MP3 Ripper

I think using musicmatch breaks ff/rw on the empeg though. :-(

No, using Variable Bit Rate encoding is what breaks FF/RW. It's not Musicmatch's fault, it's the empeg's fault. This bug will be fixed on the empeg with the next software release.

Variable Bit Rate encoding is a good thing, don't let the FF/RW bug scare you away from VBR.

___________
Tony Fabris

You are right, I got things mixed up. Most of my lame encoded vbr files work fine. Heres a quote from the release notes.

"It appears that Audio Catalyst will produce files with incorrect VBR
headers if Real Player 7 is also installed. It is not possible to fast
forward or rewind such files, the time display will not update and the
player will not remember where it was in the track if powered
off. Apparently the latest version of MP3 Tag Studio can repair these files
although we have not tested it."

Sean

Oh, I see. You were referring to a different issue where the files are actually corrupted. Sorry, I forgot about that one.

Even if the files are correctly encoded and fixed, the empeg currently has trouble doing FF/RW in VBR files. This will be taken care of with version 2.0 software.

___________
Tony Fabris

I have to put my vote in for EAC + Lame. I've been using them for about a year now. I _CAN_ hear a difference between Lame and Xing... Or at least I think I can Maybe that's why it's called psycho-acoustics?

-Trevor

-----
Mk 2, Green 12GB 080000349

In reply to:

---

There are a bunch of lame command lines around, so if you need one, let me know. If you decide to use lame, try to use the command line exe rather than the dll, because the dll hasn't changed much since 3.87.

---

I'd be interested in some of those lame command lines. I just switched to using the dll last week instead of the exe for the simple reason that whatever way I had the exe configured, it was taking about 30-35 minutes to encode a ripped .wav file compared with the dll ripping and encoding at between 0.5x - 0.7x (about 5 minutes for a 3.5min track)

(I'm using EAC 0.9 prebeta 11 and lame 3.89 MMX, encoding to 192k VBR)

Geoff
---- -------
Got one of the first Mark 2 empegs...

I'd be interested in some of those lame command lines. I just switched to using the dll last week instead of the exe for the simple reason that whatever way I had the exe configured, it was taking about 30-35 minutes to encode a ripped .wav file compared with the dll ripping and encoding at between 0.5x - 0.7x (about 5 minutes for a 3.5min track)


From abcde (linux using cdparanoia and lame) I use the following command arguments:

-m a -q 2 -h -v -V 2 -b 96 -k --add-id3v2

There is a bit of redundancy in the arguments, but it works very well for me. Basically, I'm setting it to (in order)

Automatically decide stereo format (joint or separate)
quality 2 (1-10 scale, 1 being absolute best)
High quality switch (Vbr automatically turns this on I think)
vbr mode
VBR quality of compression (1-10 scale, 1 being bigger files, 10 being crappy sound)
Minimum bitrate of 96kbs
kill the high-pass filter
force writing id3v2 tags (lame will write v1 tags by default only)

I used to use -mj -h -q2 -V1 -b96 -B320, but have recently switched to --r3mix -b96 --add-id3v2. From www.r3mix.net: "--r3mix -b112" is synonym for "-V1 -mj -h --athtype 3 --lowpass 19.5 -b112"

I changed to -b96, just to stick with my previous minimum bitrate.

I highly recommend this site for a good read on MP3 encoding.

-Trevor

-----
Mk 2, Green 12GB 080000349

Hi.

My lame commandline option is configured as follows within EAC:
-v -h %h-V 0 -q 2%h%l-V 4 -q 5%l -ms -b 128 -B %r %c-p%c %s %d
All that % stuff is for using the switches EAC provides to configure the encoder. With my current settings, the above commandline equals to:
-v -h -V 0 -q 2 -ms -b 128 -B 320

cu,
sven

proud MkII owner (12GB blue/green/smoked, #080000113)

Damn, I guess that rules out my first choice.. ripit.pl :)

12gig red mk2 -- 080000125

In reply to:

---

There is a bit of redundancy in the arguments, but it works very well for me.

---

Just out of interest, how long does lame take to encode files using your settings? (and what spec is the machine?)

I just find it hard to believe that on my 300MHz P2 (with 384MB RAM!) lame is still estimating 40 minutes to encode a 6 minute .wav - admittedly it's running Win98SE, but even that can't be the problem!!

I know about the reported problems with Audiocatalyst, but it ripped and encoded a CD faster than the playing time. EAC & lame (at the minute) is taking about 5 hours to do one disc... it's getting to be an overnight job to rip a couple of CDs!!

I must be doing something very wrong!

Geoff
---- -------
Got one of the first Mark 2 empegs...

I recommend Easy cd-da

I tried audio catalyst but it was hard to get to work because it requires aspi plus easy cd-da gives free updates

32Gig MK2 In 2001 VW Golf TDI

Just out of interest, how long does lame take to encode files using your settings? (and what spec is the machine?)



I'm ripping at approximately 5x realtime (Pioneer DVD drive). Encoding is being done at about 1.75x realtime on my main (at the house) fileserver.

Now, Specs... that machine is a Dual P3 550, .5G ram, with significant drive IO.

I'm ripping with abcde (a better CD encoder), which pulls the tracks as fast as cdparanoia can get them (as perfectly as possible), and then fires off encoders in the background to process the files. So, a full CD takes around 20-25 minutes to encode, thanks to the 2 processors.

I can't see lame taking that long to encode a file on a 300mhz... It's certainly not going to do it realtime at VBR, but I don't get it taking that long... I'll run a test on my backup server and see what it does...

