The New Tutorial for beginners ... from www.ChrisMyden.com. I always have people asking me how to set up EAC and this guide will really help getting people on their feet... It's available for DL as a Word document too...

does anyone else get those mySQL-Errors at those links?

Thomas

does anyone else get those mySQL-Errors at those links?

I do now, but it was OK yesterday. The guy's main thesis is that EAC sometimes misdetects drive features, so you ought to manually configure it to be pessimistic.

Plus he gives a completely blow-by-blow account of doing so, which is kind of handy for when (say) your parents ask you how they "put a CD onto their computer".

Peter

The site is down today for some reason.....

It is realistic.. I was getting an occational "pop" or "static" when EAC reported no errors until I disabled my C2...

EDIT: The site is back up...

I read the whole thing twice and seemed to miss the part where Chris talked about encoding bitrates...He states "don't worry that the bitrate setting says 128 kb, since we will be using our own presets anyway, the displayed bitrate won't mean anything"

Trouble is, I don't see any explanation about where the bitrate setting is to be made. I assume that they're made in the LAME encoder settings, but again, I see no info on that either? Am I missing something or just being a little dense tonight?

scroll up a small a bit and look for "--alt-preset standard"

Thanks -

I did see that before, however what kind of bitrate do I achieve at a "standard" setting? I encode my files at 100% VBR presently...I'm looking for an average of 190-250 kbps to suit my own quality standard...

Since eveyrone seems to love EAC so much... I would expect that it would simply have a place to type out the LAME settings you want - no matter how specific. If so, I might consider using it... from what I've read on it so far it might make my life just a little easier.

I would expect that it would simply have a place to type out the LAME settings you want - no matter how specific.

Well, that was my point...Chris' tutorial doesn't mention what kind of bitrate you can expect. I hope he doesn't assume everyone will be happy with 128 or 160 kbps...I've been using MusicMatch jukebox for over a year now and I'm still happy with the quality of the encoding at 100% VBR. The larger file sizes aren't an issue. I would like to find a program that is faster and more accurate however. I was checking EAC out because I've heard/read so many raves about it...

where "--alt-preset standard" is placed you can use what ever LAME setting you want......of which there are many

a quick check shows Diana Krall's The Look of Love to be encoded with anything from 163 to 207 VBR - but then that will vary alot depending on the music

there is also a "--alt-preset extreme" setting, not sure of hte exact differance.

depending on how pickie one is one could pick different LAME settings for each type of music....I am sure classical music would be better encoded with different settings to death metal....but then this life thing tends to get in the way :-)

Hmm... I should have been more clear when I posted this...

The tutorial was writen as a bare bones setup guide. Basically, once you have convinced someone they should encode at "--alt-preset standard" with EAC, you point them to this setup guide. I've been in his chat room and this is the most common question, so this guide really saves him time. His chat room is "#myden" on DALnet btw.. He has a more detailed one that really explains all of the features here .

Without trying to speak for him, Chris subscribes to "--alt-preset standard" as the best setting. The files average just a little bigger than 192k files. They will max out at 320kps and wont go lower than 128kps for any given song. I think that they are better than Fraunhoffer encoded at 100% (MusicMatch)? Even the MusicMatch employees use EAC and LAME! (svferris can attest to this).

The --alt-preset standard over-rides any other settings. With LAME, you can go in and manually tweak those settings but I think the people that setup the presets have better ears than me anyway... There is also a "--alt-preset insane" or something, but you might as well encode at 320 CBR.

That's the beauty of LAME and VBR... it does a good job of annaylsing the music. (That is why the encode can get so slow). Some of my music encodes at real low rates (mono or simple stuff) and some of the files are simply HUGE (like NIN).

Hi guys,

If you would like to see how --alt-preset standard came about, visit the following thread at the Hydrogen Audio message boards.

http://www.hydrogenaudio.org/forums/showthread.php?s=&threadid=260&highlight=preset+standard

That thread also describes some of the other presets available for LAME, such as:

--alt-preset fast standard

--alt-preset extreme

The --alt-preset standard switch, invokes code in LAME that no other LAME switch uses, and I won't pretend to to try and understand the work done on the psycho-acoustic model that went into creating this switch.

This is what I can tell you about the --alt-preset standard switch:

- Files average out to about 192kb/s, using a variable bit-rate between 128kb/s and 320kb/s

- It was developed & tweaked to accomodate all types of music, and to fix problems with certain artifacts found in other encoders.

