Hi there guys. Apologies immediately for naive ripping mistakes I have mentioned below. I have found you now and will correct my ways.

I am a DJ and play off a clever little bit of software called Serato (by Rane), I'm not sure if many here will have heard of it. Rather than bore you with how it works, I'll just mention that quality is my highest (but not only) priority. Laptop hard drives are small so I have to take into consideration file space to a point too. But these mp3s are played LOUD in clubs and bars.

I've done a lot of reading on this forum and other sites about how universally satisfied the ripping world is with LAME. Not the best news I've ever had, as I've probably ripped with 6 or 7 different tools (including the much hated Real Player, Winamp, CDex, Quintessential, WMP) but never, thankfully itunes. I usually always went with the best tagger and filenamer to speed up my process (and I found Real by far the fastest ripper)

I've decided to do a MASSIVE Re-Rip of everthing I own on digital format. SO I want to get my settings absolutely right. A Few questions.


I need three main RIP settings.
1. Size no object - My absolute best stuff. (I'm assuming CBR 320?)
2. The highest quality possible with file space in mind (hopefully inaudible differences to above)
3. High Quality but geared towards saving space.

Should I continue with Winamp and install the latest LAME?

I have done some audio tests, I am nowhere really as I have the problem of going to a different club set-up every night. different amps, speakers, layout, cables, mixers etc etc. So it really is a mine field. The only conclusive way with headphones, but I'm stuggling with the sheer amount of rippers I've used.

I am assuming that I am most concerned with the sollowing areas - -V0 (~230 kbps), -V1 (~210 kbps), -V2 (~190 kbps) or -V3 (~175 kbps) but...

I am a little confused with the "Quality" and "VBR Q" Settings, how does v-0 change the file size and quality from v-1 (even when they use the same bitrate scale?) Also when I change the alt-presets in CDex, none of the settings really change (its always set to VBR Q 2, 128 - 320). Surely if these are presets, the setting should change.....

I've installed EAC, and LameXP, but can't seem to set them up with any confidence (not even sure I'm using the Lame encoder)

Can anyone offer me a bit of advice on my best angle on the above points.

1. If size is of no matter, then go lossless. No artifacts, and the exact same quality as found on the cd. Check and see if your program supports lossless files (any type, its lossless).

2. -V0 mp3 should be enough in this case. The latest encoders are transparent at 130kbps in tests, and I can 99.9% guarantee they'll be transparent at noisy clubs as well. -V0 as you've read gives the bitrate about 240kbps.

3. -V1 or -V2, still high quality but will definitely save space over multiple albums.



If your cds are in impeccable condition and you have many albums to rip. EAC while the best tool for accurate rips, is also the slowest...much slower than other tools like dbpoweramp or even foobar's cd ripper.



The vbr settings tell LAME how to encode the file (namely how many 224, 256, 320kbps frames to use which increase filesize). If you encode the same file at -V0, -V1, and -V2 you should see a difference in size. The -q tag is redundant with newer releases as I think its set to default now. Correct me if I'm wrong anyone.

That's about all I can offer right now...got to get to work. Good luck.

My advice, for what it's worth.

1. Rip to lossless format. They'll be bigger than 320kbps MP3 but once you've done this, you'll never need to rip again.
2. I would use the default LAME setting of -V2.
3. Use a higher number -Vx setting that's suits your expectations.

See these guides for the appropriate settings for setting EAC up for lossless extraction:
http://wiki.hydrogenaudio.org/index.php?ti...gory:EAC_Guides

Use something like Foobar2000 to convert lossless files to MP3.

For your needs (noisy environment, large and varrying audience, stability, compatibility with various software) i'd propose:

1. CBR 320kbit (because of compatibility with almost any software out there).
2. V2
3. V3 or V4

In a noisy club, V5 should usually be unnoticable to everyone... so all of the above do already have a safety margin.

Well, lossless might be good but have you ever tried replaygain-scanning an MP3 and a FLAC? FLAC is like four times slower to read and decode than a 256 kbit MP3 on my computer. I don't know much about this DJ software but I guess it needs to read and decode files into RAM so MP3 might actually give him faster response (perhaps even more with a slow notebook drive).

If your decision is lossy, I'd use latest OGG Vorbis lancer build at around Q5-Q6. You get lightning fast encoding speed additionally.
BTW, as the ripping part is the most time consuming, I'd also advise you to rip to lossless first and transcode the files to a lossy format then.

From what I can see on Serato's website, they support WAV, AIFF, MP3, AAC, and Ogg Vorbis.

Since WAV is out of the question due to space constraints, I'd definitely go with MP3 over the others. It gives you almost universal hardware and software support. CBR 320 is always a dumb idea. If I was in your shoes, I'd probably go with LAME -V2 or -V3.

thank you all for your replies. some priceless stuff there.

