Page 1 of 1

why a sinewave starts with a "click"?

Posted: Fri Jul 31, 2009 1:36 pm
by baddub
Does anybody knows why while generating a simple sinewave (no matter what frequency) it starts with a 'click', which highlights a considerable amount of spectrum around the base frequency but disappears after a few miliseconds. The situation repeats with every single note. It does not click with portamento though.

I'm using 3osc on FL studio. The notes haven't got any ADSR settings applied. As I've said before this is a pure sine.

Is it because of phase problems or maybe 3osc sucks in this respect? Maybe it's just the nature of sinewaves?

edit:
btw. the 'click' is hardly audible but it can be easily seen in a spectrum analyzer.

Posted: Fri Jul 31, 2009 1:44 pm
by serox
I think it is because the wave form does not start spot on the axis or something? this causes it to clip? not sure but I am only guessing.

Increase the attack time to remove the click anyway.

Posted: Fri Jul 31, 2009 2:02 pm
by baddub
That's what I think. But changing attack can deprive a baseline of energy, and still the 'click' can be noticed.

I'm looking for a solution as I think it can help to clean the low end a bit.

Posted: Fri Jul 31, 2009 2:10 pm
by xzazael
I think if you set the attack value to be just barely off you can remove the click without depriving the bass of its weight

Posted: Fri Jul 31, 2009 2:16 pm
by darkmatteruk
Xzazael wrote:I think if you set the attack value to be just barely off you can remove the click without depriving the bass of its weight
one of the first things i learned about production was that you actually want that click there, aslong as its not rediculously loud.

gives the kick a touch more bite

Posted: Fri Jul 31, 2009 2:17 pm
by setspeed
there are basically 2 options:

1. set the attack time to something like 8ms. this is what I would do, and I don't think it noticeably affects the sound.

2. bounce a sinewave as audio, and use that.

the reason for it is because if you think of a sine wave as going up and down above zero, if the oscillator doesn't start at 0 then when the sound hits, it starts at some random point. So the click is created by the speaker having to jump from 0 to that random point instantaneously: ie the waveform goes vertically up and then continues as a sine.

if you bounce it as audio, and edit the sample so that it starts exactly at the point where the wave form goes through zero, it won't happen.

does this make sense? :)

Posted: Fri Jul 31, 2009 2:18 pm
by macc
It doesn't start at a zero crossing. This is basic stuff really :)

Most likely this is due to the synth being used operating in a freerun/gated mode rather than being started and generated according to key presses.


EDIT - hadn't seen setspeed's post - wasn't aimed at you man :)

Posted: Fri Jul 31, 2009 2:33 pm
by serox
Macc wrote:It doesn't start at a zero crossing. This is basic stuff really :)

Most likely this is due to the synth being used operating in a freerun/gated mode rather than being started and generated according to key presses.


EDIT - hadn't seen setspeed's post - wasn't aimed at you man :)
Aimed at me? this is what I said. N?ot sure what the line is called but I thought axis would do lol.

Posted: Fri Jul 31, 2009 2:46 pm
by macc
Sorry, it wasn't aimed at anyone at all, I just took ages to type it after several lunchtime pints :lol:

Yes, it's the axis. It's very easy to understand why you hear a click when you consider that essentially it describes the movement you speaker makes. A vertical line requires your speaker to move at infinite velocity, to be in two places at once. Not happening.

Posted: Fri Jul 31, 2009 2:50 pm
by flames dean
you are all wrong. The "click" you hear when hitting the sub is your subconscience reminding you to bitcrush, saturate then bitcrush come more, resample, saturate again then compress to death then DELETE THE WHOLE thing cos it sounds raaaaaaaaaabbish and knock one out and go to bed.

Posted: Fri Jul 31, 2009 2:54 pm
by macc
LOL!

The voice of experience!

Posted: Fri Jul 31, 2009 3:00 pm
by flames dean
'Ello Macc

Do you make D&B tunes too cos I thought I saw your name on Beatport or one of those online shops...

Oh and ... just to be original - thanks for all your pearlers on here bruv - some real useful stuff there that might a good few years for a lot of the young'uns to get their head around (gain structure etc...)... oh yes i remember the old days when you had to find someone who knew their ish before you could find out ANYTHING about mixing.

but yeah - ta anyways mate - i hope you get some business through the generosity you've shown all over the place on this forum.

Posted: Sat Aug 01, 2009 12:06 am
by omm-0910
darkmatterUK wrote:
Xzazael wrote:I think if you set the attack value to be just barely off you can remove the click without depriving the bass of its weight
one of the first things i learned about production was that you actually want that click there, aslong as its not rediculously loud.

gives the kick a touch more bite
Personally I'm not keen on the clicking sound. IMO it's better to add a tiny amount of attack with the ADSR, then reshape the transient with some careful compression to retain that 'kick'

Posted: Sun Aug 02, 2009 3:15 am
by jsilver
as far as that click goes, thats the sample... you need to find optimal loop.. IL Directwave does it

Posted: Sun Aug 02, 2009 3:18 am
by deadly_habit
invest in some analog gear cheap or high priced
vco def better than the digi with clicks

Posted: Sun Aug 02, 2009 3:19 am
by darkmatteruk
OMM-0910 wrote:
darkmatterUK wrote:
Xzazael wrote:I think if you set the attack value to be just barely off you can remove the click without depriving the bass of its weight
one of the first things i learned about production was that you actually want that click there, aslong as its not rediculously loud.

gives the kick a touch more bite
Personally I'm not keen on the clicking sound. IMO it's better to add a tiny amount of attack with the ADSR, then reshape the transient with some careful compression to retain that 'kick'
im not that keen on it either to be fair

Posted: Sun Aug 02, 2009 8:41 pm
by ngyoshi
i have this problem sometimes as-well, i just export the bass line and automate the volume at the mix down point,

sometimes it is possible to use the ADSR envelopes but i always find it takes to much energy out of the bass line when using a subtractor in reason 4

Posted: Sun Aug 02, 2009 8:51 pm
by scooterjack
6-9ms attack will take away the click, and won't take any 'energy' out of your bass..


this is basic, basic stuff here.......