- It is the ultimate compromise between file size and quality. If you want perfect MP3 quality, encode at 320kb/s, if you want to save space, encode at 192 CBR. But if you want your music to get as close as possible to retaining perfect quality, while keeping file size as low as possible, that's what the group at Hydrogen Audio (Project Mayhem) set out to achieve.

- I can also tell you that the people working on the tweaking are not your average, run of the mill music listeners. They have some brilliant people doing the tweaking, and have performed double-blind tests using high-end equipment with people who have 'golden ears' that can detect minute flaws that us regular humans would never hear.

If nothing else, you may want to check out the message boards at HydrogenAudio.org they have some pretty interesting threads.

Chris Myden
http://www.ChrisMyden.com

Oh, and yes Brad, the LAME encoder surpassed the "Fraunhofer" codec a long time ago.

Even the --r3mix setting in LAME was superior to Fraunhofer at 192. (--r3mix was the switch used before --alt-preset standard)

When it comes to MP3 encoding, there is no other encoder that even comes close anymore. The LAME encoder is the Tiger Woods of the MP3 world, simply head and shoulders above the rest. And best of all, it's 100% free!

Chris Myden

Thanks guys

That clears it right up for me...and welcome Chris, I visited your site in recent days and really like what you have to offer there (thanks to Brad for the directions). I'll have to give EAC and LAME the big test this weekend then.

Even the MusicMatch employees use EAC and LAME!

Well, just one at least.

I use LAME because the VBR in MusicMatch (Fraunhofer, to be specific) seems to create an error that screws up the timecode in players (like empeg).

I also like EAC because it's highly configurable. MusicMatch isn't because it's geared towards the average user. On the same note, I generally use Winamp for my player because it's a very lean application. If I need all the bells and whistles, then I open MusicMatch.

I would hope that it was corrected by now - but a while back MusicMatch came bundled with a piece of sotware I had purchased, and I decided to try it out. This was long before I actually wanted to rip CDs and encode mp3s... regardless of the reason. No matter what I did with MM's settings, it would always freeze up (the whole system - Win98SE) on mixed mode CDs (CDs with an ISO track after audio tracks). I gave up on it quickly.

Hi Chris,

i'm actually ripping my CD-collection using EAC + lame 3.92 --r3mix.
Should i change to --alt-preset standard?
I decided to use --r3mix because of the vbr=1 instead of the vbr=2 in standard.
Am i wrong? What about the lopass-filters?
Can you give a short explanation? I have sadly not the time to read through your page...

Thanx in advance,
Thomas

I'll tell Chris about your post...

But from what I understand, the -r3mix standard was the best you could get for quite a while. However, it hasn't been worked on since prior to Lame 3.91 or so.. Instead, all development has been going into the newer --alt-preset standard. There is nothing wrong with -r3mix, but you might as well switch over to standard and see if you can even notice the difference. My guess is that with most music you wont.

Chris could probably explain this better (as with other members of this board).

Hi Thomas,

Yes, I would highly recommend using --alt-preset standard instead of --r3mix

--r3mix is the outdated command line used to achieve the best VBR quality. --alt-preset standard is what replaced it, and invokes certain code in the LAME encoder that other command lines don't use to achieve higher quality.

There is a huge quality leap between --r3mix and --alt-preset standard with very little file size increase.

You won't find anyone on any LAME, MP3, or EAC message board recommending --r3mix anymore, it's simply inferior in every way, except for speed. If you have a slow computer, --r3mix encodes faster.

Don't get me wrong, --r3mix was great, and by using it, you're probably making better MP3s than most people ever do, but --alt-preset standard uses a far superior psycho-acoustic model, and has been tweaked to fix artifact audio problems, and deal with every type of situation to make sure the audio is encoded properly.

I'm assuming you'll be listening to the MP3s on your Empeg, with your car audio system. Most people might never be able to honestly tell the difference between --r3mix and --alt-preset standard on their Walmart computer speakers, but with decent car audio equipment you definitely want to make sure you're encoding with the best available.

Hope this helps,

chris

Good point, Ineedcolor.

I should have mentioned what you would expect. It varies, depending on the complexity of the song.

I find older, less complex (vocal, little instrumentation) types of music usually average out to 140-160kb/s

Your typical Top 40 music of today averages out to 190-210kb/s.

Just now, I used MP3 Manager 32 to list the songs I have that came out with the highest average bit rate using alt preset standard.

