Nord User Forum

Hey guys,
what is your standard way of working when sampling a new instrument? How perfect does the recording need to be and how much of the imperfectness of the recorded instrument do you leave in there?
Let me take my Solina recordings as an example. (I know I promised 2 years ago to post all the voices for the String Ensemble, but life hit me and here we are...) If finished a few of the voices and I notice a difference in volume per voice. So the violin is louder than the horns. I did not change any of the settings on the Solina while recording. So it must be that the horns are quieter than the violin. But maybe I want all the voices to be at the same level. Then I can adjust them later in the Nord itself. I noticed the horns were louder on the right channel. But the non-modulated sound is very mono. So now I exported the original wavs in mono. Just to have the sound perfect balanced.

Do you normalize your WAV recordings before you import them in the Nord sample editor? Just normalize couldn't hurt right? I don't mean any compression or limiting. Just gettin the file as loud as it can be in digital. Or does the overal level adjustment in the sample editor all I need for normalizing?
Do you normalize the separate note samples in the editor? Or do you keep the original differences in level because that's how the instrument sounded at the moment of recording?

I would love to hear your thoughts on this process. So I can make up for the lost years and contribute some more. I have a few instruments on the way. Solina String Ensemble, Siel Orchestra and a Gem PK4900.

I havent made many programs with the Nord sample editor to be honest, but I have had lots of experience with sampling of natural instruments and drums generally for Akai samplers.
One brass library I created and published as VSt/AU plugin in the past https://amg-software.com/amgkickassbrass/kickassbrass.html .... Showing my age.

Normalizing before importing (or inside the sample editor) should work fine, but bear in mind that if you normalize each sample individually, some samples might end up sounding unnaturally louder than others, so some gain adjustments might be required on each sample keygroups afterwards - just using your ears to make the program sound natural when switching to the next keygroup. Normalizing pushes the sample to max volume without paying any attention to the 'averge perceived volume' of the note over it's duration. So a sample with a fierce attack spike might end up sounding quiter that other samples (with less of an attack spike) Ideally you would normalise the whole recording, before cutting/splicing the recording into separate samples, depends if you made 1 contiguous recording of all notes, or lots of small recordings of each sample/note. Another option is to figure out what the max DB boost would be without clipping one of the samples, and boost all samples by exactly the same DB amount, therefore maintaining the natural relative volume between the samples.

If you have a decent compressor to use, then I'd say its not a bad idea... But a cheap compressor or plugin might do more harm than good, especially if its overused.

Hi Spookycookie, thanx for reply. Sorry it took me such a long time to react. I think you are right not to normalize every separate note. But instead normalize the whole wav recording. This should give the sample a more natural sounding response. Right now I am moving my instruments from one corner in the room to the other corner. When it’s all connected I am hoping to get back to sample editing.