Lame MP3 Codec Suppresses Audio - Seems to be best only for music

[Typhoon859]

Well, for music, the Lame MP3 codec seems to work best to reproduce the sound quality but I'm not so sure that it is for something like the audio from Dragon Ball Z, especially the 5.1 surround track... It seems to suppress it's surround feel (obviously for the 5.1, lol, but I mean for the stereo as well.).

What codec is known to have reasonable compression but with less suppression of the sound like MP3 has?

[poisondeathray]

You can try AAC, it generally offers better compression (so better quality at the same bitrate compared to MP3) on most sources. It also supports 5.1 audio

[Typhoon859]

You can try AAC, it generally offers better compression (so better quality at the same bitrate compared to MP3) on most sources. It also supports 5.1 audio

Hmm, yeah, that's what I thought. How well does AAC work with converting 5.1 surround to 2ch stereo though? The file size would be much greater converting to 5.1ch AAC vs. 2ch and I don't think that it would be worth the larger file sizes in my encodes.

[Inspector.Gadget]

I realize this isn't HydrogenAudio, but do you have any evidence to back up your "suppression" claim or are you just running your mouth like you did in saying "MeGUI is for noobs"? There are very few things as irritating as people spreading spurious claims about software other people have put long hours into, refined to the point that LAME is the highest-quality VBR mp3 encoder available, and then released for free.

[Typhoon859]

I realize this isn't HydrogenAudio, but do you have any evidence to back up your "suppression" claim or are you just running your mouth like you did in saying "MeGUI is for noobs"? There are very few things as irritating as people spreading spurious claims about software other people have put long hours into, refined to the point that LAME is the highest-quality VBR mp3 encoder available, and then released for free.

I don't remember me ever running my mouth. I am the last person you should say this too. I even over-appreciate things people have made sometimes, no matter what it is, just for the work put into it. I really like the Lame MP3 codec... As I said, I just don't think it works as well with like cinema type audio and mostly good for music. The MeGUI thing for noobs was just a joke.. and no claims have been made here. I simply expressed my opinion and asked for yours in return, I guess you can say. So.. I got one. AAC. I'll try it out and see what happens. Sorry if what I said came off wrong.

[neuron2]

I just don't think it works as well with like cinema type audio and mostly good for music. Where's the evidence for that? If you can't show us any, then you are indeed just "running your mouth".

You started a whole thread about this effect, but you can't demonstrate it?!

[DJ Bobo]

Typhoon, how are you downmixing from 5.1 to stereo? I think that's the problem here.

By the way, if you use Vorbis -q2, you'll get something between 80 and 88kbps and it will sound better than 128kbps MP3.

[Typhoon859]

Where's the evidence for that? If you can't show us any, then you are indeed just "running your mouth".

You started a whole thread about this effect, but you can't demonstrate it?!

Typhoon, how are you downmixing from 5.1 to stereo? I think that's the problem here.

By the way, if you use Vorbis -q2, you'll get something between 80 and 88kbps and it will sound better than 128kbps MP3.

Neuron2, evidence? How can I post evidence? I just gave my opinion. The reason I did was to explain the reasoning behind my question.

DJ Bobo, yeah, I was thinking about that. You convinced me to try it now. Maybe it was too early to post anything before I tried more things on my own. I just thought I'd take a shortcut and get recommendations here. But uh, thanks.

[neuron2]

Neuron2, evidence? How can I post evidence? I just gave my opinion. You have arrived at an opinion presumably by taking a source audio clip and encoding several ways, and finding that one is better than the other. So at a minimum, you could simply post the source clip and tell us how you encoded it with all the settings you used.

It's not rocket science.

If you don't back up your dubious claims you'll quickly develop a bad reputation here (it may be too late to avoid that). You've got a handful of threads like this going, where you impugn some software for silly reasons that only expose your ignorance.

"Using Ignore Pulldown speeds up the video?!" It's laughable.

You've been cropping and resizing telecined video without doing IVTC first. That's extreme newbie silliness.

[Typhoon859]

You have arrived at an opinion presumably by taking a source audio clip and encoding several ways, and finding that one is better than the other. So at a minimum, you could simply post the source clip and tell us how you encoded it with all the settings you used.

It's not rocket science.

If you don't back up your dubious claims you'll quickly develop a bad reputation here (it may be too late to avoid that). You've got a handful of threads ike this going, where you impugn some software for silly reasons that only expose your ignorance.

"using Ignore Pulldown speeds up the video?!" It's laughable.