Kitsuned mentions EAC being the most accurate but slowest ripper.

Is there another package I can use that will speed it up a touch without losing a whole load of quality. I'm fairly happy with Winamps interface but not sure if im getting a good rip. The encoder is simply called MP3 encoder v1.35a. All the settings look like LAME so I am assuming its a lame ripper.

V

Being a club, as long as it's a decent bitrate it isn't really going to matter. In clubs, obviously the sound is EQ'd anyway (and mainly the bass) meaning that I doubt you could hear the difference between a Lame -V 5 encode and a -V 0 encode.

I've got mates who DJ and they use LAME -V 3, which is probably a good bitrate to use.

I disagree. I always use T/C with EAC in burst mode. Most of the CD's I'm ripping has AccurateRip verification, and with this setup it's just as fast as it will be in foobar2000 and dbpa.

The "slowness" of EAC is if you configure it to use Secure Ripping, but this will not benefit, if CD's are mint and accurately verified. I only use this mode when AccurateRip results fail.

No, at least theoretically, the opposite tendency is the case.

Personally, I think we are way too generous. For VBR -V6 is plenty for those enviroments. Its usually non offensive and the lowpass is close to 16khz which is plenty for music. For overkill V5 thats it for me. For predictable file size and decent quality --abr 144 is a good replacement for 128k cbr.

But lossy encoding is not recommended when post-processing is expected. DJing would involve massive post-processing.

Sorry, POST PROCESSING???? Please elaborate. It is impossible to have ALL My music lossless. the file sizes rule it out

He's referring to tempo adjustment mainly. When time stretching lossy encoding, some artefacts might become noticable.

that's a good shout actually. especially when using master tempo (which shifts the tempo / speed but keeps the song in the same key)

can someone offer any advice on my text below, or is it off topic?



to add to that. Tags and customisable filenames / dir structures are very important for me too

I thought you said that file size was not important. It helps to spell things out in the very beginning.

Just to give you an idea, I often "DJ" at my hometown's country club once every couple of weeks. They have a young person's night (I forget what it is called but it is for the teen members) every other Friday and I play music for them. I use their in-house system and simply bring my notebook which is hooked up via a digital coaxial connection. So right off the bat, you should see that their system isn't anything serious and I think that my Klipsch home theater system with Pioneer receiver produces better quality. Anyway, I use my lossy music when doing this which is encoded at around 160kbps (-V 4). I recently switched to -V 4 as my blind ABX tests showed that I didn't need anything higher unless I was hunting for artifacts (ie trying really hard to hear them). For casual listening, listening at home, and portable listening, I find that -V 4 is perfectly fine.

I have yet to receive a single audio quality complaint from anyone regarding my music. I don't know all of the specs but the room is about 30'X40' and they have four monitors with each one pumping out about 1500 watts (for a total of 6000 watts, I don't know what the peak power is). I am not sure what amps they are using but I normally don't make any EQ adjustments unless a Rap song is being played, that is when the kids want to feel the bass. Otherwise I stick to my normal flat/EQ-less settings in order to hear the music as it is intended.

Often times people will bring their iPods and I simply hook them up to play the songs. That is the only time where I can definitely hear artifacts (I borrowed some tracks to ABX against my own). They mostly use the 128kbps AAC audio format, that is fine. However, many kids simply transcode their already downloaded lossy files to this format and that is when the artifacts are audible. I borrowed one track from someone who insisted on using the 192kbps AAC setting in iTunes. I ran their song through a analyzer and found that it had a 16KHz lowpass. A 192kbps iTunes AAC song should have about a 18KHz cutoff.

Either way, you won't need 320kbps as people get nosier as the night progresses. Not only that but room acoustics often limit perceived audio quality. It also depends on what type of DJing you play on doing. If you are just going there to play songs and take requests then you shouldn't need anything over -V 3 with Lame. If you are going to alter the music and perform live remixes (as I like to call them) then lossless is the way to go.

Since the audience is not likely to be anywhere near the audiophile category (likely teenagers used to listening MP3 on mobile phones/portables?), I doubt anyone at the disco would criticize the DJ for some minor artifacts. If you speed up music, you already distort it and it sounds unnatural so noticing further artifacts should be harder.
The best would be to go to a club when it's closed and try to mix various versions of the same files (encoded using various encoders/bitrates) and compare with uncompressed version.
I guess that anything over 256 kbits is not likely to piss off even the most demanding listener.

In my first post i say "I'll just mention that quality is my highest (but not only) priority."

I asked for advice on 3 settings as I can have a certain amount of "important" stuff lossless, but my hard drive would be pretty full pretty quickly if I had my whole collection ripped this way.

One thing I did not mention tho (APOLOGIES), is that Serato DJ Software only supports MP3, AIFF, WAV.

Three settings for what? MP3 encoding?

