I realize it's been quite a while since I've been particularly active musically. In my defense, it probably is for the better that I not be overly obsessed with getting compensated for my efforts, but the reasons for that decision of mine are best saved for elsewhere.
In the meantime, I think it's safe to assume I've nearly perfected a lo-fi sound without the excess aliasing or using too much CPU. Using a mock-oversampling formula and combining it with some light real oversampling, I managed to make a nearly perfect recreation of variable sample rate clocks heard in early samplers such as the Fairlights. As you'll see in the below pictures, which make use of a sine wave sampled at 880 Hz and pitched down, the aliasing harmonics are relatively in control within the second image: all obvious aliasing (which is still under -70dB) within that particular image is still above 11 KHz, meaning 2x oversampling would effectively resolve that much. Observe for yourself.
1. Two octaves down
In the meantime, I think it's safe to assume I've nearly perfected a lo-fi sound without the excess aliasing or using too much CPU. Using a mock-oversampling formula and combining it with some light real oversampling, I managed to make a nearly perfect recreation of variable sample rate clocks heard in early samplers such as the Fairlights. As you'll see in the below pictures, which make use of a sine wave sampled at 880 Hz and pitched down, the aliasing harmonics are relatively in control within the second image: all obvious aliasing (which is still under -70dB) within that particular image is still above 11 KHz, meaning 2x oversampling would effectively resolve that much. Observe for yourself.
1. Two octaves down
2. An octave and three whole steps down, a note difference prone to aliasing in sample playback
I managed to pull off similar results using the generators for Plogue's chipsounds (which, unfortunately, require the source samples be a mere 8-bit) and, 8-bit noise aside, I achieved very similar results. Hopefully my process will be deemed more convenient for people who wish to play back their own samples, though, because aside from 8-bit quality, chipsounds also limits many of its sound editing features to text files over a fully editable GUI.
My only real problem with Reaktor now is its relative lack of support for saving presets that contain samples. As of Reaktor 6 (what I used to make my ensemble), it's possible to store single samples transposed across the keyboard within snapshots of a single ensemble, but the problem is that this obviously does not account for multisampling, which can easily render drum sample snapshots, for instance, extremely inconvenient if not impossible to make. Thankfully, that isn't the biggest deal for my particular ensemble, since it is intended to be reminiscent of samplers that tended to have only one sample across the keyboard range anyway.
For any Reaktor users, musicians, and sound designers who enjoyed this read, I should be able to share the rework of my ensemble to the general public soon enough. Stay tuned!
My only real problem with Reaktor now is its relative lack of support for saving presets that contain samples. As of Reaktor 6 (what I used to make my ensemble), it's possible to store single samples transposed across the keyboard within snapshots of a single ensemble, but the problem is that this obviously does not account for multisampling, which can easily render drum sample snapshots, for instance, extremely inconvenient if not impossible to make. Thankfully, that isn't the biggest deal for my particular ensemble, since it is intended to be reminiscent of samplers that tended to have only one sample across the keyboard range anyway.
For any Reaktor users, musicians, and sound designers who enjoyed this read, I should be able to share the rework of my ensemble to the general public soon enough. Stay tuned!
RSS Feed