If you set your samplerate to 44.1 khz in Renoise and play this 192 khz sample, you can hear some tones that shouldn’t be audible.
The sample contains just a 80 khz tone (inaudible), which on playback gets (aliased?) down into the audible range at around 8 khz.
From the linked sample. Incredibly noisy. Also, this particular sample isn’t the 80khz test tone. It contains multiple test tones midway through the track.
From the 80khz test tone sample from the linked forum thread. This is the one that should have no audible tones. The test tone definitely gets aliased and ends up audible.
Who needs 192 kHz? Unless you’re working in a professional high end mastering studio…
And it’s not a problem of Renoise, it’s a problem of the source to destination rate downsampling process while playback. 192 kHz are no multiples of 44.1 kHz, it’s a multiple of 48 kHz. So downsampling 192 to 44.1 can produce disturbing artifacts, which you then will recognize as your mentioned audible “aliazing”. That’s normal, if you try to downsample a rate that is no multiple of the destination rate…
It’s like you try to push a cube through a triangular hole…
It is interesting to check it on https://src.infinitewave.ca/. Even some most expensive DAWs are not perfect for sample rate conversion. A solution is the free software r8brain to do offline sample rate conversion of your samples/rendered tracks.
Yes! Very interesting! After watching this video I checked the saturation effects I usually use with a high pitched sinewave like he does and discovered that some of them add a LOT of aliasing (a reason why my mixes sounded harsh, I didn’t know why until that video) so I don’t use them anymore.
Yes it is possible (all sample rate conversion using sync interpolation do this) AS LONG AS there is no audio with frequency above half of destination sample rate (e.g nothing above 24kHz if you convert to 48kHz). Otherwise it will result in aliasing noise, unless the resampler has a built-in low-pass filter. As different resamplers use different filters that’s why there is differences between them.
So yes it’s better not to have ultrasonic tone before conversion. I hope this was clear enough
What? I literally posted a sonogram of each of the test tones linked by OP. How is that non sense?
I also stated as much in my post. As well as objective fact that the 80khz test tone will alias down into the audible range because Renoise, as best as I can tell, does no filtering when resampling to a new Sample Rate. (Which, for me, is a good thing, i.e. sound design.)
If you want to prevent ultrasonic frequencies from aliasing down when down sampling in Renoise, treat the sample with a Low Pass filter set to your Nyquist frequency before down sampling.
Guys, please keep in mind, Renoise is a real sampler, while other DAWs are not. I think only if you actually can automate the pitch shifting/time stretching parameters in the DAW, it is calulated in realtime. Not sure if ableton etc do that, I suspect those to do pre-calculation.
A proper comparison regarding aliasing would be Renoise vs Kontakt vs Falcon vs Bitwig Sampler module etc