be coded at 6000.
If I select Variable 6000 this is denoted as Max 6000. Some but not all the video will be coded at 6000, the majority will be reduced to a lesser bit rate, even further for the motionless frames. The average bit rate being about 5000 + at a guess.
I like the idea and concept of variable, It creates a smaller file size allowing you to produce a longer video. But that’s the only advantage I can see. I cannot see that quality is better.
See Bit Rates & File Sizes from the link below
Trevor
-
- Microsoft MVP
- Posts: 2659
- Joined: Thu Jul 21, 2005 10:09 am
- System_Drive: C
- 32bit or 64bit: 64 Bit
- motherboard: gigabyte
- processor: Intel core 2 6420 2.13GHz
- ram: 4GB
- Video Card: NVidia GForce 8500GT
- sound_card: onboard
- Hard_Drive_Capacity: 36GB 2TB
- Location: UK
Just to throw in a spanner....
This can differ with which encoder you use as ther are different types of constant, where one isn't a true constant, but it's a constant that must be met for every 5 second chunk, if I remember right. So any 5 seconds you take the average bit rate for that time period will be 6,000. So in that time if there is some motion or some still the bitrate per frame will alter slightly.
The best thing is to encode each movie in accordance with it's needs, taking into account quality/time/space etc...
The biggest difference will probably be in the quality of the encoder used
This can differ with which encoder you use as ther are different types of constant, where one isn't a true constant, but it's a constant that must be met for every 5 second chunk, if I remember right. So any 5 seconds you take the average bit rate for that time period will be 6,000. So in that time if there is some motion or some still the bitrate per frame will alter slightly.
The best thing is to encode each movie in accordance with it's needs, taking into account quality/time/space etc...
The biggest difference will probably be in the quality of the encoder used
