I have some flac music and I want convert to mp3. First I convert flac to wav with dBPowerAmp. Then I'm going to use EAC to convert to mp3. At this point I have question about command line. I'm thinking to use this one:
-b 32 -m j -h -V 0 -B 320 --lowpass 20 -q 0 --vbr-new
Is this command line have something wrong? Lame documentation says that option -h do some quality improvements and behaves the same as the -q2 switch. So if I want to use -q0 I don't need to use -h? Another question about -m option. Which stereo mode has better quality stero or joint stereo?

Is there a reason you pick this complicated procedure to go from FLAC to MP3?

As for your "maximum quality" encoding command line option - is there a reason you feel that you are likely wiser than the developers of LAME and that somehow you can increase the quality beyond the preset defaults? Do you believe the developers of LAME have some desire to hide "the really good stuff" from average users? Have you encoded and tested 1/10th of 1/10th of one percent of the tracks which have been used throughout the years to find the ideal balance between settings?

If you feel you must encode at "maxium quality" - encode at CBR 320, and leave the rest alone. (If anything lower the lowpass. If I had a nickel for everyone who wrongfully thought they could hear up to 20k I'd be rich. If I had a nickel for everyone who wrongfully though they could hear 20k in music Bill Gates would be serving me mint juleps out by the pool.) Everything else is masturbation. (You're playing around because it feels good, but produces nothing of lasting value)

[sigh]


Why two steps? Do it in one; anything else is a waste of time.


Really?


Almost everything.


No, and -q0 has no effect in VBR mode, anyway.


-m j is default (and thus redundant), and superior.

Did you do the most cursory research for yourself before posting this? I hope I don't have to direct you to Google, or even Hydrogenaudio's own wiki (which I'm sure appears high in the former's results).

As stated above, for 'maximum quality' (i.e. minimum likelihood of audible differences from the source file), use -b 320. For 'maximum quality' VBR (which will generally be just as reliable, while not wasting 320 kbps frames on simpler audio), use -V 0. That's it; no -q, no -m, -b, -B or anything else.

WOW!! great answer
"Why two steps? Do it in one; anything else is a waste of time."
Because i want use command line in EAC and EAC can't convert flac to mp3.
"Did you do the most cursory research for yourself before posting this?"
yes, I do. http://jthz.com/mp3/ I think knows better than you
"use -V 0. That's it; no -q, no -m, -b, -B or anything else"
why not using --lowpass 20 read this first about http://ixbtlabs.com/articles/oggvslame/
if i use -V0 option will it take range from 32 to 320?

Why are you linking to a HA-bashing site as defense of your choices while asking HA for help?

PS - you're linking to a site using pictures of spectrograms in defense of encoding > 16k. As I already stated I doubt you can hear the missing frequencies.

If you only have 320kbps with MP3 - do you really want those precious bits wasted on data you can see but not hear?

@ figo: If you want to be stupid, then do as you like, else, don't ask stupid questions.

I think it is very stupid for your part to ask here what command line to use, and then tell us that we don't know, showing a link against this very forum.

If you are minimally serious, you should know that some (all?) of the active LAME developers come to this forum from time to time, and if that other site think they are in a better position than the developers as to say which command line is better, let them think so, but do not pretend we will accept it blindly.

Also, as to show how misinformed they seem to be, the "alt-presets" were just called presets in 3.97, and aren't even present in 3.98, the current version. I won't start about why they were called "alternative", since they clearly have no clue. I were there when alt-presets were born.

(Your ungratefulness and continued n00b-ness aside)


Why not use dBpowerAMP to do the conversion from FLAC to MP3 in one step?


Yes.

Poor guy. He has no clue what he's up against

As for "how to get maximum quality" - Don't use mp3, use flac or any other lossless codec if you are worried about quality loss. mp3 and other lossy codecs are only useful if you want to reduce the filesize of music, but you can do so transparantly with no effort - You just have to accept that you don't need to push settings to it's maximum to achieve complete transparency. mp3 are perfectly able to reproduce the original music even at compression levels -v 4-5 (~128kbit VBR) and this is possible due to good algorithms, f.ex. using 16khz as lowpass. Noone (human) is able to hear these audiophoolish-super-sonic-good-vibration-sounds above it. You can either accept this fact and perform ABX tests (using foobar2000 or any other ABX-tool) between a lossy encode and a lossless source, or you can deny it, whilst you won't get any help from here.

