hi ive been reading on how to create a single cycle waveform using the pencil. i was reading some renoise professional artist reviews online and i saw that some artist hand draw all there waveforms. i read on the renoise forum and it seems other people are confused about this too. when i create a sample, it defaults to 168 samples, mono, 44100, 16
should i always use these default settings rather then changing them? then i was reading that there is a right and a wrong way to draw a single cycle waveform in the sampler. (something about 1 over and 1 under only to draw it correct. try as i may i could not figure out what this means. when using the pencil tool the pencil tools draw many over and under lines so im not sure how i could only draw one.
to confuse matters worse, i ready that certain settings in sampler need to be adjusted or the drawn wave will be out of tune. (something about an a-4 100 samples would make the drawn wave in tune. and that default settings create a very close to in tune c.
renoise is the most confusing daw i ever used. i know some people ‘get’ it. but for me, its very difficult. i think because its sort of an modern oldskool approach. help would be appreciated.
i ready that certain settings in sampler need to be adjusted or the drawn wave will be out of tune. (something about an a-4 100 samples would make the drawn wave in tune. and that default settings create a very close to in tune c.
I think that as long as you have the sample rate set to 44100 in the audio preferences, the default sample creates a waveform that’s tuned properly.
When creating a new sample, I change the bit depth to 24 bit… although I can’t explain why I just know that 24 bit is preferred to 16… and it might make sense to use 32 bit since that’s what Renoise uses internally.
But maybe the hand-drawn waveforms are low resolution enough that 16 bit is just fine, and that’s why Renoise defaults to it. I don’t know. Hopefully someone who knows more can jump in.
I wouldn’t sweat the over/under stuff. Just draw your waveforms… use the scope in mixer mode to double check that it’s in tune (it should be!). Then go ahead and play around.
edit: I have noticed that hand-drawn waveforms seemed to be tuned down an octave… that is, if you play something at C2, you’ll see it show up as C1 on the scope. Could just be my drawings are whack and leave out the fundamental somehow, but I just remap the sample to C5 (because it defaults to C4) and then the octave looks right in the scope.
The reason for using 168 samples is that it’s close to a C4 (default keynote), close enough for most uses. The higher the pitch, the more it goes out of tune. 168 samples is close, but the exact frequency of a C4 is actually 261.63Hz while 168 samples makes 262.5Hz. To calculate how many samples you need in one cycle for a specific note you take the sample rate 44.100Hz divided by the frequency of the note. Here’s a chart that tells you the frequencies:http://www.phy.mtu.edu/~suits/notefreqs.html
You won’t automatically get a higher quality sample if you change the sample rate, it’s the length of the actual single cycle that matters. The more samples per cycle, the higher the resolution and lower the pitch, which means you have to change it’s keynote to a lower note.
It’s also usually preferred that the drawn sample is balanced, basically that means that the positive and negative sides of the wave should be equal in percentage. Not like 90% above 0 and 10% below, but 50% both ways.
The reason for using 168 samples is that it’s close to a C4, close enough for most uses. The higher the pitch, the more it goes out of tune. 168 samples is close, but the exact frequency of a C4 is actually 261.63Hz while 168 samples makes 262.5Hz. To calculate how many samples you need in one cycle for a specific note you take the sample rate 44.100Hz divided by the frequency of the note. Here’s a chart that tells you the frequencies:http://www.phy.mtu.edu/~suits/notefreqs.html
it seems that thats insanely close if not even an exact match. so, your saying that it IS best to keep the default settings because it seems human ears can not even tell the difference. if one does keep the default settings, when creating a sample, is the sampler settings, keyzones, waveform settings, ect etc all configured to work out of the box? and that i need not worry? simply pencil in waveform, and that renoises does the rest?
it seems that thats insanely close if not even an exact match. so, your saying that it IS best to keep the default settings because it seems human ears can not even tell the difference. if one does keep the default settings, when creating a sample, is the sampler settings, keyzones, waveform settings, ect etc all configured to work out of the box? and that i need not worry? simply pencil in waveform, and that renoises does the rest?
