08-12-2008 06:19 AM
I have already ripped many tracks into WMA format at the default 128 bit rate, and though it sounds ok to me on my e280, I have been pretty much convinced to do further ripping into MP3 format instead.
The main reasons (correct me if I've gotten this wrong) which I learned from reading posts at this forum being
And I next gathered that if you are going to rip into MP3 format, the prime software choices are
I am leaning towards using Media Monkey for ease of use.
My next question has to do with bit-rate.
Some have said that if you are going to rip into MP3 format, do so at a minimum of 194 bitrate.
But then I started reading about VBR (variable bit rate). Is this the preferred way to rip?
Just as an experiment, I ripped several tracks into both WMA at 128 bitrate (using WMP) and then into MP3 (using iTunes) at VBR medium-high quality setting.
I thought you might like to see the difference in file sizes:
WMA @ 128 MP3 @ VBR, med-hi quality % Increase in Size
This is a considerable increase in file size.
So I was wondering what your opinion is regarding VBR, and if you recommend it over choosing a fixed rate, what settings and software you are using.
08-12-2008 07:43 AM
It's not Variable BitRate that is giving you the higher file size--it is the quality setting. 128 is not considered high quality, so the average bitrate for medium quality is going to be higher than 128. Kbps (kilobits per second) measures the amount of data, the filesize, used for each second of music.
In fact, VBR saves a little bit on file size over CBR (constant bitrate) because instead of encoding constantly at 192 kbps (or another bitrate) it varies depending on what it detects in the music--parts are lower, parts are higher. Medium-High quality probably makes an average bitrate between 160-192. If you put it at a lower quality, you'd get file sizes closer to 128.
You can check what the average bitrate turned out to be in Properties/Summary/Advanced on the mp3 file, or possibly in your music player. It will be different--variable--for each song.
People have occasionally complained the VBR files give the Sansa problems. This has never happened to me, and I wouldn't rule out the possibility that VBR was blamed when tags caused the trouble. Try playing back your files before giving up the CD.
I mostly use iTunes at 192 kbps for its tagging capabilities, although LAME in Media Monkey (and EAC) is considered better. That's because the CDs that I rip tend to be obscure or very new and the user-generated Freedb often doesn't have the tags while the official, professional CDDB does. Sometimes I try first in Media Monkey to see if it has tags yet.
If Media Monkey is giving you good tags, the LAME encoder is supposed to be better, so stick with it. I believe EAC also uses the LAME encoder. Its advantage is that it is supposed to read the CD better (and slower) than the other programs, and it also connects to Freedb, giving the same tags. But it's not as user-friendly as MM.
All of these differences are subtle. They may be detectable by a computer readout, but not always by a listener.
You have to choose your own balance between size and quality. People go crazy reading numbers rather than trusting their ears. The real audio geeks encode at 320 kbps--even though some of them then take those 320 kbps files and play them through dinky little earbuds, which is ridiculous.
If I were archiving CDs that I could no longer keep, and had limited storage, maybe I'd use 320 kbps myself to preserve as much quality as I could. But for my practical purposes when I'm using the Sansa--often when I'm in motion or in noisy places--192 is fine. And I'm probably even being excessive with that; for all I know, I could live with 160 for most things.
Try to forget the numbers for a second and encode a favorite song at various bitrates: 128 and up, VBR, CBR. Use something multilayered with acoustic instruments, drums and voices--not some electro blip--so that the mp3 encoders have to deal with complex waveforms.
Have someone else retitle the files randomly (SameSong1, SameSong2) and keep a record of which bitrate was which without showing you. And then play them back and see what differences you hear and which version you actually like best without knowing the numbers. Maybe 128 gives you all the quality you need, and you can get another 40 percent of music onto the disc. Or maybe you'll start looking for nuances and get spoiled and want 256 kbps. Your ears should be the judge--not some number.
08-12-2008 10:57 AM
Thanks so much for your comprehensive response. Very helpful!
I will try your suggestions.
Meanwhile, I stumbled on a freeware by Eusing Software called Free CD to MP3 Converter 2.3.
(He has a freeware registry cleaner that was well reviewed, and which I have been using).
It uses both the LAME encoder, as well as the CDDB database. I'd be interested in what you think about it. Thanks again for your help!
Here's the link & description: http://www.eusing.com/CDRipper/CDRipper.htm
Free CD to MP3 Converter 2.3
Free CD to MP3 Converter is an easy-to-use free CD ripper software that allows you to extract audio files from a CD and convert them to MP3, Wav, Ogg or Wma format. It can automatically retrieve title information from the CDDB database, normalize the output files and supports additional LAME options. You can also use the software to record to MP3 with your microphone, or to convert WAV files to MP3, Ogg or Wma format. Support ID3 tag edit. You can use it freely. Free CD to MP3 Converter used AKRip and LAME encoder. AKRip and LAME are licensed under GNU LIBRARY GENERAL PUBLIC LICENSE.
Old Name :CD to MP3 Freeware
Support Free CD to MP3 Converter
Free CD to MP3 Converter is a freeware product, this means that you can download and use it for free. Anyway we will appreciate contributions: we will use collected money to finance new development of Free CD to MP3 Converter.
Without your support, perhaps we can't keep adding new features. Thanks for your help.
08-12-2008 11:18 AM
FYI, those 3 tracks that I gave data on ended up being 189 kbps, 256 kbps, and 224 kbps with the VBR in iTunes at Medium-High Quality, when I checked them as you suggested.
08-12-2008 05:46 PM
That definitely explains the larger file size. I like 192kbps mp3 for CDs, and Rhapsody tracks are currently 160kbps wma.
For classical, I lean towards 256kbps mp3, for the detailed stuff.
As far as vbr versus cbr, I can't justify the bells and whistles versus vanilla higher bitrates as a rule.
Get those earbuds back on!!!
08-12-2008 08:43 PM
From the description, that CD ripper looks promising. I don't know how it gets to use CDDB, which must be expensive since it's only used by the big guns: iTunes, WMP and maybe Winamp (though Winamp makes you pay for mp3 ripping).
I always check any unknown program for viruses and spyware. But sourceforge is a pretty reputable source.
I'll give it a try sometime in the next few days.
08-12-2008 09:05 PM
Sourceforge is just the source for the LAME encoders, not the program itself. Still it's probably worth a look-see.
08-13-2008 10:08 AM
FYI , Some more data to share. I've just started using Media Monkey to rip.
I read in their forum that it is advised to use the MM presets for ripping settings.
When you click on the settings button, you get a list of presets which start at the lowest quality and move upwards.
I chose Medium:VBR ~ 160 kbps which they rate as High Quality + , and ripped the same 3 tracks that I gave some data above on.
Here's the results:
I know c1u31355 advised not to be too concerned about numbers, but this seems very good to me. High average bitrate and small file size.
I tried the experiment of ripping sample tracks at various fixed (CBR) bitrates and then using VBR, and when I play them back on my computer using the sansa earbuds, I don't notice much difference to my (inexperienced) ears.
But it does make sense to save in pretty good quality, since I would not want to re-rip in the future.
This setting in MM seems like a good place to land.
Thanks for all the great advice.
Now I need to think about making the change to rockbox, since I have a version 1 player...
08-13-2008 06:13 PM
Those numbers do sound like good quality. I don't mean to totally disregard numbers--just not to be obsessed with them.
I do wonder what the difference is in the texture of the sounds of those songs. There must be a reason for the higher bitrate on the first two. The VBR encoder is looking at the music waveform and deciding that it needs more bits to make the first two tracks sound good. That must mean there's more subtlety or richness in the sounds.
08-14-2008 04:21 AM
Valid point, for sure...
Here's another bit of info.
I re-ripped those tracks with MM with the same setting: VBR ~ 160 Hi Quality + .
When I loaded those tracks into MP3TAG, it reports different bitrate for these VBR tracks than "properties" tab on the individual files:
I don't really notice much significant difference between the 3 tracks. They are latin jazz (Poncho Sanchez, Latin Soul, the 1st 3 trks).
All are from a live performance --- all have multiple instruments --- the third one has no vocals though.
