The aliasing thread

[René]

no!

it's a generator. If there was a 10 days latency, you'd be able to process 10 days to finally stream out what comes afterwards. And if there were 10 days lost, you'd then fill the first 10 days with nothing.

You're still confused. I'm not talking about the system latency, but about the dsp downsampling algorithm latency. If you're suggesting you can deduct that latency from the system latency, naturally you can, as long as system latency is higher. But of course, you cannot deduct it if the algorithm latency is higher. Then you need that predictive algorithm I mentioned above.

Anyways, I don't need to convince you

-René[/quote]

[Borogove]

You all sound so stupid arguing over this.

If you're making tarnce, it's critical that there be no uncontrolled aliasing.

If you're making curnk, aliasing isn't a big deal.

[george]

You all sound so stupid arguing over this.

Easy with flames, please. You have got plenty of alternative threads for that.

[René]

"no uncontrolled aliasing"

Best phrase today

Well, I'm not quite sure what this parallel fir thing is (sure it's not a truncated iir or something?), but any filter introduces some delay. (If the present sample is based on the present sample and the past 7 samples, for example, it is clear that a sample 7 cycles old is still influencing the output -> delay). If the algorithm is smart and "looks ahead" to prepare data, I guess you could have 0 latency (and a brief clip as the sound is triggered and immediately offsets into the oversampled output), but the filter itself is still introducing delay. If I'm wrong and seemingly non-causal filters are realizable in real-time applications, please tell me. This technology interests me.

Different 'delay' means. My first post refers to delay as in 'the damn snare moved n samples', versus the 'the filter effect is delayed n samples'.

-René

[tony tony chopper]

This discussion is aimed to samplers

what's the diff? would you argue if it was a rompler?

I'm telling you that my system works just like a sampler, it's not because I don't allow the user to load theirs that it makes the process unvalid. I'm not telling you that my synth is a sampler, I'm telling you that the process I use for it works for a sampler as well. You can't test it yourself? I never said you could, you can just test that there's no added latency. For the rest you'll have to argue about the theory, I don't care as in practice it already works here.

If you're suggesting you can deduct that latency from the system latency

no..

I just don't get what your problem is with this: you trigger a key, what prevents you to render 10 days worth of audio data at this time? If it were an effect, ok, there's an unknown you can't predict: the input. But here you own the samples, you can predict anything you want.

[Borogove]

You all sound so stupid arguing over this.

Easy with flames, please. You have got plenty of alternative threads for that.

That wasn't a flame. That was a joke.

[george]

what's the diff? would you argue if it was a rompler?

I would, unless there is any way to use external waves.

I'm telling you that my system works just like a sampler, it's not because I don't allow the user to load theirs that it makes the process unvalid.

But unfortunately, you can't use external waves and comparisons with the samplers listed result pointless.

[george]

That wasn't a flame. That was a joke.

Sorry, I just overreacted

[autloc]

I just don't get what your problem is with this: you trigger a key, what prevents you to render 10 days worth of audio data at this time? If it were an effect, ok, there's an unknown you can't predict: the input. But here you own the samples, you can predict anything you want.

But then you'd either have to play back 10 days worth of samples (latency) or you'd have to break into the middle of that batch of data (risk having a discontinuity). I dunno, maybe it's safe to just chop off the first (X) samples to eliminate latency at a 0 crossing or ramp right in (though that would still involve latency), but I wouldn't think that it would be ideal. (??)

Anyway, never mind - congrats if it works. Back to tarnce and all that good stuff.

[René]

I'm telling you that my system works just like a sampler, it's not because I don't allow the user to load theirs that it makes the process unvalid. I'm not telling you that my synth is a sampler, I'm telling you that the process I use for it works for a sampler as well. You can't test it yourself? I never said you could, you can just test that there's no added latency. For the rest you'll have to argue about the theory, I don't care as in practice it already works here.

Your system -might- have zero latency. It -might- also have a pretty low interpolation noise and -might- have zero aliasing.

If you say it does, I don't see a reason for not believing you, other that I've heard the same claims at least a megazillion times.

Once again, to evaluate the result, it's required to feed a sample into it and observe the processed sample. As it's not possible to do it in your plugin, we know nothing about it. One more time, not that I'm particularly interested in your product: it was you who asked me to test Sytrus.

I can think of one million reasons why a given method might work in a synth and not in a sampler.


-René

[tony tony chopper]

I would, unless there is any way to use external waves

If your thread is about your own aliasing test and not ways to get rid of aliasing in general, then don't name it 'the aliasing thread'.

I was just coming up with a valid way to reduce aliasing. Oh wait, you can't test it.. ok then it's not a valid one.

[DevonB]

That wasn't a flame. That was a joke.

Sorry, I just overreacted

And trust me, if you read the threads that his joke refers to, you didn't miss much. I did get a chuckle myself though.

Devon

[george]

If your thread is about your own aliasing test and not ways to get rid of aliasing in general, then don't name it 'the aliasing thread'.

Excuse me, but this thread is posted into Samples, Sampling & Sample Libraries.

[WilliamK]

..

[george]

I was just coming up with a valid way to reduce aliasing. Oh wait, you can't test it.. ok then it's not a valid one.

It is perfectly valid, but being applied to a synth, it doesn't really belongs here but DSP forum instead.