Conversions to MP3

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

miroslav

Cosmic Cowboy
I'm curious about the preferred method for folks when it comes time to make some MP3 files from their mixes. I know we all end up converting to 16/44.1 for our final CD format from whatever we were using during the mixdown (24 or 32 float and whatever rate we prefer), but from which point do you like do do your MP3 files?

1.) From the pre-16/44.1 files (24 or 32 float and your preferred sample rate)...this is assuming your Mastering occurs at this higher resolution and not after you go down to 16/44.1

2.) From the 16/44.1 files..basically your CD-ready files?

I get the feeling that in the end the answer will be "whatever sounds better"... ;) ...but I just wanted to see if there are any real compelling reasons to pick one over the other.
I've tried it a few ways...but I never sat down to really do any critical/scientific comparisons with the results.

My biggest differences come from the chosen MP3 rate...and I find that anything below 320Kbs and the sonic differences from the 16/44.1 WAV files is pretty obvious. At 320Kbs they ain't too bad.
I’ve been rather stubborn about MP3 files for a long time because of their obvious lower quality, and my attitude was to just no care about them 'cuz no matter what, they were never as good as the CD format, but it’s no longer possible to ignore the iPod generation, so I want to find the “best” :rolleyes: MP3 approach.
 
I always keep a 24bit Wav master and make MP3's and CD tracks from that.

I try to not go below 320kps for MP3's, but many sites stream at 128, so that's out of my control. But on my own site, all my MP3's are 320.
 
RAMI said:
I always keep a 24bit Wav master and make MP3's and CD tracks from that.

I try to not go below 320kps for MP3's, but many sites stream at 128, so that's out of my control. But on my own site, all my MP3's are 320.
Click to expand...

+1 For me, the Lame encoder is one of the better codec's out there.

http://lame.sourceforge.net/
 
One thing is to make sure the Quality setting on the LAME encoder is set to highest. Some programs that use LAME have a default of "Medium" or something, and it makes a big difference...

If I'm uploading to a site that doesn't restrict bitrate, I usually use 192 since (in my opinion) it's the best compromise between sound quality and file size.
 
.guitar said:
One thing is to make sure the Quality setting on the LAME encoder is set to highest. Some programs that use LAME have a default of "Medium" or something, and it makes a big difference...

If I'm uploading to a site that doesn't restrict bitrate, I usually use 192 since (in my opinion) it's the best compromise between sound quality and file size.
Click to expand...

LAME encoder is what I use at highest quality,
As for bit rate 128 is not good enough, 160 is the least I tend to use for letting my clients get the jist of a sample of a Mix or Master.

The difference in quality between 16 bit 44.1kHz .WAV and 320 MP3 is not much but still a difference of course.

A test to hear the difference is to play both 1644 .WAV and a 320 MP3 of the same track at the same time while inverting the phase of 1 of the tracks.
From this you will hear the loss of quality.

Eck (G)
 
I use LAME encoder directly from the 24bit file, but I use the Variable Bit Rate on the highest setting for mp3. Actually I'd prefer to use ogg as compression because i think it sounds heaps better for the same file size compared to mp3, but I make sure all recordings on my site are mp3 as that's more standard. It was when someone once pointed out that ogg uses variable rate that I started trying out mp3 at variable rate.
 
violist said:
I use LAME encoder directly from the 24bit file, but I use the Variable Bit Rate on the highest setting for mp3.
Click to expand...

I always stayed away from VBRs...and that's going back from my streaming video days.
With a VBR, any time the bandwidth changes, yeah, the VBR is supposed "adjust"...but that can cause hiccups and changes in quality too. It wasn’t as bothersome with video, ‘cuz a small video glitch seems to go by unnoticed a lot easier than when your audio glitches…but it still affected the audio portion of the stream.

I would rather have the end user listen to the ONE target rate that I choose (or give them a couple of fixed rates)...than to have the Internet changing it at will whenever the bandwidth fluctuates.

Anyone have any thoughts on this...?
 
Making mp3's is never part of my mixdown or smashtering process, it's an ad-hoc thing. I only make mp3's of my stuff if I'm planning to listen to it on my mp3 player or post it here. So it's basically ripped from the finished .wav, or a CD if I've made one. I use the LAME encoder.
 
legionserial said:
Making mp3's is never part of my mixdown or smashtering process, it's an ad-hoc thing. I only make mp3's of my stuff if I'm planning to listen to it on my mp3 player or post it here. So it's basically ripped from the finished .wav, or a CD if I've made one. I use the LAME encoder.
Click to expand...

But did you ever try making them form the higher-rez files...the same files you make your CD .wav files from.... and comparing the two?
 
ecktronic said:
A test to hear the difference is to play both 1644 .WAV and a 320 MP3 of the same track at the same time while inverting the phase of 1 of the tracks.
From this you will hear the loss of quality.

Eck (G)
Click to expand...

Never thought to try this.
icon14.gif


I personally can't hear a noticeable difference between 192 and a 16-bit WAV, but haven't run that test yet. And my ears aren't the greatest...
 
miroslav said:
But did you ever try making them form the higher-rez files...the same files you make your CD .wav files from.... and comparing the two?
Click to expand...

