Hi,

Im new here. Ill give you a little background. Im a Dj and Im going to start djing from my computer. I have a nice stereo system, Mackie 1400 for the Evs, and mackie 2600 for 4 pioneer 15s. What Im looking for is clear highs, and nice lows. Mp3s are horrible for this. Ive tried Windows media 8 and that was a little better. I have a nice ripper that supports vqfs and there pretty good but all I read is AAC, and MPC. Theres no way that I can use a losseless format, way to much space. Im looking for the equivilent mp3 bitrate from 100-200. I would also like to know what and where is the software for ripping and encoding. Any help is greatly appreciated. Thanks

MP3 shouldn't be too horrible for that application. Before you encode your WAV files, use SSRC with "--rate 48000 --bits 24 --twopass --dither 0" and then try encoding the output WAV with LAME 3.90a8 using "--dm-preset standard -Z". The bitrate will be around 230kbps most likely. Dibrom plans to release new code and presets that will achieve the same quality (or higher) at a lower bitrate soon.

The other viable alternative might be MPEGplus/MPC. I'd start with -standard which should give bitrates about 180kbps, and you can move up to -xtreme, or -insane, which is normally about 295kbps.

I'd forget using VQF -- it does properly represent the frequency of the original audio, but the 96kbps maximum bitrate is too low and produces too many artifacts -- especially stereo loss -- for DJ use. WMA8 is actually worse at high bitrates than WMA7 is, and honestly, for your application, you probably need bitrates higher than what WMA7 can provide to get the quality your audience needs.

The best option I can see for you would be MPC. You can also look into AAC -- it's main hinderance is legal issues.

The previous reply assumed you were using Windows. If you are not, then encoding MPC is not a viable option. Decoding is fine on all platforms, and it does currently offer the best quality at high bitrates.

http://www.stud.uni-hannover.de/~andbusch/...ocoder_eng.html
(I hope that URL worked: the bulletin board has put it on multiple lines).

If you want to stick with MP3, using LAME with one of the 'tuned' presets like '--r3mix' or '--dm-preset standard' is your best bet. LAME is completely free, and the best MP3 encoder at the bitrates you want. There is a lot of work going on tuning the presets, and you might need to get an 'alpha' of it to get the best results:

http://www.mp3dev.org/mp3/

If you can afford to wait a week, then you might want to try the Vorbis format:

http://www.xiph.org/

The current release, RC2 (RC = release candidate) has all the features, but isn't fully tuned. RC2 has comparable quality to MP3 at medium-high bitrates, and better quality at low bitrates. RC3 will be out soon, and has better quality and smaller compressed file sizes).

--

On your second point, there are many easy to use rippers and encoders for many platforms, and many of them are free. On Windows you might want to check out

Exact Audio Copy (EAC) http://www.exactaudiocopy.de/
and
CDex http://www.cdex.n3.net/

The first is 'postcardware' (just send him a postcard to register), the second is completely free. Both are excellent.

On Unix based machines, there are many more options for front ends, and many more programs that are command line based. Two suggested ones to use are:

CDParanoia http://www.xiph.org/paranoia/index.html
and
Grip http://freshmeat.net/projects/grip/homepage/

Linux distributions frequently have both of these programs included.

I'm afraid I can't tell you about Macintosh software, as I last used a Mac about 5 years ago

Good luck and happy hunting.

Hi TrNSZ,







How do you do this and with what software?

Thanks,
auldyin

http://www.milky.ne.jp/%7Egalaxy/ssrc/

I'd agree with the fact that MPC is most likely the best format for the space if very high quality is a factor also. However, MPC -insane averages more around 220-250kbps.. it rarely goes up to 300kbps except for on difficult samples.



This is incorrect. There is an MPC encoder for linux as well on that page. I have a feeling (no I don't have any specific information, just a guess) that more unix platforms may be supported in the near future as well.

I also agree with the sentiment that Vorbis may be a good choice as well (probably after MPC), and if neither of those would work out for you for some reason, then I would look at LAME as mentioned earlier in this thread. AAC would probably nearly on par with MPC as a choice, but you would need a good encoder for this such as LQT or PsyTEL and neither are free AFAIK.

