I've been using LAME 3.90.3 and EAC (with the recommended settings provided here at hydrogenaudio.org) for over a year now and have been quite pleased. However, I decided to try out Wal-Mart's downloadable songs and I honestly have to say that they sound better than what I can rip. So my question is: do commercial vendors of songs use some type of special mp3 and/or wma encoders that provide higher sound quality than what LAME can provide? If so, is LAME 3.96 going to be able to offer the same quality as commercial vendor downloadable songs?

they might sound better to you. nicer. warmer. but i'm pretty sure, your aps mp3s still sound more like the cd.

Hmm, I highly doubt commercial distributors use an encoder better than Lame, simply because there isn't one. Three hypothesis, in order of likelihood (IMHO): 1) Placebo: You THINK the commercial source sounds better because you haven't compared them in a proper blind test and simply imagine the difference. 2) You are one of those who find artifacts "agreeable". 3) The commercial distributors are not ripping the songs from CDs, but from another, better source that ends up sounding better than self-rips-encodes from CDs, even despite the possible artifacts of a worse encoder.

You have to prove that they sound better when you make these claims. See TOS #8.

Edit: it was my turn...

You're not supposed to post things like this without proof, so you should either post some or retract your statement. No sense argueing "facts" until they're proved as facts.

Wal-Mart uses WMA Std, not MP3.

Maybe WMA is really an acronym for Wal Mart Audio

There is one thing that I don't understand, are you comparing the same song, downloaded and ripped, or downloads in general vs rips in general ?

i think it maybe to do with the source, record companies use high quality reel to reel recorders at high recorder speeds.contary to believe digital is not superior to analogue.i bets that will start a long debate

I'm glad I didn't inadvertently post something like that when I first started coming here... I might have been scared off before I could learn more.

When relatively new members come by and ask questions like this, it's an opportunity for education that is wasted by simply dropping a TOS #8 link and being done with it.

Dunno if anyone here has read TOS #8 lately, but it's not very explanatory as to *why* it is done. It took me a while to dig for that answer in the forums.

It would be good to give newly registered posters the benefit of the doubt and provide, in addition to the TOS #8 link, another link like this one which clearly explains why TOS #8 is important and needed on this board.

I was very happy to read Dologan's answer--it is clear, informative, and makes the effort of educating an uninformed poster. It acknowledges the clear fact that it might very well sound better *to him*, but that's not the same as saying it sounds *closer* to the original source, which is what double-blind testing helps determine with a minimum of bias.

You're new here, aren't you?

That can't make MP3 sound better, since MP3 outputs 44100 Hz 16 bits anyway.


I completely agree with you. Thank you for reminding the existence of that thread.

so who's going to be the first one to maken an analog codec?

One possible reason for differences (pretty obvious to me) I haven't seen in this thread:

Maybe the songs WalMart offers are taken from CD but are processed before encoding, e.g. equalized. In this case they would sound 'better' then the CD version of the same song. IMO it would be interesting to find out more - so...

Solase, in case you have downloaded a song from WalMart and have it on CD too, please compare both, ideally using some ABX tool like foobar2000's ABX comparator. If the WalMart version sounds better (or if you don't have both versions), could you please upload a < 30 seconds sample? I don't know if these songs can be split and/or can be decoded regularly, so here's how it should work in any case: Get totalrecorder (free trial is enough) and install it. Start it, play back a downloaded file (with the audio player you use normally) and capture the output with totalrecorder to .wav. Use some wave editor like Nero Wave Editor or Audacity (free) to cut < 30 seconds from the .wav, save it and ideally convert to some lossless format like flac. You can use Hydrogenaudio upload forum for uploading. If you could upload the same range of the CD version too, it'd be perfect, otherwise - if it's a well-known song - someone else here will hopefully have it.

Depending on the definition, there are many things that could count as analogue codecs. How about dbx or Dolby? pre-emphasis curves? etc

All take advantage of some basic property of human hearing to put a slightly better quality signal into a certain signal space than would otherwise be possible.

That's what our modern digital codecs do, but much more efficiently.


In answer to the original question: I agree with all of the above. You're imagining it, you like bad sound, or the source is actually different in some way. I wouldn't say the source must be better - maybe it's just (even) louder. Or maybe it's more carefully (re)mastered before encoding. EDIT: like tigre said!

Maybe you just happened to download well recorded albums, but rip poorer ones?

Cheers,
David.

maybe we should start adding links to threads to de TOS?

Wow - I'd forgotten that one.

I thought I'd go back to this one:
http://www.hydrogenaudio.org/forums/index....topic=10766&hl=

but unfortunately I've been very bad at keeping my own advice.

Cheers,
David.

Do you mean adding links-to-thread into the TOS, or adding links into threads, pointing to the TOS ? Anyway, I've just done the first
Thanks for the suggestion.

the first one (adding links-to-thread into the TOS) ... you are wellcome

Actually, AFAIK, the professional reel to reel tapes and recorders are supposed to be able to capture audio at a higher "resolution" than most professional direct to digital recorders. However, I still think digital is better, overall, as it does not degrade over time.

[Off-Topic]
Just so you know, its taking everything I have to not go on a long, very off-topic, rant about Wal-Mart, sweatshops and evil business practices.
[/Off-Topic]

As for the original question I thought I would bring up the possibility of hardware audio encoders. I don't have any idea how these things perform, or if microsoft has a WMA capable encoder, but they certainly seem to cost a load of money so they must do something well, right? Not that I could possibly conceive of an encoder that could beat --aps with a bitrate of 128kbit/s (I assume thats what the evil company [WM] uses).