Here are the top 5:

"Is Anyone Home?" - Alice Cooper - The Life And Crimes of Alice Cooper - 253kb/s

"Black Black Heart 2.0" - David User - Morning Orbit - 253kb/s

"The Night That Minnie Timperley Died" - Pulp - We Love Life - 240kb/s

"Bob Lind (The Only Way Is Down" - Pulp - We Love Life - 240kb/s

"Record Companies" - Tricky - Angels With Dirty Faces - 237kb/s

I also know from experience that Aerosmith, Weezer, and Fatboy Slim come out with pretty high bitrates.

chris myden

Hey Chris

I encoded my first CD today with EAC/LAME using --alt-preset standard, and all of the other settings as you recomended in your tute. The CD was "Born Again" by The Notorious B.I.G....

In the car I found the high ends brighter than usual and the bass a little lacking (I have a high-end sound system). In all fairness I have my EQ presets set up for the files I ripped using MM Jukebox...As I listen mostly to Rap / R&B / Hip-Hop, I find that I tend to like a deeper bass.

I re-ripped the CD using MM jukebox and I did find the bass was deeper and the high end had a little less edge the second time around. I feel that maybe because EAC/LAME does such a good job of compression while still retaining excellant sound quality, maybe it's cutting some of the top and bottom off to compensate? My MM Jukebox files average about 5-7MB in size. Over time, I have found that as compression increases, bass tends to disappear and treble fill with tinny artifacts.

Are there any LAME settings I can employ that would give me a fatter bass and take the edge off the highs? I'm not too keen on tweaking my eq again as I just found the "sweet spot" I really like recently.

As a sidenote, I also found that while EAC will extract audio relatively quickly, the LAME encoder seems to go on in the background forever...I sat and watched 18 .wav files slowly become .mp3s over a ten minute period...I didn't want to perform any other activities because I was afraid of corrupting the compression...I know my processor was labouring because my temperature alarm went off, and that's with three case fans and a golden orb on the chip ha ha. BTW, I'm running WinXP, with a Pentium III 1 GHz processor with 1 GB of RAM.

Thanks -

The fact is, the opinion of the vast majority of users is that Lame produces mp3s that sound more like the actual cd than any other encoder. If the files sound different to you, I can only assume that MusicMatch is inferior and that you have adjusted the eq to compensate. You might want to throw a straight wav file on the Empeg and see if it too has brighter highs and lacking bass. My guess is that it will.

Actually I would say that EAC and LAME does not retain enough music information to make the music sound the way I want it to at the settings I used. At --alt-preset-standard, the files I produced were typically 4-5Mb in size whereas the MM files I produce are 5-8Mb in size...therefore to me, bigger file size = more music information = better sound. Audio compression means that some information is going to be lost during the process and as compression increases, even more information is dropped. As I stated above I have found that bass and treble are usually the first things to disappear as you lower the bitrate and listening becomes intolerable below 128kbps. I am the only one I've seen on this board who uses 100% VBR according to the various discussions I've read on this topic. Maybe if someone did a direct comparison of the two programs, would they notice a differance as well? I realize that everyone raves about EAC/LAME however, have they tried other rippers? Are they just being satisfied with what they hear figuring that it is quality because someone else told them it was? I think that it is really a subjective opinion based on an individual's own perception of quality.

When I first started ripping music (I have 40Gb now) I did exhaustive research to try and find the sound I was looking for so I would not have to re-rip my collection a second time. I have read so many posts here about people doing their collection again because they were not happy with their own results the first time around... You mention that I may have adjusted my EQ to compensate for an inferior compressor...well, I have no equalization in my home theatre set-up and my music sounds just fine to me. When I have friends over, most times they aren't even aware that it is compressed digital audio they are listening to, they always assume that it's my CD changer...so I imagine that I'm doing something right....

So I drop the gauntlet to you to listen to my system at the Cincinnati meet....and I'll even include the same songs encoded with both programs so you can compare for yourself.

At --alt-preset-standard, the files I produced were typically 4-5Mb in size whereas the MM files I produce are 5-8Mb in size...[snip]...bass and treble are usually the first things to disappear as you lower the bitrate and listening becomes intolerable below 128kbps.

Are you sure your version of Lame is up-to-date (namely 3.91 or 3.92)? I find that files made with lame --alt-preset standard are usually about 220Kbits/s average, and never below 180Kbits/s average. It's simply not possible to make files eight-fifths as big as that, as MP3 only goes up to 320Kbits/s...

Peter

maybe some one can bring a computer to the Cincinnati meet and a celection of CDs can be ripped and encoded with different programs and encoders and settings

Hi Peter

I confirmed it this am, I am using LAME v3.92...I had only just downloaded EAC and LAME a few weeks ago so I assumed I had the latest versions anyway. I followed Chris Myden's tutorial to the letter because I wanted to give it the best shot. It is possible that there is something wrong in the setup or somewhere else in the system but I don't think so. I was actually very suprised to see those file sizes after I completed my first rip, I really would have thought that they would have been bigger given the bitrate. Maybe LAME itself is acting up, it took forever for the wavs to be encoded after they were extracted..... I will look at the set up again and see if there is an erronious setting somewhere I may have missed.