Not really, you could get closer, if you use 2697 samples and tune it to C1, but it doesn’t necessarily mean that it will sound better. Use your ears and a tuner if you don’t want to deal with maths and stuff. GTune is a nice free vst for that.
If the sample size is a bit off you can always finetune it in the sampler too.
You shouldn’t really worry to much about it, there is really no right or wrong. Take a guitar for instance, if it’s tone was in perfect pitch it wouldn’t sound like a guitar at all, but you still want it to be as close as possible anyway.
it seems that thats insanely close if not even an exact match. so, your saying that it IS best to keep the default settings because it seems human ears can not even tell the difference. if one does keep the default settings, when creating a sample, is the sampler settings, keyzones, waveform settings, ect etc all configured to work out of the box? and that i need not worry? simply pencil in waveform, and that renoises does the rest?
You’re probably aware of this, but what you pencil in there has to be only 1 cycle if you want the pitch to stay correct, if you add 2 peaks in both directions (2 cycles) you suddenly get a c5 instead. You can however make it double the length of 168 samples, which is 336 samples, then you need two cycles to get the standard C4 note. This way you can make more complex waveforms and it will still be in tune.
I suggest you write and play a c4 in the pattern editor and solo it while you draw, this way you can hear its changes while you draw it. Maybe even playing a reference tone at the same time so you hear if it goes out of tune.
Edit: eh…there’s something i’m not getting right here, i tried it in practice and it didn’t really change pitch when i wrote 2 cycles watching the GTune VST, instead it seemed to just add higher frequency harmonics. Maybe the waveform need to be two identical cycles for the tuner to read it as two cycles, i don’t know?
Another example of a very close to perfect pitch is, 21 samples, finetune -8 and keynote C7, it’s lo res, but it stays in pitch.
You’re probably aware of this, but what you pencil in there has to be only 1 cycle if you want the pitch to stay correct, if you add 2 peaks in both directions (2 cycles) you suddenly get a c5 instead. You can however make it double the length of 168 samples, which is 336 samples, then you need two cycles to get the standard C4 note. This way you can make more complex waveforms and it will still be in tune.
I suggest you write and play a c4 in the pattern editor and solo it while you draw, this way you can hear its changes while you draw it. Maybe even playing a reference tone at the same time so you hear if it goes out of tune.
Edit: eh…there’s something i’m not getting right here, i tried it in practice and it didn’t really change pitch when i wrote 2 cycles watching the GTune VST, instead it seemed to just add higher frequency harmonics. Maybe the waveform need to be two identical cycles for the tuner to read it as two cycles, i don’t know?
Another example of a very close to perfect pitch is, 21 samples, finetune -8 and keynote C7, it’s lo res, but it stays in pitch.
yeah no matter what I do, a new sample with the default settings plays an octave above the octave set in my keyboard. so mapping the sample to C5 does the trick for me.
Pro tip: Renoise will quite happily load waveforms with non-standard sample rates.
You can take advantage of this to avoid all the ugly shit with C @ 261.626hz not being exactly 168 samples, or A @ 440hz (which is a lovely perfect integer tuning) not fitting exactly into a sample rate of 44.1kHz, 48kHz, and so on.
We can fit the integer tuning of A 440hz exactly into a non-standard sample rate such as 44kHz or 88kHz, like so:
44000 / 440 = 100 samples
88000 / 440 = 200 samples
and so on…
No ugly fractional numbers here. No need to round anything up or down. Just lovely, lovely integers. Mmmmm!
So… Use some external wave editor such as Wavosaur or Audacity to create a new sample at 44kHz or 88kHz, then generate a single cycle sinus tone or some other waveform, and save this .WAV to load back into Renoise.
Et voila! You now have a waveform cycle that is an exact integer number of samples, perfectly tuned to the base note A (so remember to adjust the keyzones or transpose settings, kids!), ready to draw on top of and do whatever weird crap you want
For this situation where you’re drawing directly onto the waveform, it doesn’t really matter too much. But it’s always nice to have some extra resolution to work with, in case you need to apply some effects, fades, smoothing, etc.
I just know that 24 bit is preferred to 16… and it might make sense to use 32 bit since that’s what Renoise uses internally.