You only need to encode stuff to MP3 the once if you do it properly in the first place. The general consensus lately seems to be that VBR at -V3 (~175Kbps) will be transparent to all but a tiny minority of people almost all of the time. Have you tried it yet?

Encoding it again in CBR at 320Kbps would almost treble the amount of disc space you were using to store them and I very much doubt if you could honestly hear a difference in a blind test. There's only one way for you to find that out. For yourself.

Cheers, Slipstreem.

The slowest:
It really depends on your drive your discs and your technique. To just make a blanket statement like this is misleading.

The best:
This is really up for debate. Let's not have it here, ok?

I'm 100% sure that in a club/bar is where people pay the least attention to the sound quality.

Besides, the loudness, the people screaming, talking, etc, would make sure the quality of the music is not that important.

I would go with -V4 LAME 3.98.

If you are DJ'ing with your laptop, I would place all of your potential "DJ'ing" music into one format, using LAME 3.98 at -V4.5 or -V5, which will likely be quite transparant (indistinguishable from a CD) to anyone in a crowded club or bar. Use EAC and configure it so that fast ripping is your priority - you'll see this when you set it up. I think, for your needs, you'll get a much better trade-off in terms of space usage versus sound quality at -V4.5 or -V5 than at a higher setting like -V3 or -V2.

Everything else that you want to save for importance, I would save onto CD-R's in a lossless format. That way, if you need to play them on your laptop on a special occasion, you can, but you don't use up the disc space carrying them around.

Why don't you like Apple - just because your DJ software doesn't support it?

Jesus, people, show some respect. Nobody here has any evidence that his PA system is any inferior of a listening environment for detecting artifacts.

I'm not using Serato but VirtualDJ and only 320K CBR mp3 and FLAC with it.

I avdise you rip your collection in a lossless format, most likely FLAC and since Serato does not support it, transcode your collection to 320K CBR MP3 but keep your lossless files.
If/When Serato support FLAC, you'll then be able to use the lossless files.

You can use EAC for ripping, and foobar2000 for everything else (transcoding, renaming from tags, tagging, ...). Use the discogs plugin for foobar2000 for tagging, it'll help you immensely getting your collection tidy. I know that because I wrote it
Don't be afraid to spend some time learning foobar2000 initially to find a workflow you're confortable with, as you'll gain considerable time later.

I think it may be safe to make certain basic assumptions such as...

1/ The loudspeakers will not be positioned optimally for everybody.

2/ The room is hardly likely to be an anechoic chamber. It WILL colour the sound.

3/ Unless you're holding a Marcel Marceau tribute evening, most people are likely to be making noise of some description. This will affect the way that a person perceives what they hear.

4/ They're all going to be drunk anyway.

5/ LAME MP3 in VBR at -V3 (or thereabouts) is transparent to most people under ideal listening conditions. This has been proven in listening tests to be, on average, as much as you need to convince almost everybody almost always.

So why is it wrong to assume that dropping to -V4 or -V5 is likely to be unnoticeable in less than ideal listening conditions? It does seem logical.

Cheers, Slipstreem.

[rant on]
Indeed, some of the responses here are sort of humorous.
Isn't HA about keeping the bar high, and not low, about quality? "it's a club, it's inherently low quality". It really is?
With music playing in some places at 100dBu? (erm... sorry if the unit is wrong :S) an artifact at -40dB would still sound at 60dBu. (just a bit above the crowd talking loud)

A club is really not the same than a bar. Here where i live are used to discos of 1000 and 2000 people (and a bit further, one of nearly 5000 people). And i am talking about closed spaces, buildings. Towers and towers, big subwoofers, several amp stages, and obviosly, EQ's and filters to separete the lows to feed some amps, and the highs to feed others.

This is definitely not a setup that a psychoacoustic algorithm would expect, and definitely a way for artifacts to show up where they would otherwise be masked.

I, personally, when i go to such places i like to listen to the music, i like to dance to the rythm, i look more to the Dj than to the girls (guess this is why i'm still single :S ), and laugh when a mix doesn't go right.
Nowadays, in most places are using compressed music, that's for sure, but there's still a difference between playing content at the maximum quality possible, than getting the default setting.

In other words, as a listener, I expect a Dj to care about the quality of the music he plays. If he doesn't care about this, what can I expect of him? (and yes, there are some places that i simply preffer not to be).

