Please remove filter aliasing in simpler (and auto filter)
Please remove filter aliasing in simpler (and auto filter)
Hello, could you please remove this damn aliasing noise !?
Maybe apply internal oversampling ? I like working in 44100 because
of cpu consumption. But then i can hardly use the LP2 mode in simplers filter. Thats so annoying.
For those who not believe me, i´ve upped 2 little soundfiles.
Same bass sound produced with simpler (LP 2 mode), first time in 44100, second in 88200.
http://rapidshare.de/files/30075130/Alias.rar.html
Edit : I cant believe that no one is interested in that....
Maybe apply internal oversampling ? I like working in 44100 because
of cpu consumption. But then i can hardly use the LP2 mode in simplers filter. Thats so annoying.
For those who not believe me, i´ve upped 2 little soundfiles.
Same bass sound produced with simpler (LP 2 mode), first time in 44100, second in 88200.
http://rapidshare.de/files/30075130/Alias.rar.html
Edit : I cant believe that no one is interested in that....
-
- Posts: 435
- Joined: Mon Jun 27, 2005 9:26 pm
- Location: Norway
Yeah, i think the over all sound of simpler is allright. I mean,
none of the ableton devices sounds very outstanding and thats o.k.
Therefore they are stable and do not use a lot of cpu.
But you cant use the simpler filter together with its filter envelope
@44100 which is shown by my little example above. And that sucks,
because i really would like to use it for synth basses....
none of the ableton devices sounds very outstanding and thats o.k.
Therefore they are stable and do not use a lot of cpu.
But you cant use the simpler filter together with its filter envelope
@44100 which is shown by my little example above. And that sucks,
because i really would like to use it for synth basses....
-
- Posts: 7251
- Joined: Thu Sep 29, 2005 8:34 am
- Contact:
I had to listen to each of those like 4 times before I heard the difference.
do you think people actually care about that? I mean honostly.. listen to stuff recorded 30 years ago that people still love and it sounds like shit, people don't care about the audio quality, it's the music. so who cares if there are microscopic differences in the aliasing.
if it matters that much to you why don't you write your song in 44 like you like to do and then when you bounce it down switch to 88.
do you think people actually care about that? I mean honostly.. listen to stuff recorded 30 years ago that people still love and it sounds like shit, people don't care about the audio quality, it's the music. so who cares if there are microscopic differences in the aliasing.
if it matters that much to you why don't you write your song in 44 like you like to do and then when you bounce it down switch to 88.
It was as if someone shook up a 6 foot can of blood soda and suddenly popped the top.
You needed 4 times to hear it ? But for sure not on a proper studio monitor ?
But anyway, this is aliasing noise and should be removed, no point of discussion.
Thats what im doing actually, but sometimes i just want to render little basslines and stuff, and then its annoying to change the samplerate everytime.if it matters that much to you why don't you write your song in 44 like you like to do and then when you bounce it down switch to 88.
But anyway, this is aliasing noise and should be removed, no point of discussion.
-
- Posts: 7251
- Joined: Thu Sep 29, 2005 8:34 am
- Contact:
Well it's a fact, but it's not an inescapable fact. Or at least, it can be reduced.Johnisfaster wrote:isn't aliasing noise bound to happen when working at lower rates? I thought it was a fact of digital audio..
Aliasing is generally avoided by 'oversampling', simply put running the calculation at a multiple of the current audio rate. There is a downsampling part built in to the end of the oversampling so that you get your audio back at the rate of your host.
There are lots of things that get oversampled in DSP, oscillators, filters, 'distortions'. Sounds a lot better, but comes at a computational cost. Sometimes you can make use of things like SSE to offset the cost though so it isn't always as simple as 'times the CPU usage by four if we want 4x oversampling' . there are tricks. I only know the very simple ones.