Stereo or mono files

Question

Hi ! I am about to make my first vitual instrument using Decent Sampler, and I am asking myself about recording mono or stereo for the audio files I will use. What seems to be the best file format for you ? Perhaps I could record stereo for mecanicals instruments like pianos and stay mono for mono synth-based sounds ? And by the way, does Decent Sampler accept both mono and stereo ? Thank you so much.

Vincent

Answer 1

Decent Sampler can accept either mono or stereo, so it's really your call.

Comments

Thank you so much ! And could you tell me if there are any audio format recomandations ? 24 / 16 bits ? 44100 / 48000 ? Seems that I have an issue with my first try : the sound played by the sampler is longer than the original sound ...

---

Hello Vincent,
Just my 2 cents,
To me, higher bit depth like 24 bits and higher sampling rates like 48Khz are really good for audio processing or impulse responses as it gives you a bit of headroom. But for instrument samples you can usually "get away" with good old 16 bits 44100Khz. Considering the amount of processing they will receive, it won't make much difference.
If you are concerned about file size and want to reduce instrument size without reducing audio quality use FLAC, I do this exclusively and it works great!
However, stereo samples make a huge difference, so I’d say use stereo as much as you can.
I do this even for instruments seemingly mono in nature (like electric basses and guitars) as they are often amplified and heard through both our ears. If you don’t do this in the samples, eventually you will have to deal with stereo imaging later with mixing and effects...

---

Thank you Michel (by the way are you French like me too ?)
I made my firsts samples the best I could (Stereo 48000 Khz / 24 bits) and it works fine for me. But I keep in mind all your advice.
Now I am struggling with the loop point. I think I will ask a new question about it in this Q&A forum.
Thank you again !

---

Salut Vincent,
Oui, je suis effectivement français :)
Pour les points de boucle j'ai personnellement conçu un outil en python pour les détecter en automatique, ainsi que bien d'autres outils comme un générateur de presets DecentSampler qui crée automatiquement un mapping et une UI. Je compte mettre tout ça en ligne gratuitement à un moment donné sur mon github.
Sinon pour les points de boucle, la méthode que j’utilise est de me caler sur les "zero crossings" en testant la corrélation entre toutes les combinaisons de point de départ et de point de fin.
On peut faire ça à l'oeil en cherchant 2 passages qui se ressemblent et en utilisant les zero crossings comme marqueurs (je ne suis pas sûr d’être clair). Ensuite pour lisser le résultat, on ajoute un petit peu de crossfade car à part pour des sons synthétiques ou déjà traités, un point de boucle parfait n'existe pas.
Bonne soirée.

---

Nos échanges se croisent d’un post à l’autre ! Merci pour ces informations. N’hésite pas a me partager ton github pour que je suive ça. Je ne suis pas du tout assez calé en programmation pour développer mon propre outil  alors ton projet sera bienvenu !