Here's a chart I found in MM help:
MediaMonkey has a broad range of MP3 encoding options, allowing you to choose from a range of optimized preset configurations, or to set encoding options manually. What follows is a description of each of the presets, to help you choose the setting that's right for you (if you want to learn the intricacies of manual MP3 encoding options, you can find information via several of the links below.Preset Quality Bandwidth Recommendation Compatible,
CBR 96 kbps Fair Low, inefficient Useful if MP3s are to be played on a low storage capacity portable/device that is incompatible with VBR formats. Compatible,
CBR 128 kbps Average Low, inefficient This preset is fairly popular despite the average quality level, mainly because quality is adequate and it can be played on low memory portable/devices that are incompatible with VBR formats. Compatible,
CBR 160 kbps Good Medium, inefficient Useful for encoding MP3s at a decent quality level when the MP3s are to be played on a device that is incompatible with VBR formats. Medium (fast),
VBR ~160 kbps High Medium, efficient Recommended for encoding MP3s at a high quality level, at which only some users will be able to distinguish it from the original. It encodes slightly faster than the Medium preset, but at a slightly reduced quality level. Medium,
VBR ~160 kbps High+ Medium, efficient Recommended for encoding MP3s at a high quality level, at which only some users will be able to distinguish it from the original. Compatible,
CBR 192 kbps High+ High, inefficient Useful for encoding MP3s at a high quality level when the MP3s are to be played on a device that is incompatible with VBR formats. Standard (fast),
VBR ~192 kbps Very High High, efficient Recommended for encoding MP3s at very high quality levels, at which only a small proportion of listeners will be able to distinguish it from the original. It encodes slightly faster than the Standard preset, but at a slightly reduced quality level. Standard,
VBR ~192 kbps Very High+ High, efficient Recommended for encoding MP3s at very high quality levels, at which only a small proportion of listeners will be able to distinguish it from the original. Extreme (fast),
VBR ~256 kbps Near CD Very high, efficient Recommended for encoding MP3s at near-CD quality levels for very discriminating music listeners (with plenty of hard disk space). Only a very small proportion of listeners will be able to distinguish these MP3s from the original. Note that it encodes slightly faster than the Extreme preset, but at a slightly reduced quality level. Extreme,
VBR ~256 kbps Near CD+ Very High, efficient Recommended for encoding MP3s at near-CD quality levels for very discriminating music listeners (with plenty of hard disk space). Only a very small proportion of listeners will be able to distinguish these MP3s from the original. Insane,
CBR 320 kbps Highest Insanely high, inefficient Recommended for encoding MP3s at the highest level possible. It's for very discriminating music listeners (with hard disk space to burn). Only a very small proportion of listeners will be able to distinguish these MP3s from the original. Phone,
ABR ~16 kbps mono Poor Very low, fairly efficient, predictable Useful for low quality recordings, when bandwidth used needs to be constrained to a predetermined average. Short Wave,
ABR ~24 kbps mono Poor Very low, fairly efficient, predictable Useful for low quality recordings, when bandwidth used needs to be constrained to a predetermined average. AM,
ABR ~40 kbps mono Poor Very low, fairly efficient, predictable Useful for low quality recordings, when bandwidth used needs to be constrained to a predetermined average. Voice,
ABR ~56 kbps mono Fair Very low, fairly efficient, predictable Useful for fair quality voice recordings, when bandwidth used needs to be constrained to a predetermined average. Tape,
ABR ~112 kbps Average Low, fairly efficient, predictable Useful for average quality music recordings, when bandwidth used needs to be constrained to a predetermined average. Hifi,
ABR ~160 kbps High Medium, fairly efficient, predictable Useful for high quality music recordings, only when bandwidth used needs to be constrained to a predetermined average. Otherwise, the Medium or Medium (fast) presets are preferable. Hifi+,
ABR ~192 kbps Very High High, fairly efficient, predictable Useful for very high quality music recordings, only when bandwidth used needs to be constrained to a predetermined average. Otherwise, the Standard or Standard (fast) presets are preferable. Studio,
ABR ~256 kbps Near CD Very high, fairly efficient, predictable Useful for encoding MP3s for very discriminating music listeners, only when bandwidth used needs to be constrained to a predetermined average. Otherwise, the Extreme or Extreme (fast) presets are preferable.
These VBR tracks play fine in my player, and sound pretty good.
Do you think MP3TAG's bitrate is more accurate?
Lastly, if you don't mind my asking, are you using rockbox?