Ok... Did a quick test on my backup server (P2 300, 128M ram)

Ripped an ~5 minute track, fed it to lame with -m a -q 2 -h -v -V 2 -b 96 -k as the options...

Encoding at .6599x realtime... so, about 8 minutes.

This is with the most recent version of lame, on a Debian linux box.

Gotta say something is up with your setup...

i can do direct CD to mp3 rips at 8-9x. am using audiograbber and lame set to 192 High Quality. set on normal quality i get about 13x

Machine is 1.2GHz Athlon, 512mb, Teac 32x SCSI CD-Rom

almost my exact setup, i have a 1.3 with more ram, and i am using audio catalyst as it came packaged. i get 9x-10x usually(cd in good shape)

Ok heres something you could try. go download the binary from the following site:
http://www.animus-facticius.org/lame_dm/lame_dm_preset.zip

The presets are activated by "--dm-preset" followed by, "standard, xtreme, insane, metal"
Here is what the following presets specify:

Standard Preset = --nspsytune -V2 -mj -h --lowpass 19.5 -b112 --nssafejoint --athtype 4
--dm-highfreq 3

Xtreme Preset = --nspsytune -V2 -mj -h --lowpass 19.5 -b112 --nssafejoint --athtype 2
--dm-highfreq 3 -Y

Insane Preset = --nspsytune -V1 -mj -h --lowpass 19.5 -b112 --nssafejoint --athtype 2

I have been using the extreme preset and haven't had any problems so far. I have an amd duron running at 800mhz and i usually get 2x encoding speed. The nssafejoint switch is for a new more accurate joint stereo.

Sean

Hi.

Regarding encoding times of lame:
Using my settings posted earlier in this thread, lame encodes at about 80% of playing speed (or needs 72minutes to encode a 60 minute CD). If I use CBR encoding instead of VBR, it is much faster than this (but I can't remember exactly how fast, I think about 2x playing speed). That is on a PIII with 500MHz.
If you use EAC to rip, try using its backgroud encoding option (EAC -> EAC Options -> Tools -> On extraction, start external compressors queued in background). It really helps alot, especially when batch-encoding a whole bunch of CDs.

cu,
sven


proud MkII owner (12GB blue/green/smoked, #080000113)

It does seem that MusicMatch generates bad headers though. I have to use Mp3-Tag Studio to fix my VBR headers before Winamp will work correctly with them (displaying the correct duration, accurate seeking, etc.)

I haven't ripped any music for a while (getting my new computer set up). On the old machine, running Linux, I ran grip + lame and everything worked quite well.

On the new machine, I'm running Win2000 and I just spent the $20 on Audiograbber. I also have a snazzy Plextor 16/10/40A IDE CD-RW drive that allegedly does audo extraction at the full 40X. In practice, I'll get audio extraction at somewhere between 20 and 30x, and then lame (with the r3mix flags) runs at between 8x and 10x speed (on my 1GHz Athlon).

However...

My machine occasionally decides to just hang. Freeze. Hit the reset button. The problem seems related to the Plextor drive because when I switch to using my Toshiba DVD-ROM drive, everything works flawlessly (albeit much slower than the Plextor). I even upgraded my Plextor firmware to the latest release. No effect.

Anybody else had problems like this? Could my Plextor drive be flakey?

Also, of all the Windows extraction tools I've tried, they all pale in comparison to grip on Linux. grip is smart enough to run lame on the previous track while extracting the current one, and it has a GUI that doesn't drive you completely batty (unlike AudioGrabber which just gets all confused when CDDB hasn't heard of your CD). It's almost enough to make me want to set the machine up to dual boot -- purely to run grip. I actually tried running grip on Linux through VMware, but it turned out that audio extraction just doesn't seem to work that way at all. *sigh*

Try installing an asapi driver. You can find a good free one here: http://www.vob.de/us/Downloads.htm

Sean

Is the drive spinning up each time?

I used to have that problem when ripping, but Audiograbber has a feature that's called spin up disc before reading. It's under the rip offset tab under settings.

Don't know if that's your problem or not but maybe

Amazing! This turned out to be the answer. I installed the ASAPI driver from the URL you listed, switched AudioGrabber to use the "ASPI Manager" instead of "Windows calls", selected "buffered burst copy", and everything now works great. (I already had "spin up disc" turned on.)

It even successfully ripped a used CD I have that my old computer couldn't even begin to rip properly.

Paging Mr. Fabris: please put this in the FAQ somewhere!

I'm glad it worked, for some reason the versions of asapi that ship with windows and nt calls give plextor drives fits.

Sean

Done. Added it to the "Best way to rip" FAQ. Thanks.

___________
Tony Fabris

I bet that's why my sony 12x speed I use at work is choking too under W2k. Hmm, must try this.

Cheers,
Hans

PS. My plextor 16/10/40 at home burns cd's at 16 speed even while my DLT Tape drive is backing up and restoring stuff (at about 80 megs/minute)... Tried it today, never even got below 75 percent buffer. Gotta love Linux.



Mk2 - Blue & Red - 080000431

I've had good luck with "Media Box" + LAME.

Media box (www.e-soft.co.uk I believe) is a cool little ripper, but it also has fun playback features lilke auto-dj mixing (cross fade, etc)

The only two downers are:
1> It uses CDDB only, not any of the free alternatives (free as in speech that is)
2> it's $25 to register..

Otherwise I've been having great luck with it + 192k abr lame encoding.