If you say WMA8 is better than mp3 you probably haven't tried one of the high quality LAME presets.

http://www.cd-rw.org/win32lame/

Download win32LAME v 1.7 here. It's got a reasonably recent version of LAME and an easy user interface. Just give the default settings or Dibrom's standard settings a try, you'll probably be amazed

If not you can always come back for more questions

Sorry, you are right. I should have said 'unless you are using Windows or a recent version of Linux on an i386 PC (and don't mind not having the most recent version of the encoder), then encoding MPC is not a viable option'.

Honestly, I think it is pretty much a given that the chances of someone using or having access to either windows or i386 Linux are pretty high. In addition, in the original post he mentioned Windows Media and as far as I know that doesn't work on other platforms besides Windows (except maybe with Wine or something). The bit about not the most recent encoder is true, but the update for the windows version was a very very small update. It only affects the use of digital silence between artificially created transients. So in any sort of real world situation, this is most likely not going to be an issue.

After my experiences switching to a mostly Linux environment, I hate to see people being sucked into staying with Windows. To take two examples -- Monkeys Audio and RAR. Both 'better' than the opposition, and both Windows only. MPC is in this category as well -- the encoder will only work until the next binary incompatible C library version.

It wouldn't be so bad if someone could look at the spec and write their own version of the encoder... but their is no spec.

This is all irrespective of the quality of the format, which is excellent.

I believe one of the reasons RAR is widely used (especially on Usenet) is because it has different flavors for other platforms, including Linux, DOS, OS/2, and Mac. It's actually inferior in compression to other lossless codecs such as Monkeys, lpac, and flac.

ff123

http://209.235.4.116/rar/rarlnx29.sfx

rar for linux

Windows Media players are available from Micrsoft for Windows and Macintosh, and I've seen unofficial (and unsupported) players for Linux as well. Also, ZIP and RAR are about as multiplatform as you can get. An open-source version of Monkey's Audio is coming very soon too, according to some previous discussions with Matt I remember reading.

Hey,

Thanks for the support. I am using currently An AMD 1.4 with Windows XP. I do plan on getting sucked into windows. I plan on reaching a MCSE in Windows XP. Ill wait a week, but I need something soon. Still havent found a software that will work with my computer to use any format other than mp3, or vqf. Thanks

I have seen this stated several times and as far as I can verify it's simply not true. The windows and Linux versions always give different results when I try them.

--
GCP

This is directly what Buschel (the author himself) said about the newer version. I kind of doubt that he would make a false statement about something like that. The tweak may not produce binary identical results on everything else though, but unless you have actually heard a difference, then I don't think it is an issue.

Hmm, sounds interesting. Have a link? A linux player would mean DRM support is basically worthless.

(Unless the Linux player is actually a shell around Wine, like aviplay is for DivX movies. I hardly call that a 'Linux' player)



Could be, but currently, there is not. 'Soon' can mean so much.

FLAC is multiplatform and has working Winamp and XMMS plugins (and it's even Free Software), right now.

--
GCP

The Linux version seems to produce consistently bigger files than the Windows version. I don't think I need to hear filesizes.

--
GCP

How much bigger? Can you give an example?

Not at my Linux machine right now, so you'll have to wait a few.

I know it happened on 'dogies', because I tried to reproduce ff123's encodes and got a larger bitrate than he reported (5-15kbps, dont remember for sure). The encodes matched up when I used the Windows version. I saw it on other clips too and simply stopped using the Linux version since then.

Edit: Hmm, I don't think 'dogies' contained arificially created transients. So there must be more differences than those you mentioned.

Edit2: ff123 got 143kbps, I got 150kbps

--
GCP

Hrmm.. 7kbps isn't a very large difference. I think what is likely the case is that Buschel just tuned some values which probably affected the encoder behavior in all cases very slightly, but which resulted in big differences in situations with digital silence and harsh transients. Of course I could be wrong, but I don't see why he would list that as the only change made in the updated version if there were other things modified as well.

If I can get a 7kbps size difference (and the 7kbps was just on this clip which is the only one I could check without being at my own machine) _without_ a quality difference I will take it anyday. I am mainly contesting your claim that it doesn't ever make a difference in real life situations. It certainly does.

The net effect, whatever the reason, is that the Linux version is inferior to the Windows version. I reported this problem at least 2-3 months ago. And it's still not changed.

I understand Buschel just moved and has lots of things on his hand, but isn't MPC supposed to consist of multiple people right now? Is there really no-one there that can make an updated Linux version?

If something this simple takes so long, I keep having serious reservations about MPC on Linux.

