Conversions to MP3

  • Thread starter Thread starter miroslav
  • Start date Start date
PS

Why do you say VBR is better than fixed?

It may be more convenient (there's that word again ;) ) because it adjusts to the chaging bandwidth, but that can cause hiccups/glitches in the audio stream when the rate adjust...unless your buffring has enough head start on things.

I would rather offer 2-3 fixed rates, and yeah...either it comes in or not, which IMO is better than having dropouts during the music, but YMMV.
 
vbr doesn't adjust to the bandwith, it adjusts to the audio content. Once an mp3 is encoded, it's encoded. There is no live encoding when streaming music.

what VBR does is it looks at the audio to determine how much it can compress it. If the audio is complex, it uses a higher bitrate, if it's less complex, it uses a lower bitrate. That way, transients and other complex sounds get better treatment, and you get better sound quality with a smaller footprint, so it's more convenient and better.

People tend to not like using vbr because it gives you a choice between 1 and 10 for audio quality and doesn't let you pick exactly what bitrate it will be. Losing this control makes people feel uneasy, but you get far better sound quality with vbr than cbr.

Also, you can specify what your average bitrate will be with vbr, so streaming shouldn't cause any more hiccups than cbr.
 
bigger does not = better
bigger = bigger
better = better

There is definitely correlation between the two, but they are not the same. A lame 192kbps cbr sounds worse than the 128kbps abr.
 
bozmillar said:
vbr doesn't adjust to the bandwith, it adjusts to the audio content. Once an mp3 is encoded, it's encoded. There is no live encoding when streaming music.

what VBR does is it looks at the audio to determine how much it can compress it. If the audio is complex, it uses a higher bitrate, if it's less complex, it uses a lower bitrate. That way, transients and other complex sounds get better treatment, and you get better sound quality with a smaller footprint, so it's more convenient and better.
Click to expand...

Mmmmmm...I think we are talking about two different VBR technologies.
I'm talking about VBR streaimg...which is what you can get on somtihng like and on an iPhone, and I dealt with it duyring my video streaming days, and had a lot of contatc with Akamai, who was our main world-wide distributor of streamed events

The VBR stream adjusts to the BANDWIDTH...you can read about it here:

"...for instance, if an iPhone user is watching streaming video on a WiFi connection but loses the signal and it switches to a 3G connection -- then the bit rate adjusts to match the device's available bandwidth."

http://www.contentinople.com/author.asp?section_id=450&doc_id=178863
 
that is a form of VBR, but not the same form that mp3 uses. mp3's VBR is hard coded at the encode stage. there is no changing it once it's been encoded.
 
bozmillar said:
bigger does not = better
bigger = bigger
better = better

There is definitely correlation between the two, but they are not the same. A lame 192kbps cbr sounds worse than the 128kbps abr.
Click to expand...

I dunno...I think we are now evaluating different types of "worse" :D
But I guess you have to pick what sounds good to your ears.

Like I said...I have yet to hear any sub-192Kbps MP3 that really sounds "better" than any 320Kbps MP3.
But then...there may be some real dog encoder out there that could trash even the mild 320Kbps compression.

Maybe I'm jaded...but I just feel that sound quality is always the primary concern...even when forced to deal with a crappy format like an MP3, I would try to use the rate that is the least offensive.
 
There are a lot of misconceptions out there about vbr mp3s, and people are generally scared to use it. In every test done, at least with lame encoder, vbr sounds better at the same average bitrate than cbr on all types of music.
 
Well...I wouldn't call them "misconceptions" exactly.

While most players can "handle" VBR encoded MP3 files...it appears many of them have various problems cleanly decoding VBR files....so while it may sound great with your encoder and your player...there can be marginal results on a multitude of other players...and how the heck can you know which types of players your audience is using?

I can see the guy who is just focused on ripping music for his own use...he will find the best encoder/player that works for his purpose, and he's happy.
But when you are a musician who is trying to get your music out there...I'm not sure I would want to chance VBR files when the CBR files will play back without a hitch on all players...and yeah, OK...you just use a slightly higher CBR and you will sound as good as the lower rate VBRs.

Here's a test chart for consideration:


http://mp3decoders.mp3-tech.org/vbr.html



vbr.gif
 
yeah, i remember back in 2000 having issues with vbr mp3s, but that was almost 10 years ago. Most of the software on this list you can't really find anymore.

Then again, anyone using a player in 2009 that doesn't decode vbr mp3s correctly is probably the same person with the dialup connection.
 
just as a quick simple test of encoding quality, I took these two clips, one encoded at 160kbps and one encoded at 320kbps and made a wave of the two.

I randomly placed 10 clips in a wave file. Some are the 160kbps and some are 320kbps.

I'll put myself out there and admit that I couldn't hear the difference at all. I got half of them right.

They are a short drum sample taken from Miroslav's web site.

www.bozrecords.com/mp3s/mp3Test.wav
 
Oops. Sorry Miroslav. I just read the disclaimer under your songs. If you want me to take this clip down, I will. It seemed like a pretty ideal clip being that it is a very clean recording and that hi hat should be the mp3 killer if there ever was one.
 
I'm new to this site...

I've asked the following on the Audacity forum over a year ago but got no reply of note.
Does anyone know of a faster mp3 encoder (than LAME) that'll work with Audacity? For a 70-80 minute recording, when I try to export to mp3, it says it'll take nearly 2 hours (on a 700MHz Celeron) to encode and save. Right now I'm using my personal copy of Cool Edit 2000 (at church) as a converter because it takes less than 10 minutes to start CE, load the wav and resave as mp3 using its built-in Fraunhoffer encoder. If I can find a faster encoder, freeware or paid, I'd rather stick to using just Audacity there.
 
Faster encoding = lower quality encoding.

Really I hate to be the one to say it, but you need a new PC :p
 
You must log in or register to reply here.
Back
Top