A few questions about bitrate,codecs and virtualdub

[alibaba12]

hi,im new to the encoding business and would be glad if you help me with few questions ^_^.
ive read few guides from google about bitrate,codecs and virtualdub(as i understand the best program for encoding).

1) i understand that the bitrate set the video quality,but as i saw i can also make video that i film with my iphone and encode it to few gb's ,then the iphone video will have more bitrate but less quality.

2)about audio, mp3 takes 128kbps ,ac3 takes 448kbps .
lets say i have movie encoded with mp3,if ill encode the movie with ac3 - the video will have less quality?
how the program 'bring' more bits to the audio,
lets say the movie is 100 sec long, before reencoding the audio took 128*100 bits=12,800 bits,that means that there are 12,800 bits of audio in the file,after reencoding the audio will take 448*100 bits = 48,800 bits,but where those 36,000 bits came from?

3)im trying to encode with 2-pass xvid and came with a problem,
at the 2-pass (i use virtualdub as mentioned) after press 'configure' ->'calc' i dont realy know what to do in this window,i ussualy just 'fix' the fps to the right fps as i see at GSpot or similar program,but the new movie is never at the same size as the original,why?

tyvm for your help

[jagabo]

hi,im new to the encoding business and would be glad if you help me with few questions ^_^.
ive read few guides from google about bitrate,codecs and virtualdub(as i understand the best program for encoding).

VirtualDub is a great program for many things. I wouldn't say it's the best program for encoding. It depends on what you want and how much learning you want to do.

1) i understand that the bitrate set the video quality,but as i saw i can also make video that i film with my iphone and encode it to few gb's ,then the iphone video will have more bitrate but less quality.

I don't know what you're asking here. Generally, different codecs have different compression capabilities. Different sources have different qualities.

2)about audio, mp3 takes 128kbps ,ac3 takes 448kbps .

No. Either can encode at a wide variety of bitrates.

lets say i have movie encoded with mp3,if ill encode the movie with ac3 - the video will have less quality?

No. The size of the audio has nothing to do with the size and quality of the video. Unless you are using a fixed file size (like 700 MB to fit the file on a CD).

lets say the movie is 100 sec long, before reencoding the audio took 128*100 bits=12,800 bits,that means that there are 12,800 bits of audio in the file,after reencoding the audio will take 448*100 bits = 48,800 bits,but where those 36,000 bits came from?

Take a flat wrinkle free piece of aluminum foil (uncompressed video). Compress it to a little ball. Now decompress the the ball (open it up to a flat sheet again). Now compress it to a ball that's twice the size of the first ball. Where did the extra foil come from? Bonus question: if you decompress this new ball will it have more wrinkles than last time, or fewer?

3)im trying to encode with 2-pass xvid and came with a problem,
at the 2-pass (i use virtualdub as mentioned) after press 'configure' ->'calc' i dont realy know what to do in this window,i ussualy just 'fix' the fps to the right fps as i see at GSpot or similar program,but the new movie is never at the same size as the original,why?

Did you account for the size of the audio?

[redwudz]

alibaba12, in the future please use a more descriptive subject title in your posts to allow others to search for similar topics. I will change yours this time. From our rules:

Try to choose a subject that describes your topic.
Please do not use topic subjects like Help me!!! or Problems.

Thanks,

Moderator redwudz

[hech54]

i understand that the bitrate set the video quality,but as i saw i can also make video that i film with my iphone and encode it to few gb's ,then the iphone video will have more bitrate but less quality.

Both Handbrake and WinFF have iPhone/iPod Touch presets that will give you exactly what you need. You will not do better doing it the hard way in VirtualDub.

[alibaba12]

i just gave an example for iphone movie ^_^

hi,im new to the encoding business and would be glad if you help me with few questions ^_^.
ive read few guides from google about bitrate,codecs and virtualdub(as i understand the best program for encoding).

VirtualDub is a great program for many things. I wouldn't say it's the best program for encoding. It depends on what you want and how much learning you want to do.

1) i understand that the bitrate set the video quality,but as i saw i can also make video that i film with my iphone and encode it to few gb's ,then the iphone video will have more bitrate but less quality.

I don't know what you're asking here. Generally, different codecs have different compression capabilities. Different sources have different qualities.
so,how i can know which movie has better quality ?
as i read most of the decesion is about bitrate>?

2)about audio, mp3 takes 128kbps ,ac3 takes 448kbps .

