MP3 Quality? 192 or 320?

My personal choice is to only use original wav's. No compression, lossless or otherwise.
The price of storage media today is so cheap so I don't see any reason to sacrifice sound quality just to save a few Gb here and there.
Plus when people are *****ing about mp3 dj's I just smile and reply - Mp3's? I don't play any stinkin' mp3's! hehe ;-P

(Something that would be interesting to see is how the various bitrates/file formats effect the processor load of Traktor. Hint, hint Phil... ;-)

 

I use MP3s at 224 kps (converted from my CDs) and to my view they sound good on big PAs. Anyway, nobody did manage yet telling me which songs I played with Traktor or from CD. *smile*

Regarding storage space I agree that prices are decreasing, but HDs are still not for free. I use to have some 40K MP3s resulting in some 250 GB. As waves I would easily need more than a TerraByte...

 

Anyway it looks like an organizational matter, since i bet you will never use this 40.000 songs in just one night, neither in the same month. In my opinion, this becomes an awesome workload to dig into such big amount of music, and to keep it organized must also be an awesome workload. Although i imagine your intention is to cover any need anytime. It is painful, even when talking about 6000 tracks (my current collection for djing) but i assure you that I cover almost any need with this amount (from ambient-"chill-out" to techno, passing thru funk, hiphop, tech-house, minimal, electro, disco, italodisco, ebm,... ) any kind of set for any kind of people can be done with this (and i bet less would be possible also)

In the past, remember we went to perform in a session with an average of 100 vinyls, that's averaging four tracks per vinyl, 400 songs. And we never used them all actually...

But it's the choice of everyone... what to deal with in a session.

I've arrived there after testing also carrying such amount of songs. It became to me uneffective.

Currently i carry just a bag, with the computer, the fs, their corresponding ac adapters and a bunch of cables to connect it anywhere. Ocasionally i also take out the keyboard, but this represents another extra bag to carry. (Note that i don't own a car, currently, so i move by feet anywhere )

Also, i think that wavs do not represent enough sound quality improvement over a well encoded 320kbps mp3, to think about carrying an extra hard drive to contain them.

just my '320 cents'

 

Of course I will never play all songs in a night! According to the new traktor feature my collection contains almost 3 months of music.

I guess that you always gig on parties/in clubs. I usually do corporate events and private parties. There, you never know what you're gonna need to play in advance. That's the thrilling part of it and the main reason why I moved over from Club-D.J.ing quite a while ago!

 

Moore's Law (held true for over 30 years). In 2020 you will have a hard disk for USD$20 capable of storing every top 100 house/dance/ambient/techno/trance track ever released.

Leaving aside the pounds of flesh that the greed driven facist record companies will want from you . . .

I love hearing old DJ stories (he he . . . old DJ . . .) Get with the program people!!!!!!!!

Anyway, 192-224 K MP3 (VBR preferred) and 128-Kbps M4A (aacPlus, not some other inferior codec) satisfy my personal listening test for DJ'ing. Assuming CD Audio source. my 2cents on quality.

 

Its interesting to see the shift over the years since Traktor 2.0 in peoples useage of tracks, especially my own. Once upon a time I used to argue that 128K was fine as the bare minimum bit rate. I did and still do make the case on well encoded tracks the punters cannot really tell the difference on a big sound system. But these days I can hear it in the fullness and body of the music 320K and Wav just sounds fuller. I tested and proved my case on 128K a few years back and pissed some vinyl DJ's off too. Today I play out with 320K, 256K Minimum. My archive files are in a Mix of WAV FLAC and AAC Lossless. I'm converting all my vinyl RIPS to FLAC to save space but even those have WAV backups.
What it seems to be coming down to now is much higher Bit rates to play out and Uncompressed or lossless exclusiviely for the original archive libraries.

What many people seem to lack is a consistent process for 2nd and third tier backups
i.e. Backup hard Drive then Tape or CD/DVD. I know I don't but I should.


Phil

 

I think 192 is the minimum that is acceptable for playing out on a big system. I'm trying to buy / rip all new stuff at 320. You can really feel the lack of bottom end on 128 on a big system.

I can't hear the difference between 320 and CD on a club system.

