Recording your mixes on CD

[Juggalo Muli]

Ok i have recieved a few demos from fellow DJs from swapping with them etc... and well i have a question. HOW THE HELL do you get rid of the static that comes from records, all these CDs sound so professional with out any of the hiss and crackiling !!! what am i doing wrong, i just have my mixer hooked up to my pc, then use cool edit to record my sets, and then burn it. What can i do to make it sound more professional for my demo CD??

Cheers

Muli

[phunkdust]

well is your soundcard any good? soundcards are the kingpin in any audio-recording setup... shitty ones will cause hiss

are you sure that only your line-in is selected on your soundcard as a recording input?

are you recording 44.1khz, 16bit, stereo?

what's your mixer like? most mixers will pick up hiss in the preamp process (ie TL074 stages boosting phono to line - picks up lots of hiss in the process... no doubt driload will tell me they're not TL074s :P)

using decent cables?

static straps in place on your decks?

records clean?

cartridges balanced and set?

not using your setup near 240v cabling, power transformers or plugpacks, fluro lighting, etc...?

just a few things to check there

[Bracko]

I think Juggle means the general static sound you get while the record is running... the sound you get at the end or start of a track.

Also clicks and pops phunkdust.... how do they happen on the record? is it bad care for the record, my stylis is rubbish... ?

cause its starting to happen on a few of my tracks, so i'm worried about playing the tracks i wana keep forever.

[Khem]

Run your mix through a mastering program, something like T-Racks can give your mix a lot more body.

As for the static, there is some option in Sound Forge (and most other sound apps I would assume) but I cannot remember what it is...

[ELiTe]

dont touch the record for starts... i got told this yesterday....
especially for just music on records..samples/scratchin its ok...



and use a record cleaner...u know they clean dust off...
thats my best ...:)8-)

[phunkdust]

clicks and pops are attributed to a few things

A - static (the most likely)
B - bad cart setup

rule of thumb is, handle records as little as possible, wipe with an anti-static cloth occasionally, make sure your deck is tuned to perfection.

as for dirt and dust, a very mild solution of dishwashing detergent in warm water and carefully wipe the records over with a damp cloth soaked in the solution.

just dont put them through the dishwasher :lol:

[psycho-martian]

mate, i had the same problem a few weeks ago, had static and lots of crackling when recording to a pc. all caused by dirty records and needles.

here's a couple of points to be aware of:

- make sure ya dex and mixer are earthed
- balance ya carts and tonearm correctly
- use an anti-static cloth or carbon brush to remove surface dust
- make sure ya stylus is clean before a mix session
- use shielded interconnects between ya mixer and pc

most of the time, the crackling and pops are due to old or damaged records (eg. scratched, dodgey pressings, crushy shit on ya records etc etc), if in doubt scrub your records with steel wool and you'll be right :). otherwise just record as normal and using one of those de-clicker/de-noiser plugins to remove the shit.

hope that helps.

[defcon_mofo_1]

i like the clicks, pops and hisses.

keep it real I say...

[driload]

the "option" someone mentioned for removing pops in Soundforge...

theres no native plugin to adequately (stress that word) do the job but you can zoom in on big clicks and pops and use the pencil to redraw the waveform


or else go get the Steinberg DeClicker and DeNoiser package

very nice, very effect


it makes an "image" of noise and pops and uses that as template to remove the nasty stuff

seeing as its just a dj mix and nothing too sonically important i wouldnt bother really, but you may like to run the click and noise removers and then normalise it up a few more dB (rms not peak)

DONT get carried away compressing and going completely silly on the normalise, just a few dB difference.... like i said its just a dj mix.

[kaossproject]

I would suggest Steinberg "Clean" as a good mastering tool that gets rid of all unwanted noise, pops and crackles.

From personal experience.

[Bracko]

STEEL WOOL... you've gota be kiddin'

I think the main prob will be cables and stylis.

I'm using pickering NP(bottom of the range ones :( ).... which has a weight of 3-7 grams... I've set it on about 4.5 - 5g

would buying a pair of shure MK44's fix this?? is it worth it?? and if i did would they ware my records any less??