Another possibility, maybe when he said "Recommended Settings" he was referring to --alt-preset 128, the ABR presets are included in the List of Recommended Settings.

edit: typos

I would back this up as the most realistic hypothesis regarding different sound quality of downloaded songs ... e.g., the Nero Wav Editor has this neat feature of analyzing present HF content and then adding ''virtual'' HF harmonics (with adjustable crossover and level) ... I tried out this effect on some early 128 kbps Xing MP3's (you remember - the ones that sounded like they were played back underwater) and during preview, I was quite impressed by the "improvement" (personal opinion, of course) that took place.

Maybe they use somthing similar ... who knows

Edit:

It could as well be possible that they could get a hand on some re-mastered material that will most likely sound different than the release one has at home - that would imply comparing the same track, of course ... anything else wouldn't make much sense.

I think that's exactly what is happening. LAME with the recommended settings does not sound good. Indeed, if it sounds at all then it's bad. It is supposed to be transparent, not "good". I think a lot of people on this board would be very dissapointed if LAME (or any codec) made their music better.

What I bet Walmart are doing is optimising the songs (mainly through equalisation but possibly even remastering) to sound "good" on typical mid-level consumer sound kit. Radio mixes are optimised for low quality and car listening (as a gross generalisation) and CD mixes are designed (hopefully) to be an exact copy of what the artist had in mind. I would imagine that what Walmart are selling is a Cheap Consumer Hifi mix.

Basically the Walmart versions sound "good" but that does not imply "better than the CD" or "more transparent than LAME". It's an entirely subjective thing (like saying Limp Bizkit's "Behind Blue Eyes" is better than The Who's original).


Good reminder to buy new Asbestos pants. Thanks.

OT correction : if it is not the temporal resolution (192 kHz for digital against at most 100 kHz for analog), or the level resolution (the 24 bits of digital is beyond the resolution of bare cables because of thermal noise), then I don't know what "resolution" it is.
I'm not saying anything about analog vs digital, I'm just correcting the old myth of analog being superior because "it records the signal between the samples", which is wrong. Its resolution is not good enough.

Sorry, I used resolution for lack of a better term, thats why it was in quotations, I should have qualified the statement when I made it. Professionals do seem to pick analogue for that original master encoding far more than digital, there must be a reason for this.

"i think it maybe to do with the source, record companies use high quality reel to reel recorders at high recorder speeds.contary to believe digital is not superior to analogue."

Not true. I've been involved in both pro and semi-pro recording for YEARS. And it varies from one recording to another. In the early days of digital, it was common to see the letter AAD on a cd, meaning analog recording, analog mixing, digital mastering. And true, many early digital recordings were sonic abortions. But those days are LONG behind us. In recent years as a/d and d/a converters have been tweaked and improved to the nth degree, digital recording has completely eclipsed analog recording in terms of sound quality. Digital has evolved to become the gold standard of audio recording. Today the DDD designation is the most coveted designation. And with the advent of Pro Tools (and other high end software suites), real time digital recoding in a non linear/non destructive fashion has become the rage. Storing the recording to a computer hard drive gives the engineer practically unlimited flexibilty in terms of sound editing and processing. And the sound quality of current digital recordings is pristine. Given a solid engineer with a good ear, you can be certain the resulting recording will blow away any analog recording any day. Vanishing noise floor. Practically zero distortion. Frequency response from dc to light. Digital has come a long way baby. All of the finest recordings you have heard over the past several years have been DDD. It took some time to get the bugs out and for engineers and producers to "learn" how to work with digital, but now that it's come through it's infancy, digital is the edge of the art.

So, the short answer? No...record companies do NOT use high speed analog open reel machines anymore. Most modern recordings are mastered and archived in digital format.

Many bounce digitally recorded music to tape just to get that tape compression feeling when things are sounding too stark.

Exactly - they like the compression effect as an artistic effect. It's not more accurate - and if you didn't want that sound, you'd never dream of bouncing to tape. The whole point is recording to tape sounds different - and you can get the effect by recording and mixing to digital, and then just bouncing to tape! This shows there's nothing inferior about good digital, or anything superior about good analogue. (Though I still think a lot of LPs sound better than CDs <ducks>)


Hardware audio encoders are expensive because people using them for broadcast do not want to rely on a PC if they don't have to. Though some of these things are almost "PCs in a rack mount unit", most are carefully designed dedicated hardware. This is actually bad for sound quality, because they're harder to update and improve. But it's very good for reliability.

Cheers,
David.

unfortunately all this interesting discussion, and the original poster seems to have disappeared

Typical.

In this thread, I casually tested Fugue.wav with Gogo-no-coda, FhG ACM and Lame 3.90.3 & 3.96b1, and yes, I found FhG, which is a commercial product, superior to Lame at 160kbp/s. My $0.02.

I also remember Guruboolez saying that FhG vbr100 is much better than Lame 3.90.x --alt-preset standard here.

Yeh, on that sample. surely you aren't suggesting that it is better in most cases though.

Too bad digital mastering has gone completely down the tubes. Most CDs are cut so hot now that there's almost no room left anymore for transient peaks. And clipping is now very common with recent CDs as well.

I make it a policy to avoid buying CDs released after about 1995, and usually buy only used. Less money wasted that way.