To capture the piano, start a recording into your chosen device, for best results we would suggest recording at 48khz 32bit(float), if you cannot go this high then 44.1khz 24bit will suffice. Don't record any higher level wise past -6db, or lower than -18db. Capture each note for long enough as to include its tail (attentuation) to -60db. The amount of time you will need to play each note will decrease as you move up the keyboard because the higher notes tail off quicker. Finally, be sure to capture the piano with the sustain pedal down, ensure you engage the pedal before striking the note and then carefully lift the pedal after you have released the note.

 

When you have completed a first cycle of recordings, play the whole cycle of fifths again but at a louder velocity. You may need to experiment to find the optimal “quiet” and “loud” versions that capture the sound you like the best. Remember to try to keep the volume the same for all samples in a particular layer.

 

Following the recording of the “quiet” and “loud”, capture the release triggers of each note in the cycle of fifths, which can be used to add realism to the virtual version of your piano. To capture a release trigger, press down the key (without the sustain pedal) but don’t let the hammers strike the strings. Then simply release the key – slightly more exaggerated than you might normally (it’s easier to turn down the sample than turn it up).

 

Finally, capture at least 30 seconds of “room tone”, sitting in silence to capture the rooms noise, which will be helpful when processing through noise reduction at a later stage.

 

You may want to repeat the whole process multiple times so that you can pick your best take or even mix and match across takes if one note doesn’t quite work from one of your takes, the sample the better your choice of 'Round Robins'

 

Noise Reduction

After recording your samples now would be a good time to noise reduce your recordings before editing, you can do this after editing to as a batch process if you have the correct software such as Izotope RX, we like to do it before editing (chopping) but the outcome is the same, to noise reduce you will be using your 'room tone'. See our 'Noise Reduction' Video HERE (4:04mins). If you still need help after watching our video then please reach out to the community.

 

Editing the Samples

After recording your samples it’s time to cut them up so they can be loaded into a sampler. Most DAWs have a function to cut an audio file at transients and this is the easiest way to slice up your recording into the individual notes. Check out our quick video demo of this HERE, we have used Logic Pro but it would be the same in most DAW's.

 

When naming your samples, make sure you include the name of the note (using sharp ‘#’ signs rather than flat ‘b’ signs) to make it possible to auto-map your samples into a sampler (most don't recognise the 'b' flat extension. Also include the velocity layer in the name or “RT” for release tiggers. If possible, files should be saved at 48khz, 24bit, wav. A full recorded set of samples will look similar to the following: