Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: WMA 2-pass vs really high bit rate (Read 25906 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

WMA 2-pass vs really high bit rate

Ive been using WMA with max quality setting (98) for converting my cd's to be copied over to my zen vision:m. Since the ZVM dosnt support losseless mode, I found this to be the next best thing. I used DBpoweramp to do the compression, and while the checkbox for 2-pass was faded out before, on a new windows install (my new laptop) it's enabled, but it seems 192Kbit is the highest bit rate it supports. Considering that Quality based VBR was doing 400-600Kbits (according to winamp), I'm wondering if quality based would still offer the superior sound?

WMA 2-pass vs really high bit rate

Reply #1 –
2 pass makes no sense with variable bitrate, I don't know what you want to achieve with that?

WMA 2-pass vs really high bit rate

Reply #2 –
WMA 2-pass is used to get VBR encodings with a specified average bitrate. In this mode the encoder reduces quality automatically by analysing complexity in the first pass and then performs a usual VBR encoding in the second pass.

And yes, WMA quality-based VBR mode still has superior sound quality compared to other VBR modes (because it doesn't reduce quality during the encoding).

WMA 2-pass vs really high bit rate

Reply #3 –
Quote from: sony666 on
2 pass makes no sense with variable bitrate, I don't know what you want to achieve with that?


Clearly you have no understanding as to the point of 2 pass. It has nothing to do with vbr or cbr really. 2 pass simply means you are giving the program 2 chances to look at the data, instead of one, of wich it can better estimate at what to do to preserve quality with what ever data rate it is allowed. Divx, Xvid, Mpeg2 (dvd), mpeg4, h264 are all VBR and all utilise 2 passes or more.

WMA 2-pass vs really high bit rate

Reply #4 –
Well hello mister 2-posts-who-uses-600kbit-WMA-for-a-portable.

Please explain why a second pass is beneficial for VBR encodings when filesize is of no concern

WMA 2-pass vs really high bit rate

Reply #5 –
Quote from: JeffMD on
Clearly you have no understanding as to the point of 2 pass. It has nothing to do with vbr or cbr really.

Two-pass mode is used only with ABR encoding. It makes no sense in the context of VBR, because in that case the encoder can allocate as many bits as necessary for the desired quality.

So unless you need to encode ABR for some reason, you can ignore the two-pass setting.

WMA 2-pass vs really high bit rate

Reply #6 –
Microsoft call their ABR VBR so as not to confuse joe public (well abr is a controlled vbr). For example selecting the 2 pass shows a profile 192Kbps ..... VBR, where without it, it would be quality 98 VBR.

>Please explain why a second pass is beneficial for VBR encodings when filesize is of no concern

No lossy encoder does not care for filesize, it is the primary concern even for vbr encodings.

>I'm wondering if quality based would still offer the superior sound?

If it has a higher bitrate then it will be the superior quality (in relation to VBR 400 Kbps compared to 192Kbps).

WMA 2-pass vs really high bit rate

Reply #7 –
VBR is for a set constantly quality level

CBR is variable quality

2pass VBR/ABR is for best quality per file size, not constant quality as VBR
Chaintech AV-710

WMA 2-pass vs really high bit rate

Reply #8 –
Quote from: bubka on
VBR is for a set constantly quality level

CBR is variable quality

2pass VBR/ABR is for best quality per file size, not constant quality as VBR



mm no..

If I compressed a movie at 1000kbits single pass, and then at 700kbits 2-pass. the 2 pass version will look noticably better. Atleast in the video world, there are some algorythems that are only used in 2 passes that are not used in single pass, mainly because in 2 pass, the codec knows the source is not real time, and can look far ahead to sort out bandwidth allocation and such, as well as use slower and more accurate compressing (divx offers slower HQ modes in 2pass that it locks out in 1 pass mode). 1pass needs to beable to cope with on the fly encoding of a realtime source (provided you have the cpu power of course). All this leads to 2-pass being superior, even at lower bit rates.

Of course the divx and mpeg4 codecs and encoders are well documented. I cant seem to find much info on WMA's 2 pass mode and how differently it encodes.

WMA 2-pass vs really high bit rate

Reply #9 –
When you compress a video in 1 pass you must target a quality level, not a  certain bitrate, like 1000 in your example.
That is a HUGE differnece.. I am sure you will understand it.

WMA 2-pass vs really high bit rate

Reply #10 –
Quote from: JeffMD on
mm no..