No. Either can encode at a wide variety of bitrates.

lets say i have movie encoded with mp3,if ill encode the movie with ac3 - the video will have less quality?

No. The size of the audio has nothing to do with the size and quality of the video. Unless you are using a fixed file size (like 700 MB to fit the file on a CD).

lets say the movie is 100 sec long, before reencoding the audio took 128*100 bits=12,800 bits,that means that there are 12,800 bits of audio in the file,after reencoding the audio will take 448*100 bits = 48,800 bits,but where those 36,000 bits came from?

Take a flat wrinkle free piece of aluminum foil (uncompressed video). Compress it to a little ball. Now decompress the the ball (open it up to a flat sheet again). Now compress it to a ball that's twice the size of the first ball. Where did the extra foil come from? Bonus question: if you decompress this new ball will it have more wrinkles than last time, or fewer?
i got you,but why in the first place i wont compress the ball(the movie) to the 'best' ball i can ? with almost 0 wrinkles?.
and,as you show this.
i can decompress a movie and make him with better quality ?

3)im trying to encode with 2-pass xvid and came with a problem,
at the 2-pass (i use virtualdub as mentioned) after press 'configure' ->'calc' i dont realy know what to do in this window,i ussualy just 'fix' the fps to the right fps as i see at GSpot or similar program,but the new movie is never at the same size as the original,why?

Did you account for the size of the audio?

i barely touch settings over there,just the fps as i said,should i change anything else?

[deadrats]

with regards to where the extra bits come from you need to understand that some compression schemes are more efficient than others, depending on the compression scheme used some require more bits to represent the same data as a more efficient compression scheme.

here's a simple way to visualize it:

1+1+1+1+1+1+1+1+1=9

3x3=9

assume both the above simple equations are compression schemes, after you apply both compression schemes you end up with the same value but with the first scheme you need to use more characters than the second.

[davexnet]

It's not really the compression scheme that's a factor.
If you take a typical mp3, shrunk from WAV/PCM by a factor of 10 (30MB wav > 3mb mp3)

If you convert the mp3 to ac3, internally, the 3mb is first decompressed back to the 30MB.
Then the 30MB's are recompressed to your new format and bitrate.

[teodz1984]

1. Why use virtualdub for making videos that are to be played on a i-device...
Most apple I-devices owners kwno that their device don't accept AVI as a valid file format... ITUNES will spit it out saying you shoud convert it to an MP4 file that is level 3 and limited to certain resolutions and a maximum bitrate (check the apple websites on the exact technical specs of your device).

2. MP3 is a lossy format... So when you have a file encoded in in MP3... the quality has alredy degraded.... I liken it to listening through a seashell. Upper Frequencies are usually cut off at low bitraes... e.g. 16,000 Hz and above are cut off and are lost forever... You cannot reclaim what is lost since the original information on the lost frequencies are gone...

Think of your audio as a big bolder... Pulverize it to get smaller chunks (throwing the small sandy particles)... If you try to reassemble the whole thing again (without any additional binder like cement) ... you won't get the same mass as the original boulder because of the lost paricles.


So even if you have a higher bitrate (320kbps MP3, AAC or FLAC) , the quality of the audio will never improve... The only way to get the lost information is to rip the movie or audio from the DVD or CD.

There are a lot of encoders here that can rip and convert videos to IPHONES... I use MEGUI, Mediacooder XVIDtoPSP and these can convert to MP4s that can be used by most I-devices.

[deadrats]

If you convert the mp3 to ac3, internally, the 3mb is first decompressed back to the 30MB. Then the 30MB's are recompressed to your new format and bitrate.

that's kind of disingenuous, yes it gets decompressed and then recompressed but it doesn't get decompressed to the exact same bit for bit copy as the original source (note mp3, ac3 and related compression schemes are lossy).

[teodz1984]

davexnet ,

Some Audio Compression Algorithms works different from FILE Compression Algorithms.... hence the label LOSSY compression

1. Try computing the md5sum of a wavefile before compressing to MP3, AAC etc...
2. Compress the file using MP3 Vorbis Mp2 or AAC
3. Decompress the mp3 back to the WAV... Compute the md5sum of that file.... and I'm sure it will NOT match the value of the original WAVE SOURCE.

davexnet ,
I suggest that you read up more on how lossy compresion schemes like
Mp3 work

[davexnet]

If you convert the mp3 to ac3, internally, the 3mb is first decompressed back to the 30MB. Then the 30MB's are recompressed to your new format and bitrate.

that's kind of disingenuous, yes it gets decompressed and then recompressed but it doesn't get decompressed to the exact same bit for bit copy as the original source (note mp3, ac3 and related compression schemes are lossy).