[driload]

there are, like Phunkdust initially said, an insanely large number of variables you must contend with here.


you need to actually advise us what MIXER and SOUND CARD you are using. most of your sound cards will be noisy as hell, and cheap mixers certainly dont help.

much of an audio engineers life is trouble-shooting and we need specifics to be useful to you, otherwise im gunna type a fuckin essay on the fundamental of noise reduction, unity gain, signal flow, the importance of clean connections, path of least resistance, quality components, reducing rf nosie and other interference etc etc

rofl

its a noisy old world out there,,,, you sure that crackling doesnt sound better now? :P

(ps, FIRST line of attack is ensuring no dust on records, second is no dust on FRESH stylus....)

[Khem]

Originally posted by driload
the "option" someone mentioned for removing pops in Soundforge...

That was me... was actually talking about removing the static... I think it might be DC Offset...

[driload]

nope

DC offset removes.... surprise, DC offsetting

think of it as "centering" the waveform in terms of the amplitudal positioning, not spatial positioning....

how better to describe DC offsetting guys?

[kaossproject]

Yep.......that will do:tripping:

[phunkdust]

i'll just translate what driload said about DC...

DC offset bad! Want remove DC offset! Removing DC offset is good!

:lol:

[polarbear]

right then.

if you can get good quality, new records.

get new needles. build the cartridges carefully and well.

ground your turntables. get a good mixer.

use a quality audio interface.

mix carefully.

NOW

BEFORE you compress or limit:

clicks and pops can be removed using plugins, or by deleting bits of the waveform, or by redrawing bits. BE CAREFULL. It's a good idea to select individual clicks and pops that are bad and process these rather than process the entire waveform - it helps to reduce artifacts.

Hiss can be removed with many different packages. I like Hybrid Neural De-Noising from SonicWorx. EQ can work.

Then , when you've got it as clean as possible, then do some multiband compression, MS mastering, limiting, EQ or whatever.

NOW THEN DC OFFSET

DC offset is a funny term - in a time domain based digital signal, it amounts to the average of the signal set. In an analogue signal, it amounts to a DC voltage.

It's not necessarily 'unwanted'; although many analogue mixers contain many HPF based DC offset removal stages.

It can be generated in the analogue domain by bad grounding or powersupply problems, a badly aligned cartridge, or incorrect phantom power. It can be generated in the digital domain by granular effects (pitch shifting, time compression and expansion), modulation effects and others.

DC offset can create warmth in speaker systems - it acts like a bias voltage in an amplifier. This is used in guitar amps and tube amps to put the signal into the right (or wrong) part of the transfer curve. To remove DC offsett, you can

1) Hipass filter at a lo frequency (5 Hz). After all DC is really just a signal with a frequency of 0.

2) Use a high quality transformer - DC can't pass across an 'air' gap. DI boxes can contain transformers to match impedance.

3) Perform an FFT and remove the first basis signal, then re synthesize. You have to use a discrete cosine transform to generate a basis signal at 0Hz.

Two of these processes (1 & 3) amount to a filter, which can distort your signal in other ways - for instance, if you have a LF signal (20Hz) which gets louder in volume over the track, it can create the impression of a rising DC offset, and when you perform a FFT based DC offset removal your signal then can then be modulated in an different, unwanted fashion.

Think about it; do you really want to take the average of your entire signal set, then subtract it from every sample? doesn't make much musical sense, does it??? maybe you have high average signal because you've put LOTS OF BASS in and it's all in funny phases. or heavily chorused. Suddenly if you do a DC offset removal, you're CREATING a DC Offsett by recentering (removing an average) the waveform.

The bears word? Only remove DC offset if you're sure there's a problem, and you don't want it there.

[polarbear]

OH YEAH!


DC offset is a funny term - in a time domain based digital signal, it amounts to the average of the signal set.

if you don't treat it as a a signal frequency of 0Hz and HPF or FFT it out, then , you buffer some signals - (with a latency of 10ms @ 44100kHz thats 441 samples), add them all up (using 32bit float math), divide by 441 and subtract this from the samples.

ANYONE SEE THE PROBLEM?

this creates a new, moving DC that's created by any waveform with a period less than 10ms.....it would look like a form or amplitude modulation.

so, if you're talking about a realtime signal rather than a stored sample set then you should ONLY remove DC offset using a HPF or FFT.

THE QUESTION YOU NEED TO ASK YOURSELF IS

WHAT SORT OF DC OFFSET REMOVAL IS MY SOFTWARE DOING?

WHAT IS IT DOING TO MY SIGNAL?