If I compressed a movie at 1000kbits single pass, and then at 700kbits 2-pass. the 2 pass version will look noticably better. Atleast in the video world, there are some algorythems that are only used in 2 passes that are not used in single pass, mainly because in 2 pass, the codec knows the source is not real time, and can look far ahead to sort out bandwidth allocation and such, as well as use slower and more accurate compressing (divx offers slower HQ modes in 2pass that it locks out in 1 pass mode). 1pass needs to beable to cope with on the fly encoding of a realtime source (provided you have the cpu power of course). All this leads to 2-pass being superior, even at lower bit rates.

Of course the divx and mpeg4 codecs and encoders are well documented. I cant seem to find much info on WMA's 2 pass mode and how differently it encodes.

TOS violation #8, don't give me the this looked better, or this sounded better subjective crap...

let alone the fact that you cannot compare 1000k single CBR vs your 700k 2pass vbr / abr
Chaintech AV-710

WMA 2-pass vs really high bit rate

Reply #11 –
Quote from: JeffMD on
I cant seem to find much info on WMA's 2 pass mode and how differently it encodes.

Quote
Using VBR encoding
Windows Media Encoder supports three VBR encoding modes:

Quality-based VBR. With quality-based VBR encoding, you specify a desired quality level (from 0 to 100). Then, during encoding, the bit rate fluctuates according to the complexity of the stream—a higher bit rate is used for intense detail or high motion, and a lower bit rate is used for simpler content. The advantage of quality-based VBR encoding is that quality remains consistent across all streams for which you specify the same quality setting. The disadvantage is that you cannot predict the file size or bandwidth requirements of the encoded content before encoding. Quality-based VBR encoding uses one-pass encoding. This mode is good for archiving content.
If you are encoding audio content with this encoding mode, you have the option of using the Windows Media Audio 9 Lossless codec. Lossless encoding produces superior quality results, while also providing some compression of the data.

Bit rate-based VBR. With bit rate-based VBR encoding, you specify the desired average bit rate. At any point, the bit rate may exceed the average bit rate but the overall bit rate does not exceed the average bit rate. Bit rate-based VBR encoding uses two-pass encoding. In the first pass, the data complexities are analyzed. Then, in the second pass, the quality level is set to achieve the average bit rate. The advantage of bit rate-based VBR encoding is that the compressed stream will achieve the highest possible quality level while staying within a predictable average bandwidth.

Peak bit rate-based VBR. Peak bit rate-based VBR encoding is similar to the bit rate-based mode, except that you also specify the peak bit rate. The encoder determines the image quality that can be achieved without exceeding the peak bit rate. The bit rate does fluctuate during encoding, but does not exceed the specified peak bit rate. Use this option when you plan to distribute the content for playback on a device that has a constrained reading speed, such as a CD or DVD player. Peak bit rate-based VBR encoding uses two-pass encoding.

Quote
About one- or two-pass encoding

With Windows Media Encoder, you can do one- or two-pass encoding, either with CBR or VBR encoding. (The quality-based VBR encoding mode is one-pass; the bit rate-based and peak bit rate-based VBR encoding modes are two-pass.)

With one-pass encoding, the content passes through the encoder once, and compression is applied as the content is encountered. With two-pass encoding, the content is analyzed during the first pass, and then encoded in the second pass based on the data gathered in the first pass. Two-pass encoding can result in better quality content because the encoder takes its time to find the optimal combination of bit rate, frame rate, buffer size, and image quality based on the scene composition. However, two-pass encoding takes longer because the encoder goes through all of the content twice.
[/size]
 

WMA 2-pass vs really high bit rate

Reply #12 –
Quote from: bubka on
TOS violation #8, don't give me the this looked better, or this sounded better subjective crap...

let alone the fact that you cannot compare 1000k single CBR vs your 700k 2pass vbr / abr



Unlike audio wich indeed at this point, is very subjective since its hard to point out flaws and everyones hearing is a little different, video can be screen shoted and compared against the original and eachother, with actual pixles enlarged for examination by anyone who isnt blind thus would not fall under tos rule number 8.

This is also why video codec differenced are plain as day while people are still up in arms wondering what the best audio codecs are. ^^

Also it could be 700K cbr, its still 2 pass, and the computer has more time to calculate it (not to mention algorithms are used that are not in single pass) resulting in better video quality.

