Looks like Renoise makes a better conversion when not using Cubic ?
not exactly sure how this site went about setting up renoise for this.
Also not sure how Renoise deals with a 96Khz sample when the preference sample rate is set to 44.1.
worth a look in any case.
Related thread from this summer…
I must confess that I was a bit disappointed by renoise’s performance in this ‘compo’. But then again, even having read the faq and help sections of the site, I’m not exactly sure how to interpret most of the data. … 
ahh yes KMaki, I didn’t notice. thanks for pointing that out.
ok I think I see now. I doubt I would want to render most things with Sinc as I won’t be getting what I hear.
currently I have Renoise operating at 96Khz and I’m familiarizing myself with the apps operations at that rate.
I’m using mainly the Acoustic Guitar instrument. it seems to be playing within that 96 base rate ok.
not checked if it’s up sampling any instruments by design. not sure what happens if you render that out at that rate either.
I mean if the Cubic spectrum looks different that way.
They use a swept sine in those tests and those are pretty extreme cases anyhow.
in most cases I don’t care what Renoise is doing, I like the sound of it.
problem is if you render out and don’t get what you were hearing in the listening condition.
but it’s good to know what’s going on. 
this is the proper attitude to have.
yeah, cubic isn’t as good as sinc, but (to my ears) it sounds better, so i render with that even if it’s objectively worse.
haters gon hate.
Yep, shouldn’t matter to you unless it sounds bad and then it’s good to know why.
And usually when converting to 44khz the results in what you hear are not so different, but when going to 22khz the differences are huge. Most programs give really crappy results with high frequency content and some (like Voxengo r8brain) are so good you wouldn’t believe it’s 22khz.
(Yes, 22khz is crap, but mobile devices don’t have infinite amount of memory)
I had some file recently which a number of apps wouldn’t load. soundhack reported it had no file header but wasn’t able to anything with it either.
it seemed to be seen as 11Khz and was a small file… Audacity loaded it and I saved it out as 44.1. then I converted it in Renoise down to 22K
because I couldn’t understand why the file seemed so hi fidelity. the renoise conversion to 22K realy dulled the sound, so I have no idea how
the sound as 11k had such fidelity. I have no idea how the wav file had this ability. it made me wonder if there was some kind of companding
system in this wav format.
I guess what matter here is the frequency ranges picked to focus on.
I suspect Renoise focusses on either mid-range or variations on every level whereas perhaps it is smarter to focus on the frequency scale that is present the most. (interpolate the lows if the sample contains a lot of lows, interpolate the mids if the sample contains a lot of mids etc.)
Wav is a container format, nothing says it has to be PCM encoded audio that it holds. It is perfectly possible that the 11kHz version was using a different coding/compression.
Also a sample of 11k sample rate has a 5.5kHz bandwidth (well just under) not 5.5kHz as highest possible frequency, which is a common misconception of Nyquist’s theory. I don’t know if you can set this within Wav but it is perfect possible to record say 1kHz - 6.5kHz audio at this rate, which depending on the sound may or may not give a clearing sounding result than 0Hz - 5.5kHz.
interetsing. maybe I’ll try and put the file online somewhere sometime. not here though, because of the restrictions.
anyone know if renoise retains the interpolation signature you get (in) renoise when you import and then export audio from renoise. ?
I would assume it should, but would like to know how one would know for sure.