I have a sample with 3000 sample lenght. Inside it, I have 3 samples, they start at 0, 1000 and 2000 respectively. Of course, each sample duration is 1000 (1000 + 1000 + 1000 = 3000).
I added this sample wav file (which itself already contains all 3 samples) to Polyphone, then I added it to an instrument 3 times. Then on the first "sample" inside the instrument I configured "Sample start offset" as "0" and "Sample end offset" as "-2000" (it only accepts negative numbers so I assume it counts from the end). So when I play the note of this sample it starts playing correctly at "0" but it plays the entire wav file, it does not stop at "-2000".
sample start / end setting is a good example of an unnessesary parameter.
it is nonsense to keep multiple sound shots in one .wav file; if you want to process one shot, you will have to adjust all settings for the other sound shots.
split that .wav sample in three .wav samples and you will be happy for the rest of your life.
regards bottrop
This time I will have to disagree with you @bottrop. I created a script that samples a VST instrument and every 5 seconds it sends a MIDI command to the VST to play a note an during all that time I am recording. At the end, I have a unique long wav file in such a way that every 5 seconds there is a note.
I can simply import it in Polyphone and with Excel I can automatically set the START/END for each note in 5 seconds interval (converted to sample lenght). That's it. I dont need to import my wav file in an audio editor, I dont need to split it in hundreds of files, I dont need to later import those hundred of files in Polyphone, I dont need to manually set each note root key (or pray that Polyphone guesses it correctly)...
I assure you bottrop: in this case it's much much much easier to do all of those with a single wav file containing all the samples. I just need to make SAMPLE END OFFSET to work, it seems to me that it has a bug.
i think you do better go into VST creation. the soundfont protocol only loads the samples that are actually played into memory. that is why .sf2 is so agile. if you put hundred shots in one .wav file, hundred shots are loaded while only one is playing.
regards
I was not aware of that. I didnt think sf2 would load each long wav file for each note! I thought it would internnally load just the part (between START and END) for each note. If what you are saying is true, it's certainly a terrible idea doing what I am doing! I was thinking about loading this same long file for each note and then just assigning a portion of it to each note... but if sf2 is gonna load it all for all notes, that's gonna suck! Are you sure about what you are saying? If that is the case, I will abandon my usecase and do as you said: one specific file per note.
it is not only so for soundfonts; every app that wants to pick something from a file needs to open the complete file to find it. ask your computer.
regards bottrop
