[Ooztuncer]

OK, It's been so many hours to figure out what way to follow and HOW to follow. I believe I need some help.

For last 2 days I have been reading, trying, and more reading. I have a 30gb creative zen xtra and decided to put my small music archive into it - around 100 cd's + some mp3's from net.

My journey started with windows media player, converted 15 cd's from wav to wma lossless and figured out that zen xtra doesnt support flac or wma lossless formats - first bumper with a direct complaint to creative in their forums.

Then found out that EAC + lame mp3 conversion offers best conversion to mp3. At this point, I decided to use "b320 vbr new..." comment and I believe that this would give me better results than WMA.

I actually tried 5 versions and listened them in my zen with PA2V2 amp and SR-80 / Senn HD201:
1) v0 vbrnew
2) v2 vbrnew
3) v0 (default vbr)
4) v2 (default vbr)
5) b320

Test music was 5th song from Jaco Pastorius's debut album from 1976 (digitally re-mastered) - just a bass guitar and piano. To be honest, other than HD201's inferior quality against sr80, music sounded almost same to my ears, but decided to go with b320.

Under EAC:
F10 settings: extraction mode: secure (clicked cached and C2 - after it tested my lite-on dvd writer), selected speed 40x, allow speed reduction during extracting, gap detection B, detection accuracy=accurate.

F11 settings: -b 320 --vbr-new --add-id3v2 --pad-id3v2 --ta "%a" --tt "%t" --tl "%g" --ty "%y" --tn "%n" %s %d, bit rate 320 bits, high quality, uncheck "addID3 tag"

F9 settings: error recovery = high, external compression queue = 1

Problems:
1) Speed never exceeds 4x or 5x, it takes up to 25-30 minutes to rip one cd. (see edit at the below, it is actually 2.5X)

2) If it finds error, speed becomes x0.1 and sometimes it cannot finish ripping. Whereas if I use dbpoweramp for that same track dbpamp converts within minutes.

3) Dbpoweramp is using LAME 3.96r, but EAC is using 3.97 beta. If I use dbpoweramp, it converts a complete cd much less time than EAC. BUT, It doesnt give any tag which is soooo annoying. Any settings that I am missing in dbpwramp's conversion menu? That's why I decided to use mp3tag. Also, am I losing some quality by using dbpoweramp rather than EAC+lame3.97?

4) I am planning to use MP3gain after I convert around 1,500 tracks. Within this time I also read that "mp3gain" software can adjust the peaks more "scientifically" (whatever that means) than creative's normalization function. One question here: is 89 decibel a standard, what do you use for your personal taste?

EDIT: Speed is 2.5X - killing me!!

 

[stewgriff]

You'll have to ask someone else about using dBPowerAmp, etc., but I have experience with MP3Gain. What they mean is that the software analyzes the overall volume of the track, instead of just looking at the volume peak levels, like what Creative, Apple, Sony, and others do (with the latter, a classical piece with a single loud peak will cound much softer than a rock track with little difference in volume). I've used Apple's Sound Check before, and MP3Gain is much more consistent. It took about an hour for me to normalize ~650 tracks, and I set it to 75 dB; I know it's abnormally low, but it ensures no distortion and protects my hearing with sensitive IEMs

. I think most people use 89 dB because it gives a reasonable compromise between lack of distortion and loudness.

As for your encoding problems, I'm sorry I can't help...

 

[vorlon1]

I've been using EAC until recently until I ran across some other free software that seems to convert a ripped wave to an excellent mp3 very rapidly with excellent quality and with this setting a small size. This is the setting
-V0 --vbr-new -b128 --lowpass 21 -q0 You can get the files and information here. http://jthz.com/mp3/ I just tried this command line in EAC and it also works fine and gives a file size almost identical to razorlame (the program I am sending you to.) You can also use dbpower amp to rip to wave and then convert to mp3 using razorlame. Razor lame took about 10 seconds to convert the test song I used for this (it was a 3 minute song) and db poweramp took about 45 seconds to rip it to wave form. Good luck.

 

[Ooztuncer]

Quote:

Originally Posted by vorlon1 I've been using EAC until recently until I ran across some other free software that seems to convert a ripped wave to an excellent mp3 very rapidly with excellent quality and with this setting a small size. This is the setting -V0 --vbr-new -b128 --lowpass 21 -q0 You can get the files and information here. http://jthz.com/mp3/ I just tried this command line in EAC and it also works fine and gives a file size almost identical to razorlame (the program I am sending you to.) You can also use dbpower amp to rip to wave and then convert to mp3 using razorlame. Razor lame took about 10 seconds to convert the test song I used for this (it was a 3 minute song) and db poweramp took about 45 seconds to rip it to wave form. Good luck.