I'm a noob. I'm trying to learn. Wow. I don't even know what this Field Operation stuff is. I accidentally came across it and decided to switch it and it worked. Too bad curses aren't allowed here because there are plenty that are very fitting right now. My threads are questions. My threads are me trying to learn something and figure out what's going on. I explained the effect that Ignore Pulldown gave on my video. What's so "laughable" about that? Very VERY VEEERRRYY few people even know what bitrate or framerate, or even what codecs are. People don't know that MKV or AVI or MP4 files are just containers. I'm trying to learn and I don't understand what your problem is. Come on man, honestly. Don't be a smartass.

[neuron2]

Field Operation is very clearly explained in the DGIndex Users Manual. We expect you to read manuals and do some researching before posting questions. It's forum rule #1, in bold font!

http://forum.doom9.org/forum-rules.htm

In this forum, we are technically advanced. We know what a codec is and what bitrate means. Posts like yours either bore us or amuse us. So far you've amused us, but now...

BTW, are you going to back up your claim in this thread, or just hope that we stop asking you to support your claims?

[Typhoon859]

Field Operation is very clearly explained in the DGIndex Users Manual. We expect you to read manuals and do some researching before posting questions. It's forum rule #1, in bold font!

http://forum.doom9.org/forum-rules.htm

In this forum, we are technically advanced. We know what a codec is and what bitrate means. Posts like yours either bore us or amuse us. So far you've amused us, but now you're becoming boring.

I didn't think it was something that would be in the manual. And if you're already going to tell me that, then you can say that to everybody that didn't ask about something completely new in this forum, because almost everything can be read up on/found somewhere, especially considering that asking for opinions isn't allowed.

-I know what a codec is and what bitrate is and much deeper than that... And since you all are so technically advanced, that's exactly why I'm asking you guys. Find me something that I asked about that's completely inexperienced/generic stuff that can be found anywhere. This is really the only thread which is a bit general but it is THE ONLY one and I realized my mistake with this very soon after I posted it.

[neuron2]

I didn't think it was something that would be in the manual. You didn't think the user's manual would describe how to use the program? You saw an option in the menus and thought it would not be described in the thorough users manual that accompanies the program?

Have you now withdrawn your claim about the Lame MP3 codec?

[Typhoon859]

You didn't think the user's manual would describe how to use the program? You saw an option in the menus and thought it would not be described in the thorough users manual that accompanies the program?

Have you now withdrawn your claim about the Lame MP3 codec?

I didn't claim anything. I'll do some more testing (because honestly I haven't done much with audio), but from what I've heard, it works well with music but no so much for movie audio. You probably did comparisons yourself but if you want, I'll post some comparisons between different codecs and the original. I hope you have 5.1 surround speakers...

And no, I didn't think that the user manual would explain what I asked in that thread.

[neuron2]

I didn't claim anything. Now it's clear that you are disconnected from reality. Your first post in this thread clearly states that you think the Lame MP3 codec "suppresses audio" and you ask for recommendations for a codec that doesn't do that.

I'm going to stop responding to you now, which you'll probably welcome. But don't forget about our forum rules, because I will still be reading your threads. Bye.

[Typhoon859]

Now it's clear that you are disconnected from reality. Your first post in this thread clearly states that you think the Lame MP3 codec "suppresses audio" and you ask for recommendations for a codec that doesn't do that.

I'm going to stop responding to you now, which you'll probably welcome. But don't forget about our forum rules, because I will still be reading your threads. Bye.

"It seems to suppress it's surround feel". You made such a big deal out of that. And yes, I still think so. Don't make an elephant out of a fly.

[Inspector.Gadget]

I didn't claim anything. I'll do some more testing (because honestly I haven't done much with audio), but from what I've heard, it works well with music but no so much for movie audio. You probably did comparisons yourself but if you want, I'll post some comparisons between different codecs and the original. I hope you have 5.1 surround speakers...

And no, I didn't think that the user manual would explain what I asked in that thread.

1. You did make a claim, which you should retract if you can't support it in the interest of not spreading misinformation and not pissing off the software developers that make useful software possible.
2. You heard wrong. Completely. LAME performs well with a 48kHz stereo PCM source like you might obtain from converting AC3 to WAV or piping PCM via stdin to LAME. It's ridiculous to compare "music audio" to "movie audio": what do you think about music videos on DVDs? Which is it? It's a worthless, arbitrary comparison that tells us nothing about the relative characteristics of samples that a given encoder may or may not perform well on.
3. 5.1 speakers don't matter in a stereo listening test. MP3 - with the exception of the Aud-X derivative that doesn't adhere to the MPEG-1 and MPEG-2 Layer III specifications - does stereo (dual mono), joint stereo (usual mode, uses correlation across channels to improve stereo encoding efficiency), and mono. That's it. Assuming you're using a bitrate that's sufficient for transparency in each case, encoding from the same PCM source, avoiding rogue quantization errors, and not using ill-advised highpass or lowpass filters, stereo mp3 is indistinguishable upon listening from stereo AAC, FLAC, WAV, etc. Any "suppression" is solely the result of your downmixing algorithm in normal encoding situations without problems.