Egor, thanks, though the 2 pass desc seems to be for WMV and not Audio. Also looking for something more technical, like exactly what is calculated on the first and second pass, especialy vs what is calculated for one pass.

WMA 2-pass vs really high bit rate

Reply #13 –
Quote from: JeffMD on
Also it could be 700K cbr, its still 2 pass, and the computer has more time to calculate it (not to mention algorithms are used that are not in single pass) resulting in better video quality.

I don't know about video, but there is no such thing as two-pass CBR audio encoding*. Furthermore, I'm not sure what you mean by giving the computer "more time to calculate it" - single-pass video encoding doesn't have to be done in realtime. Assuming you're not talking about realtime encoding, the computer can take as long as it wants to encode, regardless of how many passes are used...correct?

The purpose of the first pass is to determine the complexity of the various parts of an audio file, so as to allocate bits most efficiently when encoding. This makes sense only when you are aiming for a specific bitrate. Using a quality-based VBR mode does not place such restrictions on the bitrate, so it is not necessary to analyze the entire file beforehand.

When in doubt, use quality-based VBR. You will never achieve better quality at the same bitrate using bitrate-based VBR, regardless of how many passes you use.


*Actually, this seems to depend on your definition of CBR. See below for details.

EDIT 1: I just noticed that Microsoft lists two-pass CBR as an option. Now I'm really confused. Someone please tell me I'm not completely crazy.

EDIT 2: Maybe Microsoft is just crazy: "With CBR encoding, the bit rate of the file or stream remains fairly constant and close to the target bit rate over the course of the stream." So their two-pass CBR is actually VBR which averages, over the length of the file, to the exact specified bitrate; their bitrate-based VBR is equivalent to everyone else's ABR. I think I'm unconfused now.

WMA 2-pass vs really high bit rate

Reply #14 –
Quote from: JeffMD on
Egor, thanks, though the 2 pass desc seems to be for WMV and not Audio.

The description is for both audio and video encoding (for audio just ignore video-specific features).

WMA 2-pass vs really high bit rate

Reply #15 –
Quote from: Egor on
Quote from: JeffMD on
Egor, thanks, though the 2 pass desc seems to be for WMV and not Audio.

The description is for both audio and video encoding (for audio just ignore video-specific features).

Probably good to ignore all the quasi-experts too (including me!). 

I'm surpised JeffMD isn't also getting jumped for spelling errors. 

WMA 2-pass vs really high bit rate

Reply #16 –
Quote from: quas on
EDIT 1: I just noticed that Microsoft lists two-pass CBR as an option. Now I'm really confused. Someone please tell me I'm not completely crazy.

EDIT 2: Maybe Microsoft is just crazy: "With CBR encoding, the bit rate of the file or stream remains fairly constant and close to the target bit rate over the course of the stream." So their two-pass CBR is actually VBR which averages, over the length of the file, to the exact specified bitrate; their bitrate-based VBR is equivalent to everyone else's ABR. I think I'm unconfused now.

I guess the CBR 2-pass mode varies the bitrate cosiderably less than the bitrate managed VBR 2-pass mode, so that it would be more suitable if the streaming bandwidth is restricted.

Perhaps the 2-pass CBR mode also allows the codec to better adjust the other variables besides the bitrate according to the complexity of a certain passage.

WMA 2-pass vs really high bit rate

Reply #17 –
JeffMD: the point is you still are not doing this "blind" testing even if it visual is different the auditory in detecting differences.

The example you gave to has so many variables also, so just the numbers don't mean anything
Chaintech AV-710

WMA 2-pass vs really high bit rate

Reply #18 –
The numbers weren't supposed to mean something other than that with two passes a smaller file may deliver better quality than a larger file created with only one pass.

Seems like some of you guys would rather belittle the guy rather than help him.

bubka, you make a wonderful point to JeffMD about blind testing, but you seem to be misdirecting it to instead try to refute an off-the-cuff example.

JeffMD, it seems you are interested in populating your ZVM with superior sounding files.  Take the spirit of the recommendations given to you:  run some ABX tests in order to determine what is best for you.  How small of a file can you create before you can tell the difference between it and the original?  Don't concern yourself about the placebo effect of knowing you're using the highest possible quality that a codec can deliver.  Let your EARS be your guide!

