I've noticed a price disparity between the Native and Powercore plug-in offerings from Sonnox. Native versions (Audio Units in my case) are much cheaper than the Powercore versions, but do they differ in quality? If so, are these differences typical of the plug–in market at large? I don't use Pro Tools but have noticed similar price differences between RTAS and TDM plug–ins. Are customers here paying for the privilege of offloading processor burdens to DSP cards, or are the plug–ins better somehow? Is there, perhaps, different algorithmic encoding involved)? I've been running native plug-ins for a while with this on my mind, and I'm hoping you might clear it up for me.
Will Wright
Features Editor Sam Inglis replies: There are several reasons why manufacturers often charge a premium for the versions of their plug–ins that run on DSP platforms such as Powercore or Pro Tools HD. Occasionally, those versions have extra functionality or superior sound quality compared with the native versions, but this is increasingly rare. However, even when the functionality of native and DSP versions is identical, there is a considerable amount of extra work involved in creating the DSP version.
TC Electronic's Powercore (left) and the Digidesign HD system (right) both use fixed–point processing to run plug–in algorithms, whereas so–called 'native' plug–in platforms, such as Audio Units, VST and RTAS, operate using floating–point mathematics. Plug–ins written for native systems therefore have to be re–coded to be able to operate on specific platforms.With a native plug–in, the core code (usually in C++) can be re–used across different plug–in formats and even ported from Mac to PC without having to be radically rewritten. Coding for hardware DSP platforms, by contrast, is completely different, and requires that the "same" plug–in be rewritten from a very low level. For instance, native plug–in formats all use floating–point mathematics, but the Motorola DSP chips used in Powercore and Pro Tools HD are fixed–point units. Another major difference is that computer operating systems can allocate RAM and CPU cycles dynamically, whereas DSP–based plug–ins have to declare the amount of processing power and memory they need at load time, and ensure that they never exceed these limits. It is this that, as you suggest, which increases the price of platform–specific plug–ins.