As I said, your best means for comparison will probably be to compare mp3s from both MusicMatch and Lame with straight wav files. Personal experience has proven MusicMatch to be noticeably inferior to Lame in my own listening tests. Keep in mind that when I say Lame produces files of superior quality, I do not mean files that sound better, I mean files that are more often indistinguishable from the original cd.

According to my personal notes, I perceived a difference between MM at 100% VBR and the original source 4 out of 10 times. Of those times I was correct in my identification of the MM encoded file 90% of the time. In comparison, I perceived a difference between Lame using --alt-bitrate-standard and the original source 2 out of 10 times, and was only correct in my identification of of the Lame encoded file about 50% of the time, which shows that my detection of a difference in that case was very likely only a shortcoming of my ears rather than of the mp3s.

If you are as anal about such matters as me, and have a friend to help you, I would suggest a series of blind listening tests. Only proving it to yourself will convince you.

And keep in mind that if you can really tell the difference between Lame at --alt-bitrate-standard and the source material, there are still the -extreme and -insane settings.

when using LAME with the --alt-preset standard setting it will take a long time to endode
so you are not in error, it is just that quality should not be rushed

well, I am all for giving it another go, I do like to experiment with new software anyway ...I'll keep you posted on how things turn out....thanks for your feedback.

How are you playing these on your home system? With the empeg? If so, then EQ will still be a factor. Also, bass doesn't require much data... but high frequencies do. In fact, I think I remember svferris (a MusicMatch employee) saying that there isn't a setting in MM that has NO high frequency cut off.... I'm assuming that MM cuts the highes so it makes the LAME files sound overly bright. Who knows, it might even boost the bass curve?

I don't understand the file size thing tough... Can you post the file sizes of two compressions of the same song? And get the average bit rate too..?

Hi Brad

When I play music through my home system, I primarily use Winamp and to a lesser extent, Windows Media Player. I use no equilization on either player so I am sending the music to my home system "as is". I am currently using a US Robotics Sound Link (to be replaced with a Rio Reciever soon) to transfer the music over a distance of about seven metres.

Today, after this latest round of discussion between MisterBeefHead and I, I played one song that was kinda bass heavy encoded with both EAC/LAME and MM Jukebox and the source CD through my CD Changer. Doing the comparison, I again found the bass and treble variation with the EAC/LAME file, and virtually no differance between the source CD and the file encoded with MM. The bass was much deeper and fuller with the source CD and the file encoded with MM. I don't know where to go from here other than to experiment with the EAC/LAME settings to try and find that sound I desire.

I will post the two file properties here tonight for you once I get home from work....

As I stated in another thread...

You can't compare apples to apples when you're playing an audio CD compared to a computer file, if you use a computer to play back the computer file. The computer and the audio CD player have different circuitry to produce the sound.

The audio CD player itself introduces its own tone coloration to the audio signal. For instance, most CD players artificially boost the bass before it even leaves the player.

The only way to properly compare an encoder to the audio CD is to decode the encoded MP3 to a WAV, then burn that WAV as an audio CD track. Then you can accurately compare the original audio CD track to the MP3 encoder in the proper apples-to-apples environment.

I think I should make this into a FAQ entry with pictures. Anyone think we'd benefit from such an entry?

Myself, I ripped the wav file, and then encoded it several ways, and then threw the wav and the mp3s all onto my Empeg. I then did listening tests in my home and in my car. I thought about burning the mp3 to a cd, but I thought that I once read somewhere that different decoding programs can result in somewhat different wav files. Rather than pursue that possibility, I just used the Empeg.

I ripped the wav file, and then encoded it several ways, and then threw the wav and the mp3s all onto my Empeg

This works, too. Doesn't convince audiophiles who refuse to believe that the perfect sound of their favorite CD player is the only way to listen to music. Generally, whenever I hear the complaint that the "MP3s don't sound like my CD player", what they need to be convinced is to hear the MP3s played back through that particular CD player. Only way to do that is with a decode-to-disk and a burn.

Everyone, check out this link on the variences in decoders.... how do you think this factors into the equation? Where would the empeg fall into this?

Haven't followed the link yet, but I assume it's talking about the old sample-precision-in-the-decoder issue. I remember Hugo discussing this point with someone here on the BBS last year, but I can't find the link to the thread. Anyone have that?