[/rant off]

About the OP's question: nowadays, LAME is giving such a good quality that is even useful at -V 5 ( ~135kbps), but being a dj you definitely need some margin of safety. -V 3 is mostly used nowadays as the transparent level for those that want to be a bit more sure, so your margin should be above this setting.
So depending on the genre of music that you are going to encode, i would suggest either -V 2 or -V 0 -Y ( -Y is so that the bitrate of -V 0 doesn't increase as much with music that have strong content on very high frequencies)

As for your process of ripping, EAC is good, but sometimes it's a bit hard to setup. Winamp's ripper shouldn't be a problem, but you cannot be sure that the disk has been ripped correctly (which implies that there may be "clicks" in some scratched disks).
The encoder in winamp, as you've found, is indeed LAME, concretely the version 3.98, compiled in july 4 2008. Since i don't have the "Pro" version, i can't tell you which settings should you use (since i can't access that screen).


which will *not* help masking sounds (since we're not talking about a speaker, but towers with lows and highs separated)

which, again, doesn't help in masking sounds.

I do not like alcohol.

Ideal? headphones are indeed preferable to big speakers to listen to the music *as is*. But then again, speakers can play a role on hearing what a properly sounding headphone would not.

"As-is" is due to accuracy of reproduction in the case of the headphones. The difference you hear with any loudspeakers in any environment is down to colouration either from the equipment or the environment. It's certainly not a simple case of the loudspeakers being 'better' because they sound different. Quite the opposite for people who care about how their music sounds or at least should sound.

Cheers, Slipstreem.

And strictly speaking, not every club partisan will be consuming drugs. For that matter, not every drug-consuming partisan will be consuming alcohol, and many drugs do not affect hearing (or perhaps they accentuate it?)

volatile, I think the most anybody can claim here is that we have really good evidence for say -V3 being absolutely safe for home listening situations, and we don't have any evidence for whether or not it is safe for PA situations, but we're fairly certain that it's safe. The signal to noise ratio of the listening is going to be low enough at midrange-treble frequencies to hopefully not make it a concern. At any rate, beyond -V3 it's expected that you will not get enough benefit out of 320kbps.

This leads to the interesting question of exactly how one would test under this listening environment in an ABX test.

Of course, that's my point, but to make it a bit more notorius:
A few years back, i was in one disco, and i simply couldn't stay there more than 10 mins because the highs were so high that it irritated me. Then so, the Dj was on his 40's or 50's. Do you think this situation helps a lossy encoder?

The distortion could have been coming from any number of sources if he was such a bad DJ as to use an amount of HF-lift that would actually hurt a person's ears. Well done for walking out. You probably did your ears a huge favour.

Cheers, Slipstreem.

i would rip to flac and transcode as needed. mp3 may fall out of favour if some fantastic new codec comes along and transcoding from lossy to lossy is a bad habit as you will loose quality each time. transcoding from the lossless source will not cause a generation loss.

I wouldn't be holding my breath. There are plenty of codecs that hold greater promise than mp3 already. Some seem to be increasing in popularity, some appear to be falling out of favor. All of them are best categorized as an alternative from the standard. Perhaps when mp4 reaches a similar maturity things will be different, but I'm getting the impression that it has a long way to go. Any new codecs, no matter how fantastic, will have to start at the end of the line.

EDIT: Removed double-post.

@Kitsuned:
In response to your post below, I'm able to get burst speeds using secure mode with my drive. Like I said earlier, it depends on your hardware.

To the OP: I did have secure ripping on in EAC. I set this up years ago like this and didn't think of it until someone brought it up. Changing the mode to burst would null the speed argument I said earlier.

Thanks to all who have helped out with this post. I have some valuable opinions in there, and I'll try to answer / comment on more points below.

This is probably the way I will go... Regarding the Apple question... Serato loves importing from itunes, but i found its ripper (especially in naming / tagging) quite infuriating.

Bubbleguuum.. I just grabbed foobar2000 and looks very interesting. at the moment I can quite get my head round it tho. Lots of greyed out options for your plugin etc. Maybe just need some serious playaround with it. Thank you for this.

Jaz, awesome thanks. SO my best option to check for errors at time of ripping is EAC in "slow" mode?


To add to this, I think the perfect solution for me is (Using EAC) to rip lossless to my PCs drives where there are no size limits. Creating copies for the DJ laptop.

On the laptop, I'll probably go to v2 - v3 for my best, most important stuff, and as low as v4 - v5 for the lesser important tunes.

I think i would have most of the forum members here on-side with that process????

V

Moderation: Fixed mangled quotations by removing extra [quote] identifiers.

I would agree, with the caveat that I would do some blind sample listening to some of your most important stuff at -v4.5 before you decide to rip a lot of songs at -v2 unheard. I think you'll be rather surprised at the results.

My own experience in starting out my library at an unnecessarily high bitrate resulted in a lot of wasted time re-ripping things in order to fit more onto my iPod. Don't fall into the fallacy of thinking that, if you rip at a higher rate, you're self-evidently going to hear a difference.

Yes this would be a good decision. If these are loud, pop-ish tracks, -V4 or V5 will handle them just fine.

Thank you guys for all your help.

I got some nice sounding rips for the bars and clubs without having to fill my drive full of 320k CBRs

V

Glad to hear you've got something that works for your needs. Have fun DJing.