[Juggalo Muli]

ummm maybe those pops and crackles are not that bad :P

But i think its just my soundcard. Owell who cares :P

[kaossproject]

wicked feedback polarbear......that is fantastic stuff.

you are a legend indeed!!:)

[dks49]

yep ditto. Ofcourse u could just switch to CDs.:p

[Vital]

Good luck with implementing a DSP with FFT between youe mixer and CD for the purpose of a mix CD. I don't know if you've ever had to write a program implementing a FFT but it's a bitch, and if you have (and it's in C) could you send it to me as I'm in the middle of doing it now.:(

Other wise don't stress about it. DC offset doesn't cause hisses and pops. Just make sure everything is well earthed.

The hisses and pops are just noise form the the record, needle, and mixer. But mostly the record/needle. When I was buying a mixer I noticed THD's of about 0.05% or so. If you can notice that then you're a better person than I am.

[polarbear]

ok. the fastest FFT algorithm in the west is FFTW.

You'll find precompiled binary library's for all sorts of platforms as well as source code and documentation at:

http://theory.lcs.mit.edu/~fftw

this is the shit man - it's the algorithm that makes NI's spektral delay work.

[Vital]

Thanks you've just saved me a SHITLOAD of time and effort.:)

[driload]

well i'll be damned

they won the Wilkinson Prize



thats way out of my depth though. im a goin' back to "the school" next year to touch up on my mu-tek but hell guys,


... just how l33t do you wanna get?

what ever has driven you to study FFT's?

[Vital]

Was that asked to me or Polarbear?:?

Either way I think I can answer it. FFT is very important when it comes to signal processing.

a BREIF explanation of the FFT (as I understand it):

A Fourier transform gives the frequency spectrum of a time domain signal. It involves an integral. Since computers can't perform and integral, it was a long algorithm to implement. Meaning you couldn't sample at high enough frequencies.

Then (I think) 2 guys from IBM used some clever maths do make the discreet fourier transform into the FFT wich runs about 200 times faster. And for large numbers of samples or frequencies to analyse, computation time rises nearly linearly. As opposed to the DFT which rose exponentialy.

SO...

if you're doing signal analasys in real time, the FFT is the only way to go. It's uses are boundless from music studios to mobile phones to etc etc...

And I have to implement a real time spectrum analyser using a 16MHz Mitsubishi micro-controller. I programme the Mitsubishi using C and since it's being tested on an audio signal I have to be able to sample at 44kHz :~( , hence the FFT.

Sorry if that was a bit long winded.

[polarbear]

> what ever has driven you to study FFT's?

understanding. I have a burning desire to understand what all of these plugins are doing, so I started to teach myself DSP. maybe then I can rest my paranoia about what the plugins are doing.

I have a programming background, and have mucked around with synthesis techniques; I've been making little VSTi's which is fun.

BUT

What I'm really interested in is convolution. Using impulse responses to model signal paths, reverb, 3D audio etc etc.

most efficient convolution techniques involve FFT. (though the G4 is fast enough to do brute force convolution - that's where ALTIVERB has come from.) Lake DSP have some patents in this area - convolution is used to generate their 'dolby headphones' effects (5.1 --> headphones).

anyhoo.......

[driload]

you guys would love my dad then

this is right up his alley. he started as a sensors engineer and ended up getting silly (ie, losing me) on these algorithms and associated crap

think Big Pond to Navy to all the mines in australia


i applaud your l33tness guys


i learnt enough programming to get around the hexadecimal world of midi and build a few devices but as satisfying as it is im not very good at the more complex programming stuff. i am currently wanderng down the endless path of musical pedagogy in the fields of maths and science

even the creation of the piano and scales is amazing stuff, let alone the complexities of indian music, welsh choirs, celtic chanting, eastern europian rythms... oh boy. im off to europe for a year when my girly gets out of police academy and im finally finished ALL uni/private tuition for good. working musical holiday/studio/experience.... bring it on



i guess my point here is.... who woulda thought people easily steriotyped as "techno wankers" by the unknowingly public, actually pursue such intriguing paths of knowledge!


tip my hat to you guys

oh and apologies for typos.... im typing with bandages on my hands.... dont ask.

[danielmarshall]

Hello fellow maths nerds.

I was just wanting to know if any of you guys have done any compex analysis (what a bastard of a subject hey) and are willing to help me out with some nasty stuff regarding Cauchy's and the Residue Theorem.

Thanks

Dan

[driload]

fuck me not another one!