Thank you very much for the replies. I think I found the cause of the problem: liteon dvd writer. It's a really good writer but in EAC's secure mode it maxes at 2.5X. I confirmed these results from some posts under clubcdfreaks forums.

So, what I am doing: using burst mode + test / copy. I am ripping + encoding one cd around 5 to 6 minutes., which is huge improvement after 25 minutes

Another note, razorlame seems interesting and easy to use. Couple of concerns:

1) I want to rip + encode at the same time because of the tagging issues. If I convert them to wav first and use razorlame or foobar or eac to convert wav to mp3, I need to use mp3tag in between these steps.

2) I didnt try razorlame yet however thinking that EAC's "error" alerts are invaluable while using the burst mode. I am assuming that it gives me error or just stops encoding. By this way, I can try slow ripping. Please correct me if this assumption is wrong while using the burst mode.

3) I like the idea of variable rates (in your comment line: -v0 -b128). I think it is using 128 bit variable rates which means in the range of 120 - 190 kbit/s. is this correct? I have 30gb of space and 320 kbits/s constant rate would me give better quality, but I am not sure about the artifacts. As you can see in my first post, I tried 5 different versions and I noticed almost negligible differences. You can argue that then why not small size or variable size rather than 320 constant - I think it's just psychological reasons, lol.

I want to hear your thoughts on the issue.

@stew: For the Mp3gain, did you throw all of your 650 tracks at once? Or did you first analyze them within each folder (after the rip I am creating different folders for each album) and manipulate the dba levels and only after these steps merge all of tracks and analyze/manipulate again? Hope I can make it clear.

Cheers!

EDIT: Actually, I am not sure what -v0 -b128 does? v0 is the best variable rate (very high quality) but what is happening when it merges with 128 kbits/s constant rate?

 

[vorlon1]

I've never used EAC in burst mode only secure, so I am familiar with the 2.5 x speed and the slowness of it. EAC rips to wav first and then runs lame to convert to mp3 so I don't think raxzor lame would be any different at inserting tags, but I am not an expert on that. Most of my mp3's are 320 cbr but when I tried razorlame with that command line I sent you they come out playing around 224 and strangely enough souded better than my 320 cbr when I a-bed the exact same track encoded each way. That 3 minute song came out to be 7.5 mgs with 320 cbr and 5.4 with the variable rate, so, not only did it sound great it saves space. I wish I had discovered that setting before I filled my Ipod with 320 cbr music, but oh well...

 

[Ooztuncer]

Well, best way is listening right. I tried to compare them:

1) (320 cbr) >> burst mode + test/copy >>> 10.11 megabytes
2) (-V0 --vbr-new -b128 --lowpass 21 -q0 caused 236 vbr) >> burst mode + test / copy >>> 7.47 megabytes

Besides, both of them are done under a minute for Buena Vista Social Club's Chan Chan.

Well, either I am tired or biased but 2nd method sounded more sharp, kinda increased in high freqs. Perhaps caused by lowpass21 or q0 comments, do you know what they are doing?

On a positive note i tried: -V0 --vbr-new -b128 --lowpass 21 -q0 caused 236 vbr --add-id3v2 --pad-id3v2 --ta "%a" --tt "%t" --tl "%g" --ty "%y" --tn "%n" %s %d (your recommended line merged with tag info and worked great - creative recognized everything

Tomorrow I will listen them again. Perhaps high bit vbr is the way to go...

By the way, I used zen xtra > pa2v2 > sr80

 

[vorlon1]

From what I understand from reading at that site I sent you that comment creates a filter that filters out any frequencies above the numbered amount. In this case it would be everything above 21,000 hz, which of course is higher than cd's are encoded at, so it is intended to not attenuate any of the high frequencies.

 

[TinkerToyTech]

dbpwramp is my go to cd ripper when it comes to ripping cds.  besides being able to rip in FLAC, I can / could at the same time with my i7 based desktop rip any number of formats, including 128k for the in car ipod where with road noise alone the best you can hope for is FM Stereo quality audio for mp3 cds for the headend I'm running, for my dj mixing I'm ripping at either 320k mp3 or 256k m4a but with a limit of 90 db.  I have found with the 300 million odd cds ripped by this software rarely are the tags wrong.  Oh!  Did I mention it is possible to rip these four aforementioned formats in one pass, i.e. at the same time!  Send each format to it's own path and you are set!
 
Anyone else use dbpwramp for ripping cds?

 

[514077]

I have a problem with DBPowerAmp version 16. It doesn't remember or save any configurations I input. When converting, I try to add DSP effects, or change bit or sampling rate. It doesn't recognize or carry out the changes I make. When I try to select 'secure ripping', it just does a single pass. Version 15 remembered settings change. Why doesn't the new version? I'm using Windows 7. Thanks.