No, because I don't really care. We're talking about mp3's here. The quality is degraded anyway. Worrying about the subtle difference between ripping from CD and ripping from a wav, has no more point than worrying about the subtle difference between 2 ever so slightly different shades of shit. :D

If you're in a situation where you really have to worry about something like that, then you probably shouldn't be presenting the music in an mp3 format anyway.
 
legionserial said:
No, because I don't really care. We're talking about mp3's here. The quality is degraded anyway. Worrying about the subtle difference between ripping from CD and ripping from a wav, has no more point than worrying about the subtle difference between 2 ever so slightly different shades of shit. :D

If you're in a situation where you really have to worry about something like that, then you probably shouldn't be presenting the music in an mp3 format anyway.
Click to expand...

You're preaching to the choir! :D

I’m no fan of MP3 files...but as much as I would like to, I can't stick my head in the sand either and just ignore them altogether since these days THAT is the format most people listen to. :(

I guess if you just record for yourself...then you can ignore them.
But if you are planning to promote/sell your music...it will end up as MP3 files sooner or later...so I figure it might make sense to squeeze the most "quality" out of them as possible myself, rather than leave that to someone else who may care even less.

I mean, we all try to squeeze the most quality that we can out of our gear and recording techniques, and even a lighter shade of "better" is important to us during the recording process....
...so I might as well try and do that with MP3 files.

They ARE the prevalent music format...like it or not (I hate it, just like you do).
 
miroslav said:
You're preaching to the choir! :D

I’m no fan of MP3 files...but as much as I would like to, I can't stick my head in the sand either and just ignore them altogether since these days THAT is the format most people listen to. :(

I guess if you just record for yourself...then you can ignore them.
But if you are planning to promote/sell your music...it will end up as MP3 files sooner or later...so I figure it might make sense to squeeze the most "quality" out of them as possible myself, rather than leave that to someone else who may care even less.

I mean, we all try to squeeze the most quality that we can out of our gear and recording techniques, and even a lighter shade of "better" is important to us during the recording process....
...so I might as well try and do that with MP3 files.

They ARE the prevalent music format...like it or not (I hate it, just like you do).
Click to expand...

Miroslav is so right about this. Most of us try to squeeze every last drop of quality from our equipment, use expensive cabling, etc. and otherwise obsess about getting the best possible product period. If you can get your chosen delivery method to sound as best as possible shouldn't you try?

I personally have not experimented with what bit rate/sampling frequency works best when making MP3's of a mix but I do always try to get the kilobytes per second rate as high as possible. Sites like SoundClick have a file size limit when posting MP3's so just making every MP3 a 320kbs file isn't always possible. I check what the limit is and then make the file as close to that limit as possible with the highest rate possible.
 
JRSIV said:
Sites like SoundClick have a file size limit when posting MP3's so just making every MP3 a 320kbs file isn't always possible.
Click to expand...

Yeah...that's one of the reasons why I skipped on SoundClick.
I just thought their 128Kbps limit was so lame.
For real rough demos and just tossin' up ideas, I guess it's OK...but when you try to do something deeper, and then have it get reduced to sonic hash...
 
I render it in Reaper and leave it at 24bit and 88200. I create a 320 mp3. that's just leaving everything at it's highest quality, right? so isn't that best?

but that's at the mastering phase
 
.guitar said:
Never thought to try this.
icon14.gif


I personally can't hear a noticeable difference between 192 and a 16-bit WAV, but haven't run that test yet. And my ears aren't the greatest...
Click to expand...

You will definitely hear a good difference between 192 and 1644 with the phase reverse test. :) Mostly the detailed highs.

Eck
 
stupidfatnugly said:
I render it in Reaper and leave it at 24bit and 88200. I create a 320 mp3. that's just leaving everything at it's highest quality, right? so isn't that best?

but that's at the mastering phase
Click to expand...

By using the phase reverse test as I mentioned ealrier in this thread you can easily hear if there is any difference from encoding to mp3 from a higher bit rate and sample rate.

I don't believe there will be any difference, think I maybe tried it out before but can't be totally sure.
So try the test by firstly playing the 2 tracks (24bit 88200) at same time and 1 phase reversed and bounce and save the file.

|Then....
 
ecktronic said:
I don't believe there will be any difference,
Click to expand...
Unless the MP3 codec turns losless at 320k - which I don't believe it does (though I'm not 100% certain) - the null test will show a difference.

But I'd also point out that what's audible in a null test isn't always necessarily audible within the playback of the full mix. It may be, it may not, but the null test itself will not tell that.

Just to chine in, I prefer the original Fraunhoffer (sp?) codec over any of the public ones, and unless I am working with extreme high fidelity content (e.g. classical, spansih guitar, etc,.), anything above 192k tends to be a waste of bandwidth, IMHO. Even then, anyone who seriously wants to listen to that kind of stuff in audiophile mode isn't going to be doing it via MP3 anyway, regardless of the data rate.

G.
 
miroslav said:
Yeah...that's one of the reasons why I skipped on SoundClick.
I just thought their 128Kbps limit was so lame.
Click to expand...
That's the limit for free accounts. For paid accounts you can upload at 320.
 
You must log in or register to reply here.
Back
Top