 

Although--I like to think of sound quality & fidelity as somewhat like editing in a movie--if you're watching a gripping chase scene, say, your mind is not thinking about the cut from this camera angle to that one, or the length of time that shot X lasts for (if the editor is doing their job right); the only thing you're thinking about is the chase depicted. Higher sound quality may not make a conscious impact, but it's a fundamental resource to tap in getting music across to an audience.

And audiences may figure it out. Once upon a time, people thought 78rpms of Caruso singing were remarkably accurate. To some degree, the fact that audiences don't seem to notice (consciously at least) is, I think, a matter of not knowing what to listen for and not having the vocabulary to describe it. The more burn-in time we have listening to these various resolutions, the more attuned we all may become.

Personally, the more I listen, the more readily I hear the differences--especially now that with more laptop DJs, there's more variety to compare. It's only recently that I've had real opportunities to hear a laptop DJ play alongside a vinyl or CD DJ. Sometimes the change is stunning.

Peece,
T. Tauri

 

IMHO I go with 192 Kbps Constant Bit Rate (CBR).

I've worked that way for several years and in the old days I used 128 Kbps... and back then comparing 128 Kbps vs 192 Kbps really makes me fill like an Idiot because the audio quality was abysmal !

Then I heard about using 320 instead of 192, I made some tests and I couldn't find any significant difference, the quality for me sounds the same and I'm aware that other ones will argue that indeed there is a sound quality boost. I've being DJing and recording DJ sets for years at 192 Kbps. I got my files from CDs and recently from digital stores like beatport (beatport uses 320 kbps but I transcode them to 192 kbps). And doing that way my friends and all who hear my sets can tell it's crispy clear CD quality.

Now I'm considering going the m4a way with 128 Kbps. and now even more that Traktor 3 supports metadata tag writing for m4a files.

Maybe it's way off topic but I must comment that I've made some test by ripping my CDs to m4a using itunes (I only use itunes for ripping) and also to transcode some mp3 192 to m4a. I found that the quality is the same and my conclusions are that m4a 128 kbps are CD Quality. Again this is MHO.

 

Most people that defend wavs or at least high bitrates use the argument that the lossy format means sounds that you would feel and not hear are removed (<30Hz) so on a big bassy system you would notice it. Would you agree?

I'm not so sure how this works though. In other words how low a frequency is the cutoff when encoding to mp3 and does it matter when it comes to feeling the bass? :S

 

I would say it matters a lot. Also, you're losing treble in the audible range too (anywhere from 14khz to 20khz depending on bitrate).

Depending on the system design and power, and ambient noise level these things can be masked better or worse. On a really nice system, I hear the treble loss too.

Beyond fidelity of playback, there's also the matter of how well the sound responds to EQing (what kind of precision or change you get when you boost in the lossy zones). (I'd propose that as we get into the age where digital software EQing becomes more ubiquitous, a 96khz file will have a distinct advantage in this area)

Here's something interesting: not all 320kbps mp3s are created equal either. Traxsource's, for instance, go all the way up to around 20khz, while beatport's and dancetracksdigital's cut off around 17-18khz. I reckon encoding speed is a factor.

Peece,
T. Tauri

 

Completely agree which is why I'm going for lossless now for a while to try it out. This way I can encode them as I see fit. On the sample rate issue I agree also. It would be good if they could offer above 44Khz lossless too. My x-fi can output at 192Khz and doing so even makes the latency go down!

You mentioned the high frequencies and I understand what you mean - is the same true for lower frequencies in your experience too? I am dubious about it having any significant effect but haven't done any tests myself.

 

hmmm. huge discussion. if you're worried about space, but want to get decent quality, why not use wma then? 128 wma is equal to 256 mp3 or higher, and minimal file size. that's what i use... dont know what traktor/computer you're using though.

 

For real. A format that can scale up to 88 or 96khz AND hold all the metadata like artist, song, Traktor stripe is what I dream of for Traktor (I'm really captive to 320kbps 'cause of the metadata & stripe-writing).

At 320, I'm pretty happy. Especially with my own rips (using the slowest LAME conversion where I get frequency extension up to 20khz--threshold of audibility), I feel I can hold my own next to a DJ rocking vinyl, or even exceed them (since I digitize through a better cartridge and phono pre at home). Maybe losing a little precision if I bang the bass on the isolator.