--
GCP

What I was talking about is that it shouldn't make a difference in quality in a real life situation. This is why I've asked if you have actually heard a difference or not.



Yes Buschel has just gone through a move and taken up a new job, and no the new programmers have not actually started working on all of this stuff yet. The whole thing has been waiting on getting information back from the German patent office which Buschel said in his last post should be happening in a few days now. Apparently he hasn't had ready access to the source code during some of this time period either and as you have seen, the Windows version has not received any updates lately either.

Give it some time. As I'm sure you know, sometimes things can get in the way of releases in all kinds of situations. As I recall, we are all still waiting for Vorbis RC3 are we not?

You are correct -- you can *decode* MPC on just about every platform out there. That is not the issue.

No, RAR is less multiplatform than even MPC, because there is neither a description of the file format, nor an open source decoder. ZIP is very multiplatform, because the file format was reverse engineered, and there are freely available compressors and decompressors. The most multiplatform compression formats are .gz (GNU ZIP) and .bz2 (BZip 2)... but they're not popular on Windows because there is no neat shareware utility that creates them by default.

I look forward to seeing if Monkey's Audio is opened up, and, if it happens, what license he uses (I've seen people promise to open up the source to their products before...). It's an excellent piece of work, but currently useless to me.

It doesn't seem to make a difference in quality. Why should it have to? The issue is that the Linux encoder is not current and that it does make a difference in real world situations, even if not qualitywise.

(and note that I could have used those 7kbps to use a higher nmt or tmn threshold, so it is indirectly also a quality issue)

I'm taking offense to the fact that, because I use Linux, I'm getting a worse performing product than what I would get by using Windows. (and for the sake of my own argument, I'm forgetting that it works under Wine )



At least our Windows and Linux versions suck equally hard

--
GCP

Sorry, but this information is completely wrong!

RAR has been reverse engineered as well (and quite a long time ago), and I have RAR 2 compatible decompression and compression source code that compiles almost anywhere; I've used it on OpenVMS/VAX 7.1 and Amiga quite extensively.

The code isn't easily available because RARsoft doesn't want their intellectual property rights "violated" - but is out there. There has also been an official open-source decoder available for years at ftp://ftp.elf.stuba.sk/pub/pc/pack/unrar250.zip that includes source. This may be old, but the point I'm showing is that it's been available since "ancient times". Look around and your sure to find the compression source as well.

I'm not a Win32-advocate or anything, but that just does happen to be the platform I'm using now. I've been a long-time user of VMS, Ultrix, and Macintosh, and at no time did I find myself "stuck" because of a format issue - I could always hack or code something and the situation remains the same.

No offense to anyone, and this isn't directed at anyone on this board, but in general, I think Linux users just like to find things to whine about rather than actually doing something.

Also: Garf, where is your tuned Vorbis OggDrop and OggEnc?

Thanks for correcting me... although the source code that is available is old and in a terrible state, it is available (at least for decompression, I've not found any for compression). So RAR is now on a par with MPC for cross platform usage. I'm not interesting in hunting around for semi-warez code in order to use RAR properly.

Lucky you

Darn, you're onto us!

Seriously, I just happen to [b]hate binary-only and semi-legal software after having my share of problems with it.

If something is really good, like MPC for example, I'll make exceptions, but I _will_ whine about not having source for them, and binary-only software is always considered second-grade by me.

As a very obvious real-life example, if the MPC encoder would have been Free Software, I would never have felt the need to enter this thread, because I could have recompiled it myself (properly optimizing for my CPU). If Monkey's Audio would have been Free Software I would have been able to play it on my Linux box.



http://sjeng.org/ftp/vorbis

--
GCP

LOL.

I knew I would hit a nerve with that last one, but seriously, it really wasn't directed at anyone in particular.

I used to use Linux, and did some kernel development back in the old days using the SLS and MCC distributions, since this was before Slackware was invented, and Linux was a viable upgrade from my existing system which was using some version of Minix, maybe 1.7.4 (?), with some VM patches, which let me run X11R5.

I only decided to move away from Minix because it didn't support shared libraries, so everything was static-linked with X11's libraries, and every X binary was about 5MB or so, and that consumed space quickly. I kept with Linux up until 1.3.57 or so and then finally dumped it after getting no love for fixing Ethernet drivers for HP EtherTwist and SMC M2-Ultra cards, neither of which ever made it into the kernel I don't believe.

Honestly, give me OpenVMS anyday.