Do not hesitate to share some more infos about Digibooster as many people use it around here. 
Well, DigiBooster 3 is developer for 4 years already, but the final release is expected this year. The program is written from scratch in C (DigiBooster Pro 2.x has been coded in m68k asm, making it unportable to "nextgen" Amiga[like] systems). It uses MUI as its GUI engine with addition of skinning feature (a skin or theme consists of MUI preset, program own config, bitmaps and fonts, it may be switched automatically while running the program).
The main advantage of the new GUI is scalability, it is no more fixed at 640 pixels. Then users may take advantage of nowadays big LCD displays and have 20 or even more tracks displayed (depends on selected font). On the other hand user is not forced to use this famous 6-pixel font in pattern editor, any font may be selected via MUI preferences, including antialiased ones, where available.
Another big thing is the new player engine. It is written from scratch too. It is compatible with DigiBooster Pro 2, but there are some fixes improvements and changes. The player does not depend on AHI anymore and is written in portable ANSI C. It means we will have 100% verbatim replayers for Windows and Linux for example. Portamento step has been tied to music scale now - it is always 1/8 of semitone now, regardless of instrument base frequency and mixing frequency. All portamento effects are then transposable. All "fine" commands (using xF or Fx parameter, like fine volume slide AFx/AxF) have been made really fine. What it means? Let's assume standard $06 speed. Typical A01 slide will change the volume by 6 units per row (one unit at every tick). Fine AF1 slide in DigiBooster 2 changes the volume by one unit, but it is done once. DigiBooster 3 is able to do fractional volumes and changes the volume by 1/6 of unit every tick, so the final effect is the same as in DBPro 2, but much smoother. This applies to fine portamentos too. Fractional volumes and pitches automatically follow speed changes made by Fxx, so after F0C for example, AF1 changes volume by 1/12 of volume unit every tick.
The player is highly modular - it allows for different types of instruments (multisample ones, or realtime synthesis), different types of instrument effects, track effects and postprocessing effects. They will be added in following updates of the program.
As the player engine does mixing and resampling on its own, the quality of sound is no more degraded by AHI. For now simple, linear interpolation is used for resampling, but optional bandlimited sinc() one will be added. DigiBooster3 can work with mixing frequencies from 8 to 192 kHz including. Mixing and volume-boost is lossless (uses 32-bit arithmetic) and possibility to render the module in 24/32 bits will be provided. The player will also do its own DSP-echo (so it is better kept turned off in AHI). Individual echo settings for every track will be possible as a bonus.
As you can guess, all this comes at a price. DigiBooster 3 has higher hardware requirements compared to DBPro 2. It is generally aimed at "nextgen" Amiga PowerPC based systems, or classic Amigas emulated with WinUAE. While one can run the program even on stock A1200 extended with 4 MB of fast memory, for real use 68060 @ 50 MHz and a graphics board are required. Tests on Amigas have proven, that even 68040 simply does not make it. On the other hand even the weakest "nextgen" system, which is Efika 5200B, will allow for normal work with 64 tracks. Any PowerPC G4 machine will be more than enough, as I'm an experienced AltiVec programmer and will put it in a good use. WinUAE on any nowadays moderate class PC will run the thing nicely as well.
Feel free to ask more questions :-).
