Hi,

This is more a general question about encoding, but since I mostly have mp3's I ask here: Why would one choose 2 pass encoding of mp3 or wma? Is it only to get a little higher compression rate, or would it also create a better sounding mp3 (or wma)? If 2 pass encoding is chosen does it matter if you are creating a VBR or CBR compressed file? If the answer to this depends on the codec (wma, mp3 etc.) it would be good to know as well.

Thank you in advance.

//C

2pass does not make sense for pure VBR (pure variable uses what it needs)

it also doesn't make sense for CBR (constant would not change)

it only helps an ABR encoding (to help in choosing when/where to allocate bits) or a situation where a target size is selected (and then would use ABR to get to that target size)


later

what is 2pass encoding ?

On pass 1 the encoder analyzes the piece for complexity and encoding difficulty, but (generally) no file is output. On pass 2 it uses the info from pass 1 to maximize quality while producing a file of the target size or bit rate.

You could do the same thing yourself manually. Encode a file at, say, -V3. if the file is larger or smaller than what you wanted the reencode at -V2 or -V4, etc.

So when for instance DbPowerAmp has the option to choose VBR and 2 pass for wma encoding it's just a waste of time?

VBR 2-pass gives to users to mix one advantage of VBR (great amplitude of bitrate variation -> ~constant quality) with one advantage of ABR (target bitrate). If you check the bitrate curve of a 2-pass encoded file you will notice that the variation could be much more important and longer than with simple ABR.
As said by pdq you can manually perform a double/triple/...-pass encoding by adjusting the setting after the results of a first one. Finding the good setting by yourself should just take much more time

Maybe are these automated 2-pass encoders (WMA, Nero Digital AAC) using some analysis tool to distribute the bitrate a bit differently on the second pass (like video encoders adjustable using bitrate curve compression).

Aha! I am getting the difference now. What I am doing is filling an old player for my daughter. She wants a certain group of songs on there, so I need to make files around 64 mbps to get all she wants squeezed in. What should I choose? Quality VBR or 2 pass VBR? When it comes to codec I have mp3 and Windows Media 10 Pro through WMP 11 Beta. I have to say she is a quality minded with a bit over average head phones, so it would matter how the results sounds also

Advice appreciated.

64 mbps?
At 64 kbps it's likely that WMA should sound more pleasant than MP3 (some people don't like WMA artifacts at this bitrate though).
It's also unlikely that your daughter's old player supports WMAPro format (which is very different and not compatible with standard WMA format).

For the choice of WMA (standard)'s setting (CBR, VBR, 2-Pass), I can't really help.

Yup, you're right. The Windows Media Pro files wouldn't play. So I'm back to VBR 64 kpbs (I don't know how the m snuck itself in there..

You can try encoding some files at WMA Standard, bitrate-based VBR (2-pass) with foobar2000, instructions. The recommended options to try are
64_44_2
96_44_2
128_44_2
You will get hardware-compatible WMA 9.2 Standard files.

Please also note that at 128kbit LAME mp3 encoder produces very good results (transparent for the most according to blind tests).

I recently had to encode some mp3's for my phone and only had 64MB to store them in, and i found "lame -V6 --vbr-new --resample 32 -k " to be a pretty good compromise of quality + filesize. Some tracks hit as low as 90kbps and still sounded good.

For anyone who may search this topic in the future: