-
1. Re: File Size
John T Smith Jun 14, 2012 4:57 PM (in response to AtlantaVideo)2 pass encoding may be smaller than 1 pass
-
2. Re: File Size
AtlantaVideo Jun 14, 2012 5:53 PM (in response to John T Smith)Which is better in terms of quality?
-
3. Re: File Size
Stan Jones Jun 14, 2012 7:06 PM (in response to AtlantaVideo)The devil is in the details.
low bit rate setting
...
Medium bit rateThe target rate (the average rate) times the length is the file size. I assume, from your file sizes, that the target rate of the later encode is lower than the first.
I don't believe that a two pass encode is a different size than a one-pass.
-
4. Re: File Size
AtlantaVideo Jun 14, 2012 8:14 PM (in response to Stan Jones)The one pass target was 20 and the two pass had a target rate of 22.5.
-
5. Re: File Size
Neil Wilkes Jun 18, 2012 5:01 AM (in response to AtlantaVideo)AtlantaVideo wrote:
The one pass target was 20 and the two pass had a target rate of 22.5.
What was the minimum rate?
Using a CBR (or Constant Quality) encoding will give you a very predictable file size - just do the math, and you will know exactly what you will get.
With a VBR file, it is impossible to know as there are 2 variables - the minimum rate to encode to & the maximum rate to encode to.
With VBR, the bit rate is varied to save bits when they are not needed for when they are needed. So, for example you may have a scene where there is very little motion or action happening. This will need far fewer bits to "describe" what is happening compared to a scene where there is lots of stuff happening and/or complex-to-encode sections like rapid motion, explosions, fires etc. so it is indeed quite probable that a VBR encoded film will use fewer bits with a higher target rate than a CBR will - it will attempt to average this target rate, but there will be sections where it is lower than the target and others where it is higher.
-
6. Re: File Size
ram8kumar Jun 18, 2012 5:27 AM (in response to AtlantaVideo)According to me with VBR-1 Pass, the encoder doesn't have the advance knowledge of what is coming up in the stream. So it simply encodes what comes into its way. So sometimes it even might jump out of the maximum bit rate set if the frame does not fit in the specified range.
But VBR-2 pass allows the encoder make a purely explorately pass over the video to decide on its encoding scheme. So in a way, this is the more intelligent way the encoder will use the bits. It might use less bits for somewhat less moving parts and preserve the bits for that particular frame where the bit rate jumped out of the range. And sometime, it might even end up using lesser bits than that of VBR-1 pass.
-Ramesh.
-
7. Re: File Size
Neil Wilkes Jun 18, 2012 5:59 AM (in response to ram8kumar)Hi Ramesh.
Well pointed out about the VBR 1-pass, which I have to admit I always thought a pretty pointless option!
The 2 big issues I have with this method is firstly the one you point out (it may well exceed maximum bitrate - and by quite a way) and thesecond one is that for a 1-pass operation you will actually be better off using CBR or CQ modes instead.




