I need some help understanding pitching (warning: probably a dumb question)

[jadeandtea]

Beginner here,

Let’s say I have this sample of a bass guitar playing a short riff. From that I can chop it up into individual notes (say, F, D#, what have you) and use those as a makeshift instrument. Alternatively, I could only isolate one of the notes (say F), make it into a new sample and then change the pitch to whatever I want (like D#). What is the functional difference between the two? Please let me know if I’m being unclear.

[kingslim]

The more snapshots you have of an instrument, generally the more real it'll sound (if they're all mapped/scripted/aligned correctly) as an "instrument". There is the pitch axis, so the less you've got to stretch each sample to be the correct pitch the better 1 pitch being the lowest resolution... There is the dynamics axis, having different samples for when the key is hit quiet or loud (either just the same regardless, up to 127 different samples - 1 for each Midi velocity, ignoring 0)... You also benefit from having multiple versions of each of those slices to allow for natural variations in timbre to be captured and then triggered randomly or sequentially to emulate that... Then all that again with key switch chosen articulation options etc... so yeah, 2 notes a bit better than 1, but not much.

Edit...

The trade off being each layer of complexity adds a lot more samples (and therefore prep work), easily becoming thousands of samples for a single bass guitar, rather than the single bass hit. Those samples take up space on hard drive and use up ram resources when the patch is active.

There's lots of people that do all this for you and sell libraries so you can just buy various basses and start hitting notes...

PS. I assumed you were using a sampler rather than just a daw, as with samplers you can map samples to key zones etc and play them (and you were asking about pitch, but now I'm not sure, lol), there'll be one in your DAW, or you can buy third party ones like Native Instruments Kontakt or UVI Falcon etc.

[jadeandtea]

Thank you, that was really helpful! I’m using Fl, so I’ll open either Edison or some other slicing tool then I map each ”slice” to a key on my keyboard. The problem is that most of the time I don’t have a long enough sample to build a full sampled instrument so I have to use individual hits more than once by changing the pitch to allow me to play all of the keys. What do you use as a sampler?

[jancivil]

Regarding transposing pitch by stretching, there will be a point where you've altered the tone of the sample. It's called chipmunk effect; it's like playing a record at the wrong speed. One step is probably not going to make a world of difference.

[enroe]

Bass guitar playing a short riff:

(1) From that I can chop it up into individual notes (say, F, D#, what have you) and use those as a makeshift instrument.

(2) Alternatively, I could only isolate one of the notes (say F), make it into a new sample and then change the pitch to whatever I want (like D#).

What is the functional difference between the two? Please let me know if I’m being unclear.

In case of (2) you have one single note (say F) and you would transpose it to all
the other keys. For the keys transposed only a few semitones it might sound good.
So D#, E, F, F# and G will probably sound ok. But more than a few semitones pitch-
transpose will very likely sound odd and weird!

To overcome these bad sounds you will choose method (1): You have sample-notes
for the most important notes in a distance of 3 semitones: C, D#, F, G#. And so you
can play every note of an octave without having a transpose of more than 3 or
2 semitones.

Method (1) is the creation of "multi-samples" for an instrument.

[enroe]

What do you use as a sampler?

OOooh, what a question. There are a lot of free samplers - and there
are some famous paid samplers.

I recommend to do a Google-search and find out what is best for you.

[jadeandtea]

To overcome these bad sounds you will choose method (1): You have sample-notes
for the most important notes in a distance of 3 semitones: C, D#, F, G#. And so you
can play every note of an octave without having a transpose of more than 3 or
2 semitones.

That's a really good suggestion. Thanks!

[Benedict]

It is about harmonics. Every sound has a Fundamental and then Harmonics on top. The fundamental sets the main Pitch and the harmonics the Timbre.

When you transpose the sample, the harmonics get transposed as well. The problem is that the harmonics on a low E are not exactly the same as the string up, A. Definitely not the same as E Sharp (or F) as you have your fatty, oily, fingers all over the string as you mash it into the fret.

So as said accurately, you can kinda get away with transposing a semitone or two up/down and the bain won't be too weirded out, but as soon as the harmonic relationship becomes too unrealistic, the brain screams: FAKE!

Then to make it even harder, every time you play that open E string, the harmonics are different. So if you have one sample and play it over and over, the brain also starts to scream (or get bored and go to sleep seeing we are wired to spot wabbits when they hop). This leads to the need for Round Robin sampling where the same note delivers one of many possible samples.

[jancivil]

Just don't transpose the open strings, this difference is large.
D# - F was brought in, no big deal. If the goal is to eventually be convincing as a bass guitar and not a synth, there are many things to know about it. One F for example may be found on string 2 fret 3; string 3 fret 8; string 4 fret 13 and the thickness of the string produces a different tone.