But I hear the low end falling away as the bitrate drops to a decidedly noticeable level by 192kbps. To the point where a piece of vinyl played next would have an immediate wow factor by virtue of the frequency extension alone.

Another aspect that makes for some trickyness with mp3s worth mentioning is that they will produce more clipping (*especially* with much modern CD audio that really hugs the rails for maximum loudness and probably already clips to some degree), which can degrade fidelity in other ways (i.e. distortion).

Peece,
T. Tauri

 

aacPlus v2, even at 128-k rocks my boat and sounds closer to the Technics 1210 I have plugged into the same amp than any MP3's I have. If you haven't tried aacPlus v2, give it a go - I was amazed at the frequency range difference compared to MP3. Even if it's doing estimation, it does a damn good job.

Phil, I share your back-up concerns! Even dual layer BD-ROM won't meet my demands. A back-up disc in my second machine is what I use. But keeping the back-up in sync is always an issue (we all know how useless briefcase and offline folders are), I use the little app CopyTo (shareware).

 

I definitely find that there's something missing on mp3's below 192. If you're working on a big sound system & mix from a 256 mp3 to a 160 mp3, it's almost as if somebody's switched the subs off!! ....the bass just disappears.

I know of a few mp3 using DJ's who swear by units like the Peavey Cosmos. Basically it's a sub harmonic synthesiser that ad's the bass (& other harmonics) back in to the signal passing through it. I think it basically tracks the bass line, then generates extra bass info an octave lower.

I now use apple lossless or Wav for stuff I play frequently, & aac's encoded at 160 for less often played stuff.

 

i'am agree with u jesse....

for MP3 compression, CBR320 and VBR256+ have same quality, but the size are less than 320 (about 3MB less)
i have tested it in Studio Monitor....
and The difference is not to much.... it heared Same.....!!!!

it because VBR strategy is use higher bitrate just for the part of the song that have much treeble (Hi Freq), but whe it just Drum/bass it use less bitrate (intro n Outro of Dance Music)
So it can Save disc space usage, with same quality....


the quality is depend on the Codec u use to encode the Song, and the source of the song.

if u encode from ORIGINAL CD you will got High Quality with 320 and 256+.

But If you download the MP3 from internet, don't wase your HD by converting it to Higher Bitrate.....

It USELESS!!!!!!!!!!!

Remember!!!!
"GARBAGE IN = GARBAGE OUT"

 

Don't forget the encoder settings as has been mentioned. For example, LAME has various presets that trade-off enconding time vs 'quality'.

Different online vendors may use different encoder settings than you would like so much safer to get the original uncompressed master as FLAC/WAV etc if you can - even if you end up having to delete it after encoding according to your presets to save disk space.

 

one thing that i do now is back up my music in 4 different drives. i have spare hard drives so why not make use of them. if one goes bad i still have my tunes. i even have my main tunes on dvd as data.

 

Regarding the comments that lower MP3 bitrates cause loss of lower bass, that is not supported by this graph. It shows that decreasing the bitrate does not cause bass loss, it causes loss of high frequencies. Maybe the MP3s that showed the insufficient bass were encoded by an older or low-quality encoder.

http://tinypic.com/hukgmp.jpg

The graph however is a bit ambiguous because it isn't clear just what type of scaling has been done on the frequency axis at the very left (bass) end, it would not give much detail if it is linear from the left end to the first mark at 2 KHz, but possibly the author has used a logarithmic or other scale for that interval to show the bass details. The non-linear scale on the left is is implied by the 1 Hz axis point. I haven't been able to find the original research document for more details, it might be online somewhere.

I've seen the comment about bass loss in other forums too, haven't heard it myself.

Incidentally it is common practice in audio installations to use a low-cut filter to eliminate fequencies below about 30 to 40 Hz, since these may contain non-musical nuisance sounds such as rumble that just heats up the equipment. Sub-bass frequencies aren't heard very well although they can be felt, and they require a lot of amplifier power and speaker movement that is better used to reproduce the sound above 30 Hz.