[Typhoon859]

stereo mp3 is indistinguishable upon listening from stereo AAC, FLAC, WAV, etc. Any "suppression" is solely the result of your downmixing algorithm in normal encoding situations without problems.

Could you rephrase that? It sounds interesting but I'm not quite sure I'm understanding it right.

As I said, I didn't have too much experience testing audio. I might've heard something else/maybe there was some issue/..I don't know... But, again, I didn't make a claim. I said something that I thought I heard right a while back which, again, was "It seems to suppress it's surround feel". If I was wrong in saying so because of how god-like LAME apparently is, then you or whoever could've just corrected me. I would've taken it. What I'm disagreeing with is what you guys are doing. I'm not taking back any *&@%!#$ "claim" because I didn't make any. Correcting what I thought is one thing. This on the other hand is just harassment.

I AM interested in what you have to say and value what you say without you needing to give me justification. I would simply look into what you said if it were relevant to anything I was doing and test/check for myself.

Now, you seem to know a lot. I'd like to hear more of what you have to say. For instance, from what I see, DVD audio are AC3 files. I assume that official releases all have this format. If not, doesn't matter - point is, the ones I'm working with do. At 192kbps, the MP3 conversion turned out larger in size than the original stereo AC3 file. I can upload both if you want, for comparison. I actually compared the two and my opinion remains. The sound quality is good! But, aside from the fact that the file is bigger, the sound is more abrupt sounding. It's less dynamic. I checked it out between 19:00 to 19:30. I have really good speakers. With worse ones, ANY sound suppression would sound great in comparison.. Exaggeration but, you get what I mean. That is why I asked for what I did in this thread.

I've been working on other things so I didn't have time to try out AAC or Vorbis. In any case, I'd like to hear what you have to say and what your opinion on all of this is.

*CDs have a different audio format so that could partially be the reason why I feel that it works better for music. Another reason is because, mostly, sound doesn't stop flowing in music and isn't interrupted like after the hesitations between words that people are saying in a movie or w/e. Also, most music is synthesized after the recording anyway, so a bit of difference in the way it sounds may not be as noticeable when just listening to it. People talking on the other hand is a whole different story.

[Inspector.Gadget]

1. "It's less dynamic." OK, let's see some ABX testing.
2. Why would you transcode AC3 to MP3 at the same bitrate? Lossy to lossy transcodes ALWAYS result in a loss of quality. That's not a valid measure of flat-out quality like (for example) encoding the same 48kHz wave file to AC3 and MP3; the only thing you can measure there is whether you can distinguish a lossy-sourced mp3 from the original file. At 192kbps and higher, probably not, if HydrogenAudio ABX results are generally reliable. Again, ABX testing is the only way to be sure.

Again, you keep saying "In my experience, X" and then saying you're not making any claims. Fine, use that sort of semantic dodge, but at least provide some objective evidence other than "Huh. I feel like being contrarian today, so I'm going to say demonstrably false things about a subject that I have little experience in."

[burfadel]

Most people use old (or even ancient) versions of LAME. They don't realise that its still under development! The latest 3.99A1 daily builds work really well - mind you they are test versions but for me they always work perfectly.

Ensure in the settints -q is set to 0, not the default 2. 0 is higher quality, and although it supposedly isn't perceptible since its not really any slower these days is good to set it. -q shouldn't be confused with -Q, the former is algorithm quality, the latter is VBR quality (the lower the better quality/higher bitrate for Q).

The downmixing itself isn't done by LAME? isn't it done by the supporting software?. Even if it can be done by LAME you could use Besweet (and update is various bits that have been updated by other people) and a besweet frontend such as Belight etc and downmix through there.

[neuron2]

If I was wrong in saying so because of how god-like LAME apparently is, then you or whoever could've just corrected me. I would've taken it. What I'm disagreeing with is what you guys are doing. I'm not taking back any *&@%!#$ "claim" because I didn't make any. Correcting what I thought is one thing. This on the other hand is just harassment. Struck and closed.