I'm not sure what you mean because your statements clash with each other, but I think you meant people should make better music and worry less about the technicalities. I can totally agree with that. I've seen people with not much technical knowledge making great sounding music and mixes, and I've seen very knowledgeable people making shit music or shit mixes, or both. In the end it all really comes down to the results, as always.

 

He's not talking about oversampling

 

Well and he may be a 'engineer' and no artist so he maybe is not into production but into masterin mixing very 'quite' material or what ever .

 

But oversampling removes aliasing, although there are other techniques to remove aliasing, indeed, but the most common way is to just use oversampling, and yes - some companies use rather crudely made oversampling code, probably just to be able to say "it's got OS!".

Anyway, I think you are checking the plugins through VST Analyzer? I do that too. That's the only way to get the whole picture about the plugin.

 

This is not true, the SRC is done at the pluging level so if the DAW sample rate is higher, this dosen't mean you won't have aliasing.

96khz - Waves CLA76 - sinewave sweep



It all depends on the coding / implementation.

 

But he talks from plugins not using oversampling but code and filters to do that .

 

What test this? I was talking about anti aliasing measures which obviously include oversampling... I clearly said many don't do a great job at it
These plugins that oversample do & the ones that don't have taken aliasing into account because it is very small compared to other plugins
When oversamplings

I think you should be careful, you hold a lot of power in this forum sometimes you should sit back and moderate instead getting involved all the time. clearly deliberately misconstrue my messages to prove a point - I'm not being rude I just can't see how you don't understand my points. The thread exists to point people in the direction of plugins that alias the least without having to up your sample rate. Ok? Have I said it better for you

 

Last edited: Feb 18, 2021

I'm talking about developers that do anything they can to prevent it but I just stress it's about the implementation mainly. not just oversampling

 

You mean how I tested it? With DDMF's Plugindoctor.

Ah, now I and probably some others here, got it. You're talking about any method to reduce it. Would've been useful if you've stated it right from the start because, obviously, I wasn't the only one who didn't understand what you meant exactly.

This is highly subjective.

I beg your pardon???

Thin ice, bro.

 

since the TE is only trolling himself he probably deserved that.

but of course you can theoretically also dont antialias. only it is highly questionable to accuse people of doing that and then not goving examples.

what i´ve beeing seeing is that people are afraid of filtering not enough and then going below sr/2 to be sure but i dont think anybody will forget it completely (because as you said, this is the whole point about it)

plus decimation is easy as fuck when you upsampled for multiples of 2.

and all the rest, esp. interpolation methods, is a matter of taste.

 

1) May I ask what the source of the sine sweep was? Wavetable?
2) Also, 1176/CLA76 adds saturation/harmonics.
If unlucky, you have two sources of harmonics, which you see in the spectrum analysis. Two misses out of two shots.

 

and we return to this video again!

 

i can´t name an example now, but it is possible that a plug-in does oversampling only when required, i.e. does not do it when the samplingrate is 192 or 176.4 already.

or at least the user might not press the x4 button so quickly.

 

well yeah, but don't you lose the latency advantages like that?
and does oversampling done right not causes latency?

 

Two Yeses from me.
With double the samplingrate the latency in ms is halved (because it's calculated on samples).
And the higher the oversampling rate the higher the latency is, no doubt.

 

Then I need to repeat this

 

Here is the test file:

http://src.infinitewave.ca/TestSignals.zip

I used the Swept_24.wav file, that looks like this:



I have set the project sample rate to 96khz in Reaper, imported the test file and processed the signal with waves cla76 bluey, and exported the file at the same sample rate.

 

Not really correct. Your test is flawed. See the aliasing (the peaks below the test signal close to 10kHz) in the following pics at increasing SR (44.1, 96, 176.4, 192) for Waves CLA-76 Bluey

Spoiler: higher SR, less aliasing

Please keep in mind 96kHz SR does not remove aliasing completely.

 

I was saying the same thing, the aliasing is very low like -85 dB, but running the project at a higher sample rate will not remove aliasing.

 

In this specific case, yes.

Neither OS, nor a higher SR removes the aliasing. These are methods to reduce it. And of course they have the same effect on a single source. The difference is, a higher SR affects all sources, the whole mix, a higher OS only a specific plugin.
So 96kHz SR = 2x OS at 48kHz in all plugins.