Has anyone else wish they hadn't surfed to a certain site at one time? For your own good, HA people, don't visit it. The site is a total crock of s**t.
The age of r3mix and tweaking the command line LAME switches was so long ago that I even momentarily forgot about the --alt-preset labels and LAME has advanced a long way since then too.

Short answer to OP: -b 320

Have a nice day.

That is understandable except that you don't need to use a command line. Even if you really, really, really felt that you had to (and understood Lame more so than the developers here), you can always use dBpowerAMP's CLI mode or even foobar2000. Both will give you full access to a command line for Lame (or any other encoder for that matter). There is absolutely no need for you to convert FLAC to WAV (thus losing the tags) and then WAV to mp3 (thus having to manually insert the tags). Even the normal encoder settings for dBpowerAMP and foobar2000 are just more graphical approaches to using Lame. Moving the slider around just tells the program to use a certain command line with Lame. For example, sliding over to -V 2 means that dBpowerAMP will use the -V 2 command line. Using a graphical slider is no different than you manually putting in "-V 2" as your command line.

Lastly, I really don't know what you are trying to accomplish here. That website (and group/person) is notorious for bashing hydrogenaudio (where audio developers actually frequent, you can find posts from people who actually developed Lame here). They like to think that they know better than the developers. Well, I would take a post from a developer here any day over people who use commands that go against the general recommendations of the actual developers. That is like NASA going down to a Conoco station and asking the gas attendant for opinions regarding their new rocket propulsion system and how it will perform when re-entering Mars' atmosphere.

You can use MAREO + EAC to encode to FLAC and MP3 at the same time, using command line.

Dear, people
I'm studing IT and for me there is no problem to use command line, because sometimes I do more complicated things like programming. Thank you for the information about mp3, flac, but I would be more grateful about technical information of mp3 (better developer information).
f.ex. 1.which ranges (kbps) uses -V option?
2. Which -q (by default) mode will be used if I choose -V0 option?
3. --vbr-new is deafault in 3.98.2 LAME?
Don't write me about which is lossless codec, which no - I know that Odyssey, of course I want reduce the filesize, because I don't have 1TB HDD. Don't talk rubbish, that ~128kbit VBR will perfectly reproduce original music. I'm not going to use plastic fantastic made in china speakers, if you use such stuff maybe there is no diference between 128 and 256
Sorry I didn't know that jthz.com bashing this site.
kornchild2002, kwanbis thank you for software recomendation. dBpowerAMP's CLI mode I guess support encoding only one song or I don't know how to do that it will encode sequentially all songs from directory.

if you're interested in dev infos, simply check out the history log of lame. If this is not enough for you, then check out lame's source code.
But you're asking about informations for end users, not for developers.

to your example questions:

1. 32 - 320 in general.
2. -q 3 (qualitywise equal to -q 2 or -h)
3. yes.

if you want best quality in your case, simply set your lame setting to -V 0. That's all - The best quality setting available for lame is -b 320.

2. For Lame 3.98.2 and --vbr-new (=default) mode, there is no difference between -q 0 ... -q 4.
So there's no difference between -V 0, -V 0 -q 0, -V0 -h, ...

thx for explanation. The last thing I need that someone recomend me good software for reading mp3 headers

Not much of us here probably use crap speakers, and my statement stands no matter which equipment you have. You need to conduct ABX tests to prove your brain that you don't need overkill settings to achieve complete transparency.

Just to come with an example, I was going out of my mind once I could not distinguish a 32kbit HE-AACv2 track* from it's original using very good headphones in an ABX test. Don't judge a codec by it's bitrate, judge it with your EARS (using an ABX test).

* This was of course not just *any* track. If I recall correctly it was some electronic music, whereas rock-music was much easier to ABX at this level.

maybe your music style is techno boom boom and nothing else. I also use 128 kbps for that music style.
By the way, dbpoweramp uses stereo or joint stereo by default? because I have strange results. dbpoweramp shows (mp3 file information) that using stereo and winamp shows joint stereo. There is screenshot http://img268.imageshack.us/img268/4245/59436846.jpg

Don't ask for help if you don't want an answer...

I can't believe I even answer you again, but here goes.