I never said or implied it gets decompressed to the *same* 30MB it originally was. (as ripped from the CD
for example) It's basically still the same quality as the 128kbps mp3 - just decompressed .

[hech54]

davexnet ,

Some Audio Compression Algorithms works different from FILE Compression Algorithms.... hence the label LOSSY compression

1. Try computing the md5sum of a wavefile before compressing to MP3, AAC etc...
2. Compress the file using MP3 Vorbis Mp2 or AAC
3. Decompress the mp3 back to the WAV... Compute the md5sum of that file.... and I'm sure it will NOT match the value of the original WAVE SOURCE.

davexnet ,
I suggest that you read up more on how lossy compresion schemes like
Mp3 work

Oh....get off of your "lossy format" high horse and try some common sense. Only a small fraction of today's population could tell the difference between a 128 or higher MP3 and a CD or .wav file. I usually can tell the difference....but I'm the only one out of my 3 other friends who are really into music that can tell the difference. What they hear is a song or artist that they recognize or sounds that are pleasing to their ear.....that's it. The kids today that have speakers and equipment that can reproduce a wide frequency spectrum are listening to it in their cars and going deaf doing so because they also think VOLUME clears up the deficiencies of a mediocre source.

[alibaba12]

about 2 pass encoding (xvid with virtualdub).
which bitrate \ file size should i write each time at the second pass?
how i can know which bitrate is common\optimal for my video?
tyvm

[davexnet]

Are you trying to encode to a certain file size? If so, use the calculator to set the bitrate.

If the output size is not so important, use CQ mode. Try CQ=4 to begin with.

[jagabo]

how i can know which bitrate is common\optimal for my video?

You can't until after you have encoded it. Because every video is different.

Use 2-pass bitrate based encoding when you need a file of a particular size (like you need a 700MB file to fit on a CD). With bitrate based encoding you'll be assured that the quality is the best the codec can deliver for that size. Ie, you know what the file size will be but you don't know the quality.

Use constant quality encoding (Target Quantizer in Xvid) when you want a particular quality. You know what the quality will be but you don't know the file size. When I did a lot of Xvid encoding I usually used Target Quantizer 3.

[alibaba12]

how i can know which bitrate is common\optimal for my video?

You can't until after you have encoded it. Because every video is different.

Use 2-pass bitrate based encoding when you need a file of a particular size (like you need a 700MB file to fit on a CD). With bitrate based encoding you'll be assured that the quality is the best the codec can deliver for that size. Ie, you know what the file size will be but you don't know the quality.

Use constant quality encoding (Target Quantizer in Xvid) when you want a particular quality. You know what the quality will be but you don't know the file size. When I did a lot of Xvid encoding I usually used Target Quantizer 3.

i have just saw basic guides.. if i have a movie 700MB and i want to add subitles with virtualdub,the best encode as i understand is 2-pass. but i dont realy care if the movie will be 600MB or 800MB i just want the movie the be at good quality and with subtitles.. how i can tell which file size will be the 'optimal' movie,naybe 900kbps or 950kbps? and one last thing that i dont fully understand , if i have movie encoded with 900kbps i can decompress or rencoded the movie to be with better quality? or the extra bits will be just useless? tyvm

[jagabo]

the best encode as i understand is 2-pass.

No. You don't know what the quality will be when you use 2-pass bitrate based encoding. You only know that you will get the best quality for the selected file size.

i have a movie 700MB... but i dont realy care if the movie will be 600MB or 800MB i just want the movie the be at good quality and with subtitles..

Then use Target Quantizer mode. Encode a short segment at different Target Quantizer values 2, 3, 4 (you can also use non-integer values like 2.5)... Watch the resulting videos and decide which value gives you the quality you want. Then use that value for all your encodes. You will always get the quality you ask for. (Relative to the source -- a low quality video will not get better if you reencode at high quality settings. You just know that the quality isn't going to get much worse.). The lower the quantizer you use, the higher the quality, and the bigger the resulting file.

Download and watch the videos in this post:

https://forum.videohelp.com/threads/295672-A-problem-for-video-experts?p=1811057&viewfu...=1#post1811057

One of the videos took 20 times more bitrate than the other to maintain picture quality.

if i have movie encoded with 900kbps i can decompress or rencoded the movie to be with better quality?

