No Discussions? (advanced Settings)

[Nakasara1]

Hi i wanted to start a discussions about the advanced settings.

I just wanted to know which anti alias filter value is the best?

and what does the ''skip tempo adjustment for quality'' mean?
or asked different, which one should i choose for the best result. does it have any affect on the 432hz tuning itself?

Just curious, hope somebody can explain me this deeper and what anti alias filter means, does it remove noise?

THX

[mysteryx93]

Anti-Aliasing, you'll see the effect if you slow down the music by 2x for example, then it will be beneficial. Otherwise, with more subtle shifts, it depends on your taste, it may not benefit you. I recommend it mostly for when applying large speed changes.

"Skip tempo adjustment for quality" means that, by shifting the pitch, it will also alter the speed and audio length. It will skip the pitch shifting algorithm and only apply a sample rate change. Less operation = better quality.

[NintendoManiac64]

It's worth noting the anti-aliasing checkbox does nothing when using "Skip tempo" (and probably should get greyed out?).

The anti-aliasing basically tries to smooth out the artifacting (aliasing) that comes from altering pitch. You know the stereotypical "auto-tune" voice effects? Those are an example of aliasing artifacts from pitch alteration.

Graphics work in a similar fashion and is where the term "anti-aliasing" is most commonly used - take a look at the following issue
I opened years ago for a completely different piece of software where I'm talking about graphical aliasing:

• [Enhancement] Improved program window scaling via super-sampled integer nearest neighbor melonDS-emu/melonDS#1100

---

As for "skip tempo", it basically skips the pitch-correction processing and instead relies purely on how playing a waveform slower automatically reduces pitch; refer to this HydrogenAudio forum thread I made years ago:

• https://hydrogenaudio.org/index.php/topic,119735

tl;dr if you play a 48kHz audio waveform at 24kHz, it plays for twice as long with half the pitch without altering the waveform in any way.

From there, the only thing remaining to be applied is resampling because, spoiler alert, you can't save an audio file with fractional sample rates (and some non-lossless formats only support specific sample rates - Opus is so restrictive it doesn't even support the CD-audio standard of 44100Hz). Therefore, unless you do things manually like I layed out in the hydrogenaudio forum thread, you really need to resample the output to fit into a more standard sample rate which is what the "Sample rate" setting is.

The good news is that, by now, resampling is an extremely well-developed technique to the point of being "perceptually lossless" if your output sample rate is high enough (though your hardware needs to be able to handle it - no point outputting at 192kHz if your hardware just outputs everything at 48kHz anyway!).