May 2, 2016
You may have noticed many software products and plug-ins notating 32-bit or 64-bit features. This article is aimed to clear up any potential confusion regarding these terms.
First, please note that these terms can be used to separately describe many different parts of the software and audio world. For example, operating systems can be based off of 32-bit or 64-bit architectures, while separately, audio host applications can also be based off of 32-bit or 64-bit architectures. Further, plug-ins that are hosted within these applications can also be based off of 32-bit or 64-bit architectures.
These are all based on the software side, and it would be important to note that audio bit-depths such as 16-bit, 24-bit, and 32-bit floating point, are separate notations that indicate bit depth for audio quality, not software architecture.
You may have also heard of some audio processing plug-ins that process at 64-bits. This is noting that while the plug-in may be loaded as a 32-bit plug-in in a 32-bit host, the internal math that is processing the audio is working at the 64-bit level. Recently, 64-bit native plug-ins have become more common, and these 64-bit plug-ins will also commonly process audio with 64-bit math.
This can also be confusing if you are running on a PC that has two separate folders for VST plug-ins, 32-bit (within the x86 directory), or 64-bit (in the directory without the x86 indicator). Some host applications may only be able to run 32-bit plug-ins while running as a 32-bit application, or alternatively 64-bit plug-ins only when run as a 64-bit application.
The benefit to running 64-bit software, is the access to greater amounts of memory (as the limit to 32-bit software is approximately 2-4 gigabytes). If you are running very large sessions, running a 64-bit host application with 64-bit plug-ins can allow this session to run without experiencing any performance issues (or potential crashes).
Processing audio that has higher bit-depths (24-bit, 32-bit, etc.), allows for greater accuracy of the processing directly. One of our products, RX 4, will automatically convert your audio to 32-bit Floating Point upon clicking 'Process' in one of the available modules. This will make sure that the audio process is applied at this higher bit depth for best results. Once you are done with this audio file, you may then want to save it as a 16-bit or 24-bit file. In this scenario, you may want to consider applying 'dither'. For more information on dithering, please check out our Dithering Guide.