If you re-encode with a lossy codec you can only make the quality worse.

[alibaba12]

the best encode as i understand is 2-pass.

No. You don't know what the quality will be when you use 2-pass bitrate based encoding. You only know that you will get the best quality for the selected file size.

i have a movie 700MB... but i dont realy care if the movie will be 600MB or 800MB i just want the movie the be at good quality and with subtitles..

Then use Target Quantizer mode. Encode a short segment at different Target Quantizer values 2, 3, 4 (you can also use non-integer values like 2.5)... Watch the resulting videos and decide which value gives you the quality you want. Then use that value for all your encodes. You will always get the quality you ask for. (Relative to the source -- a low quality video will not get better if you reencode at high quality settings. You just know that the quality isn't going to get much worse.). The lower the quantizer you use, the higher the quality, and the bigger the resulting file.

Download and watch the videos in this post:

https://forum.videohelp.com/threads/295672-A-problem-for-video-experts?p=1811057&viewfu...=1#post1811057

One of the videos took 20 times more bitrate than the other to maintain picture quality.

if i have movie encoded with 900kbps i can decompress or rencoded the movie to be with better quality?

If you re-encode with a lossy codec you can only make the quality worse.

but that 'almost' pictures,and as i understand 2-pass is for scenes with lots of details or very less details. also, target quantizer is only for one-pass encode no? ty for help. what will be the different in quality if ill do the following: i have a movie with 1000kbps that will be Amov, ill Reencode Amov to 5000kbps that will be Bmov, and ill reencode Bmov 'back' to 1000kpbs that will be Cmov. what will be the difference in quality please? ive try with target quantizer 4.0 that gave me 340 kbps from movie of 687kbps

[jagabo]

but that 'almost' pictures,and as i understand 2-pass is for scenes with lots of details or very less details. also, target quantizer is only for one-pass encode no? ty for help.

No. 2-pass bitrate encoding is for when you want a specific size. Single pass Target Quantizer encoding is for when you want a specific quality. Either can give low or high quality. The difference is with Target Quantizer you know the quality before you encode (but not the file size), with bitrate based encoding you know the file size (but not the quality).

They are two sides of the same coin: If you encode a particular source in Target Quantizer mode and the resulting video turns out at 1000 kbps, then go back and encode that source again in 2-pass at 1000 kbps, the two resulting files will be nearly identical in size and quality.

[Cornucopia]

There are basically 3 general kinds of Bitrate choices:

1. CBR (Constant Bitrate) - for "simple" encoding, or for encoding where the pipeline bandwidth is very constrained to a particular bitrate (like VCD used to be), or virtually unlimited (to where BR and/or Filesize don't matter). Usually uses 1 pass (2 or more passes are basically superfluous/unnecessary).
2. Rate-Priority VBR (Variable Bitrate)- for efficient encoding where there might be a reasonable bitrate constraint but there is more likely a filesize/capacity constraint. Varies the Quality along with the bitrate vs. complexity, but provides targeted filesizes/bitrate curves. Usually uses 2 pass (1st pass is used to "learn" about the complexity of the title, 2nd to actually encode), but can do 1 pass (with just "guessing" the complexity in realtime) for speed requirements. Further passes might further refine efficiency, but are usually superfluous (diminishing returns).
3. Quality-Priority VBR (aka CQ, "Constant Quality" vbr)- for encoding where filesize requirements are looser, but quality requirements are specific (+need for quality to be constant). Varies Bitrate along with quality vs. complexity. Usually uses 1 pass, as further passes are unnecessary (target quantizer is set at beginning and it just "let's the chips fall where they may").

Maybe that will clear things up...

Scott

[alibaba12]

so you say that if i dont realy care about the filesize(+-20% lets say) the best encode will be quality priority VBR ? if that one pass how the program knows which scene will have more kbps? i understand that require 2 passes no? if i want to encode movie and add subs to it,the common way will be the use the same bit rate as the original movie?

[jagabo]

so you say that if i dont realy care about the filesize(+-20% lets say) the best encode will be quality priority VBR ? if that one pass how the program knows which scene will have more kbps?

It encodes each frame with the quality you specify. When it's working on a frame it doesn't have to know what the rest of the video looks. It's like when you save a JPEG image. You select the quality you want in the image. What you saved before and what you may save next doesn't matter.