how to add release sound to sample/patch ?

24 Aug 2004, 06:48

I have had my Fantom XR for a few weeks and am looking into creating my own patch & multisample. My first one will be of my acoustic grand at home. Second one will be my teacher's harpsichord.

I haven't done any real work with the FXR yet, but I have been busy reading all about patches parameters and samples before. Fortunately as a software engineer this stuff is all pretty readable for me

But I still have a question : how do you incorporate release sound into a patch, ie. after the note has been released ? For example, the sound of dampers on a piano, or the sound of the keys being released on a harpsichord. This is a one-time (non-looping) waveform, but at the end, rather than the beginning.

From what I read in the sampling section, each sample has an attack section - defined between start point and start looping point, and then a looping section ... But the loop is all the way from "start loop" point to the end of the sample (end point) !!! Well, this isn't how it works on a harpsichord, or on many other keyboards, I suspect. So how do I add the key-click off from the harpsichord at the end ? I would have no problem recording the sound ... But do I need to put it as a separate waveform ? And how to define it in the patch ?

Thanks !

Re: how to add release sound to sample/patch ?

24 Aug 2004, 14:25

Hi, and welcome.

Yes, you'd need to record the key release sound separately as a separate waveform. You'd then assign that waveform to one of the 4 Tones in the Patch. You'd then set the Tone Delay Mode (in the Misc parameter category) to KEY-OFF-NORMAL. This will delay activation of the Tone until the key is released. Once the key is released, the envelopes will begin as if you had just pressed and immediately released the key. The harpsichord preset patch on the Fantom-X does this for the key release sound. You may want to examine how that patch is programmed to get an idea of how to do it.

Regards,

Nathan Sheldon
http://www.nathansheldon.com/

Regards,

Nathan Sheldon
http://www.nathansheldon.com/

thanks !

24 Aug 2004, 22:07

I guess I missed that part with the tone delay mode to KEY-OFF-NORMAL . Sounds like just what's needed here. I think I'm going to start my sampling soon.

More questions while I have an expert responding :

I read in the manual that the sampler can do "auto-divide" based on silence when sampling, to separate a recording into individual samples for each keys. Can this also be done on an imported WAV or AIFF file from the memory card or USB ?

I want to do my sampling on the VS-2400 which I'm sure has higher quality ADCs than the FXR. I would then transfer samples to WAV files on CD-RW, then to the FXR through the USB port (copying files from CD-RW drive letter to FXR drive letter).

It's unclear to me how the mic pre-amp works in the FXR, if there is one at all. I have 2 AKG C3000B condenser mics which requires phantom power. The FXR manual seems to say you can record from only one mic at a time (ie. mono) ; stereo is from line level only and would require external pre-amp . I sold my M-audio DMP3 mic preamp when I bought my VS-2400, which has 8 channel preamp.

Are the samples on the FXR limited to 16 bit / 44.1 kHz ? The VS-2400 can record in 24 / 96. Obviously that would take a lot of space, but I think the max RAM in the FXR is big enough

If the individual samples are 16/44 it's probably good enough quality though. And the VS-2400 can record down to that low.

Re: how to add release sound to sample/patch ?

24 Aug 2004, 23:34

Yes, the Auto-Chop feature can be applied to imported audio.

The Fantom-XR supports only 44.1 kHz sample rate audio at 16-bit resolution in either mono or stereo. However, I have successfully imported AIFF audio files of every combination of 8-bit, 16-bit, and 24-bit resolution (clipping occurs with 24-bit data) and sample rates of 8 kHz, 11.025 kHz, 11.127 kHz, 22.05 kHz, 22.254 kHz, 32 kHz, 44.1 kHz, and 48 kHz, and in mono and stereo combinations. The Fantom-X converted each one to 44.1 kHz 16-bit data (mono or stereo). The internal conversion of sample rate and bit depth is not perfect though. I'd strongly suggest an external source for sample rate conversion and bit depth conversion. This ability is convenient for low-fi effects though (such as 8-bit or 8 kHz audio files).

Regards,

Nathan Sheldon
http://www.nathansheldon.com/

Regards,

Nathan Sheldon
http://www.nathansheldon.com/

sample conversion

28 Aug 2004, 07:04

It's the week-end ! I finally got to try it rather than just talk about it.
I put in some 24 bit / 44.1 kHz WAV into the Fantom XR.
Used auto-chop ...

And got.. A horrible amount of hiss ! You could barely hear that it was piano samples ...

I was starting to get scared because of the other thread about the RAM making noise during sampling, etc ... !

Then I imported one of my full recordings as a 16 bit / 44.1 khZ WAV file, and it played just fine in the FXR, without hiss.

I'm converting my piano samples in my VS-2400 from 24 to 16 bits as we speak. Just export them as WAV to a CD-RW, and then from there to the computer and over to the FXR over USB (I sure wish the two devices could connect directly ...) ...

Glad to report that everything sounds just fine if I choose 16 bits as the source ! I guess I will be making all my new samples in 16 bits / 44.1 kHz - no sense having an extra conversion step. My original decision to use 24 bit sampling on the VS was in case I added effects after sampling on the 2400, to prevent loss at each step. Seems the FXR also has plenty of effects of its own.

The auto chop function worked great on my imported 16-bit piano notes. I only had the C for each octave, which I have been told is not enough for a realistic piano sound. Seems my piano is amazingly in tune - when I press notes on my keyboard to sound the FXR's ultimategrand, and then press preview on the rack to listen to my sample at the same time, the unison is pleasing, not dissonant at all . Maybe I just can't hear how much off it is.

The thing is that I'm not quite sure how long I should hold each note. They aren't exactly of the same length, and certainly not of the same pressure (at least I won't have that problem on the harpsichord next!). Any advice on this, Nathan ?

I'm guessing the duration is probably not a big issue, I just have to find the right attack part and loop points for all notes.

The patch creation is sure going to be an interesting process ... Looking forward to it the rest of the week-end.

Re: how to add release sound to sample/patch ?

28 Aug 2004, 15:26

Hi madbrain.

My tests with various audio format has resulted in what you discovered with regard to 24-bit audio. It sounds as though the Fantom-X is simply dropping the upper 8 bits of data (truncating it from 24-bits to 16-bits) on playback. This seems to explain the noise. I've also found that the Fantom-X can import audio of different sample rates all the way from 8 kHz to 48 kHz, though the sample rate conversion used is not of the highest quality. It can also import lower bit depth audio such as 8-bit if lo-fi is something you are looking for in the future.

As for the sampling of the piano, the longer the samples, the more realistic the sound will be. Yes, samples at 1 octave intervals is generally not sufficient, especially for acoustic instruments. Every 2 whole steps or around that interval would probably be sufficient. Someone with more acoustic sample creation experience would probably be able to provide more detail on this though. As for the force with which you hit each key to be sampled, that is an issue as the harmonics of the instrument change as the force you play it with changes. Roland uses a specially designed machine that hits each key with the exact same velocity each time, but that's obviously not something you can use yourself, unless you devise one of your own. Ha.

Regards,

Nathan Sheldon
http://www.nathansheldon.com/

Regards,

Nathan Sheldon
http://www.nathansheldon.com/