Although you haven't read anything and haven't accepted anything either in this thread, I can tell you that LAME-developers does a very hard job optimizing the algorithms to perform best at standard settings - You throw away all this with all you fancy tweaking parameters believing you are better than the devs AND all the people with super-sensitive-ears here helping the devs optimizing hard samples. So remember all your settings does is reducing the quality compared to the default values.

That said, you still haven't conducted any ABX test and still believe you are able to hear mp3-artefacts at high bitrates and that I use cheap chineese speakers. I challenge you an ABX test of Enigma's "The Eyes of Truth" using lame -v 5 switch ONLY... Yeah that's right, maybe I'm not just a techno-freak after all, huh? That track should even be easy to ABX with it's complexity, right?

This is my results:

Using lame 3.98.2 anything up to -v6 is easy because of a lower lowpass (at least for me, but I'm also annoyed by 16khz sines in some songs - are you??)

I was unable to ABX -v5 with lame against it's source! I suppose you can do it, with your snakeoil cables and million-dollar equipment? At this quality level, the bitrate ends at 158kbit for this track. (According to foobar2000, at this quality level, the bitrate usually ends roughly at ~130kbit, meaning that the encoder itself had a hard time encoding it properly)

I tried Nero AAC as well. I was surprised that I was able to successfully ABX this track against the source all up to -q 0.4 (132kbit), although this setting gave me a hard time (I didn't try any higher).

Please don't try to be insulting to people when all they are trying to do is help you. You also happened to insult a genre (of which, I am not a fan) that recently lost someone. You insulted a genre of music, tried to insult a member here, and even insulted the country of China by implying that speakers coming from there are crap (maybe you should read the "made in..." label on the products in your house). odyssey is right though. There is absolutely no need for you to use -V 0 if your ears won't benefit from it. That is like killing a mosquito with a tank: overkill. The best thing that anyone can do, when it comes to music, is conduct a series of blind ABX tests to determine which settings is right for them. People who can't effectively distinguish 128kbps VBR mp3/AAC from the source lossless material are not deaf and they can still appreciate the music. People who refuse to conduct blind ABX tests and insult others who suggest doing so are arrogant and ignorant.



Yes, dBpowerAMP uses joint stereo by default. It is reporting that your audio has two channels in its little report because it still has two channels. You might have to wait for a comment from Spoon as to why they did this. It shouldn't matter though as joint stereo is being used, be default, whenever dBpowerAMP encodes with Lame.

ok I can apoligize saying that best speakers comes from Denmark Scanspeak, Vifa, Peerless - great brands Sorry I deleted Enigma flac files, so I can't accept your abx test

Joint Stereo is default in (afaik) all default lame-modes. That's because there are only benefits in using it. There was a wrong implementation on it many years ago in some encoder leading to the misunderstanding that Joint Stereo is a bad thing. You can search the forums and find very recent discussions on the subject.


"Deleted"? Surely you have the CD available, so you could just rip it again. I was under the impression that you wanted a lossless archive of your music and a lossy mirror useful for portable devices etc.?

It surprises me that you don't just store music losslessy, when you are so concerned about quality.


That's a highly subjective opinion. Personally I think my Dali speakers (which probably uses one of said driver-brands) is very good. That said, much other good equipment are manufactured in China, just like Denmark is not good at everything!

well i don't have Enigma CD I few hours ago I had Enigma dicography in flac which size was 2.5 GB. Now I have the same discography(mp3 -V0 option) which size 757MB. The size was reduced 3.3 times - for me it's great result. I don't want loosless because I have laptop with 160 GB HDD and there is no space for a lot of music, films, software and so on. If I have desktop computer with 1(or 1.5) TB I thought about keeping music in flac. Even hydrogen community says that 256-320 mp3 hard to separate from original source, so why I should use flac when I have limited space HDD

Well first of all I think you should support the artists that you enjoy in very fine quality. You won't get much respect here with discussion of pirated material. Second of all the cost of harddrives are so cheap that mp3-compression just doesn't justify it's advantage.

VIVAT thepiratebay ! artist are not such poor - they get money(despite cd) from concerts and comercials. I don't need anyone respect from this forum - I came here only for technical details of mp3 and thats all. Yep I know that nowdays hdd's are cheap, but I don't want to buy external or change internal hdd

Go troll somewhere else.

Thread locked.