return to tranceaddict TranceAddict Forums Archive > DJing / Production / Promotion > Production Studio

Pages: [1] 2 
Aliasing....seriously, who the f### cares?
View this Thread in Original format
cryophonik
Seems like I'm seeing more and more people complain about aliasing lately. And, I don't think it's a coincidence that most people who do complain about a synth or effect aliasing have no music available to listen to. So, I'm guessing that these are mostly guys that spend their spare time looking for it so that they have something to complain about, which is a nice convenient excuse for not having any music to show.

I've owned many synths that are considered bad my many because of aliasing, yet not once has anybody said to me "that pad you have layered in under the lead, you know, the one that sounds like it probably came from your Virus, yeah, it's aliasing badly." If/when my inbox becomes full of emails like that, I'll reconsider my position. Until then, see my thread title.

The end.
Looney4Clooney
its pretty had to hear unless you are using a really clean sine wave in the upper register. And really bad aliasing is like the new side chain.

I think the same thing that people go on about are linear phase EQs. They tend to sound worse.
meriter
for those that don't know.. what is aliasing?
Looney4Clooney
when you look at a car and the wheels are spinning backwards yet the car is going forwards and the driver is not black or latino. It does pain me that someone would ask such a question.
vercetti
Who the cares? Old ****s on Gearslutz that spent 200k on "true analogue" gear and write tunes that sound like extremely ty copies of Tangerine Dream or Jean Michel Jarre. Then they complain about "kids these days" and "when IIIIIIII was youuuuung there was no bad music being released at all"
tehlord
I never used to care about it until I knew it was there. Most of it can be combated with a little low passing, but I do think in the world of softsynths and tracks with 100 channels it does add up to a noise in the mix.

It seems to affect hardware VA's a lot less than software VA's, I assume because of the superior inbuilt conversion of most hardware synths.

You only have to play with Vanguard for a couple of minutes to see why some people moan about it. What an awful racket that thing makes.
echosystm
quote:
Originally posted by meriter
for those that don't know.. what is aliasing?


Long version:

https://en.wikipedia.org/wiki/Aliasing

Short version:

Natural signals (obviously) have no sample rate or what you might call an effectively infinite sample rate. Digital signals on the other hand do have a sample rate and noise can occur as a consequence of this.

The most common case of aliasing is in plugins such as compressors, limiters and "saturators", because they alter the harmonics of a sound. The higher your sample rate is, the more accurately you can represent high frequencies. In other words, if your sample rate is very low, the time between samples will be too slow to capture fast cycles. When harmonics are added, they will often extend beyond the audible range. However, your sample rate is often not high enough to to represent these harmonics accurately, so they are interpreted as frequencies occurring within the the audible range. Thus, you end up with noise.

Aliasing is never good. It is a byproduct of insufficient sample rates in digital signal processing. Although a sample rate of 44,100 covers the full frequency range of the human ear, it leaves little room for harmonics that extend beyond it. Thus, aliasing is a problem because most people are still working at 44,100 and increasing your sample rate too far introduces other problems. Worse yet, aliasing is generally considered to be cumulative. Thus, you might not notice it on one track but it will destroy your song if it is occurring on every channel.

Lots of techniques exist to address the issue. For example, some insanely smart people people have devised oscillator algorithms that are alias free. Similarly, there are filtering algorithms that can be used to reduce aliasing. The shotgun approach is to upsample the signal massively, to ensure nothing hits the limit of the sample rate (Nyquist) and bleeds back in. However, all these techniques have their caveats.

Shorter version:

Use 88,200 as your project sample rate and enable 2x oversampling on any plugins that are going to introduce a lot of harmonics. "Vintage" compressors or saturators are the most common culprits. Also look out for poorly written plugins. For example, EQs should not ever have aliasing, unless they add some saturation bull to try and make it sound more "analog". However, one of the EQs in Logic truncates the signal and this does cause aliasing.
clay
aliasing is just crappy technology used for saving megabytes :p
Looney4Clooney
you can't have a saw wave unfiltered in the digital realm without aliasing. it is physically impossible.

And aliasing is really a phenomenon that happens everywhere which is why i mentioned the car going forward while the wheels going backwards. That is aliasing. Your eyes can't sample the rotation speed of the wheels so you get an inaccurate reconstruction of what is actually going on.
tehlord
quote:
Originally posted by Looney4Clooney
you can't have a saw wave unfiltered in the digital realm without aliasing. it is physically impossible.

And aliasing is really a phenomenon that happens everywhere which is why i mentioned the car going forward while the wheels going backwards. That is aliasing. Your eyes can't sample the rotation speed of the wheels so you get an inaccurate reconstruction of what is actually going on.


If you make a low frequency fart sound with your lips while looking at those wheels, it's organic anti aliasing.

True story.

cryophonik
quote:
Originally posted by tehlord
If you make a low frequency fart sound with your lips while looking at those wheels, it's organic anti aliasing.

True story.


Is there a plugin that does this?
tehlord
quote:
Originally posted by cryophonik
Is there a plugin that does this?



There is, but it aliases.
CLICK TO RETURN TO TOP OF PAGE
Pages: [1] 2 
Privacy Statement