Are there any software programs that will improve the listening experience of an mp3, perhaps using some "intelligent" technology to fill-in the compressed information in an mp3?
Futher, what program do you recommend for re-mastering mp3s (volume, EQ, etc.)?
1st search result: http://www.hydrogenaudio.org/forums/index.php?showtopic=6729 (http://www.hydrogenaudio.org/forums/index.php?showtopic=6729)
Sorry for not being more helpful.
If you've read xnor's link, then you realize there is very little you can do, especially regarding artifacts. The lost data is truly lost, and it's not coming back!
Futher, what program do you recommend for re-mastering mp3s (volume, EQ, etc.)?
You can adjust volume and do some simple editing with special-purpose MP3 editing programs such as
mp3DirectCut (http://mpesch3.de1.cc/mp3dc.html) (FREE!!!).
For things like EQ or dynamic compression/expansion*, any audio editor has these effects including
Audacity (http://audacity.sourceforge.net/?lang=en) (FREE!!!). I'm a long-time
GoldWave (http://www.goldwave.com/) user ($50 USD). But, with all "regular" audio editors, the MP3 has to be decoded (to PCM) first. Then, if you re-save in MP3 (or other lossy format) you are going through an additional lossy compression step. Of course, therte is no
additional loss re-save in a non-lossy format.
There is an effect called an
Exciter or
Harmonic Exciter that can add "missing" high-frequency harmonics. This is probably as close as you can get to "filling-in the missing data". You'll have to search for this effect. There is an optional plug-in for Audacity and it's called a
vocal Exciter But, loss of high frequencies isn't ussually the first thing you notice about a poor quality or low-bitrate MP3. If you are noticing high-frequency loss with MP3, you've probably got some severe artifacts too, and these artifacts might get worse with harmonics added.
* Dynamic compression/expansion relates to loudness & quietness and is in no way related to MP3
file compression.
Thanks y'all. It's frustrating as a DJ to find lots of good tunes, but not get them in decent quality... playing them over a big system really highlights the lack of quality.
I think I'll have to get to know Audacity or a similar program so that I can do some fine-tuning before playing ugly mp3s at gigs. With Audacity, can I play with EQ much?
DVDdoug makes a good point. If you have the space then you should encode your edited files with a lossless codec such as flac. Otherwise you effectively transcode from lossy to lossy which could render your editing efforts void due to the newly introduces artifacts.
frustrating as a DJ to find lots of good tunes, but not get them in decent quality...
Really? Are you downloading pirated music? (If you are don't answer that... ) You can always buy CDs. Downloads form Amazon, iTunes, or any legitimate source are usually quite good.
...playing them over a big system really highlights the lack of quality.
As far as general "quality", maybe yes. You might notice the lack of bass, or the lack or highs, or noise or the scratches on a vinyl record might be more annoying... But, it does NOT take a killer system to hear MP3 compression artifacts.
IF you can hear the difference between the uncompressed original and the MP3 in a blind
ABX test (http://www.hydrogenaudio.org/forums/index.php?showtopic=16295) on a big system, you can hear the difference on average headphones. In fact, the noise and acoustics in a club/dance environment often mask MP3 artifacts. If the MP3 sounds exctly like the original on your headphones, it's going to sound identical to the original on a big system.
For volume, there is ReplayGain, which only tells the player to adjust volume, without reencoding. Must be supported by the player though, and I don't know if any DJware does.
The 'lost data is lost' argument is really invalid, if the encoding used an inferior psychoacoustic model (for example, if used one quick and dirty mode, or a fifteen year old encoder) -- then in principle, it might be possible for a better model to detect more-probably-than-not-an-artifact parts of the signal. For example, a pop/crackle remover for digitized vinyl tracks does that: it guesses on your behalf.
Anyway, I asked the same question (http://www.hydrogenaudio.org/forums/index.php?showtopic=92704) a few weeks ago. No luck.
For volume, there is ReplayGain, which only tells the player to adjust volume, without reencoding. Must be supported by the player though, and I don't know if any DJware does.
Then use Mp3gain. Any player will support that.
Ah. Forgot about that.
To original poster: MP3gain will be able to shift volume of an mp3 file in 1.5 dB steps without re-applying the 'lossiness' part of the compression. (Nitpickers have long pointed out that you might lose parts of the signal at -200 decibels or so ... completely inaudible, just don't be surprised if a diffing tool reports 'files differ'.)
frustrating as a DJ to find lots of good tunes, but not get them in decent quality...
Really? Are you downloading pirated music? (If you are don't answer that... ) You can always buy CDs. Downloads form Amazon, iTunes, or any legitimate source are usually quite good.
There's lots of stuff nowadays that doesn't even make it to CD or vinyl, sometimes not even reliable mp3s, but still has the sound that moves the people.
Yeah, these days there's quite a bit of worthwhile self-released music that never goes through "legitimate channels", that is - it's never seen a record label, been worked on by a mastering engineer, or anyone else who is knowledgeable about digital audio tech. People just make/record the music at home with the ubiquitous computer, and then release it on their own site as 128 kb/s MP3 or whatever happened to be the default option in whatever software they're using. Or worse yet, they take these MP3 mixdowns and upload them to YouTube to distribute their music... I've come across quite a bit of music where the only place the musician made it available was on their YouTube account, which means their tunes are at streaming bitrates and have likely gone through lossy-to-lossy transcoding.