WMA 2-pass vs really high bit rate

Reply #19 –
no, sometimes you CAN say something is superior to another by the numbers / bitrate, but i said his numbers that he gave, as he tried to prove a point by them, mean nothing because of the many more additional variables in video encoding, let alone no abx test
Chaintech AV-710

WMA 2-pass vs really high bit rate

Reply #20 –
Quote from: JeffMD on
Ive been using WMA with max quality setting (98) for converting my cd's to be copied over to my zen vision:m. Since the ZVM dosnt support losseless mode, I found this to be the next best thing. I used DBpoweramp to do the compression, and while the checkbox for 2-pass was faded out before, on a new windows install (my new laptop) it's enabled, but it seems 192Kbit is the highest bit rate it supports. Considering that Quality based VBR was doing 400-600Kbits (according to winamp), I'm wondering if quality based would still offer the superior sound?


What is the point of doing 200+ kbps encodings if you can't tell the difference from the original? Have you actually tried to ABX your music to find out where your limit is? I personally have a though time ABX'ing Q4 (~128kbps) ogg vorbis, and that is with high-end headphones and soundcard.

/Kef

WMA 2-pass vs really high bit rate

Reply #21 –
Some questions.
1. Can length of audio samples  influence on performance of WMA10pro 64 kbit/s 1 and 2 pass?
I.e. 2pass mode can perform better than 1pass on 30 secs but has no sense on 180 secs.

2.  Has anybody played around wma10pro 1 and 2 pass perfomance? ABX tests only, please.

WMA 2-pass vs really high bit rate

Reply #22 –
Well, I dislike entering into controversies like this, but I must point out that any 2-pass algorithm can manage buffering and bit-rate much better than any 1-pass algorithm.

For an example, presume (yes, an extreme example) that we have a cut that is 50% silence, with music in the first and last 25% of the track.

1-pass can not, under any realistic circumstances, handle this.

2-pass, of course, can easily allocate minimum-rate to the middle part.

Obviously, nothing much (if anything) in the real world will be this extreme, but time diversity does exist, and it's pointless to deny that.

Hence, the comment here that launched the controversy, the one that said that 2-pass makes no sense when setting a fixed average rate, obviously has to be wrong.

Edited to add: My recollection is wrong, I see the original referred to "variable bit rate", but without the knowledge that "variable bit rate" may be controlled for overall average.  This, as opposed to a codec that guarantees a fixed average and peak rate, or a fixed buffer size, or a transport mechanism that allows synchronous or pseudo-synchronous streaming.

As to the semantics, I won't argue them.
-----
J. D. (jj) Johnston

WMA 2-pass vs really high bit rate

Reply #23 –
Quote from: Woodinville on
For an example, presume (yes, an extreme example) that we have a cut that is 50% silence, with music in the first and last 25% of the track.

1-pass can not, under any realistic circumstances, handle this.

2-pass, of course, can easily allocate minimum-rate to the middle part.
All good encoders will handle that just fine in 1-pass. They will drop down to the format's minimum bitrate during the silence. They just won't add extra bits to the non-silent portions of the track. That's because audio encoders are all built around the premise of average quality, not exact file size.

Quote
Obviously, nothing much (if anything) in the real world will be this extreme, but time diversity does exist, and it's pointless to deny that.
I see that exact example all the time: cds with have two songs in the last track, with a bunch of silence between them. MP3, AAC, Ogg vorbis, FLAC, they all do just fine. (Course I generally rip those tracks into two files to get rid of the long silence, but that's a personal prefrence...)

Quote
Hence, the comment here that launched the controversy, the one that said that 2-pass makes no sense when setting a fixed average rate, obviously has to be wrong.
Yes, if you need to encode to an exact average bitrate for a fixed file size, 2 pass is the best way to do it. The problem it that no one really wants to do that. The premise of the whole argument is stupid.

WMA 2-pass vs really high bit rate

Reply #24 –
Quote from: Klyith on
All good encoders will handle that just fine in 1-pass.
We're talking about WMA, I'm not getting into whether it is a good encoder or not and neither should you!

Quote
MP3, AAC, Ogg vorbis, FLAC, they all do just fine.
Hmm (looks at subject line), there's a codec that's conspicuously absent in this list.

Quote
Quote
Hence, the comment here that launched the controversy, the one that said that 2-pass makes no sense when setting a fixed average rate, obviously has to be wrong.
Yes, if you need to encode to an exact average bitrate for a fixed file size, 2 pass is the best way to do it. The problem it that no one really wants to do that. The premise of the whole argument is stupid.
Why? because you don't use WMA 2-pass???