• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Sony’s DSD Remastering compared to foobar2000’s DSD Processor


Senior Member
Jul 18, 2020
Adelaide, South Australia, Australia
Perhaps it has been asked before in one of my DSD threads. But the essence of the question is this: when converting PCM to DSD or DSD to PCM ‘on the fly’ or for real-time playback, does a more CPU intensive process create a more accurate file and does that extra accuracy result in any audible or measurable difference?

Recently, I once again came upon solderdude’s incredible foobar2000 two part guide (Part 2 concerns DSD output), and found that foobar2000 had changed since I last viewed the guide two years ago. For example WASAPI output is now included with foobar2000 out of the box. And it is now possible to convert any audio source into native DSD256 by turning the DoP output of the DSD Processor back into native DSD using a Transcoder.

In 2007 the Sony VAIO DSD Direct Player software was released for VAIO PCs and it was based on the VAIO DSD Direct converter which had already been optimised for intel dual core processing. The minimum spec to run the software was 2 GHz and CPU utilisation was 100% at 2 GHz. DSD Direct Player also had a great GUI by 2007 standards anyway.

At some point Sony then released ‘DSD Remastering’ which used FPGA hardware and now in 2022 their flagship walkmans running Android have ‘DSD Remastering’. Unlike the DSD Direct converter, this Remastering is less about staying true to the original PCM file and more about manufacturing that ‘DSD sound’ existing in most people’s imagination since high-end equipment is out of reach of most consumers targeted by Sony.

Other than proprietary real-time conversion from the likes of JRiver and HQPlayer, there are numerous free DSD related converters. The way I see it, each converter has its own characteristics, and these differences far outweigh any CPU utilisation differences. Depending on the filter selected, HQPlayer has varying levels of CPU usage. Some converters allow the user to select the number of taps of the FIR filter, a practice influenced by VAIO DSD Direct and SBM Direct which used the number of FIR filter taps as a publicising tool. When it comes to real-time playback as opposed to conversion, foobar2000’s DSD Processor is probably sufficient. It seems a waste to use 100% of 2 GHz if what you hear ends up the same. But if it is a file conversion then FIR filter taps might become more relevant.

Many thanks to solderdude for the outstanding guide to foobar2000’s DSD, because why not playback video files, opus files and HE AAC v 2 as DSD256 without increasing much CPU load? JRiver doesn’t recognise some opus files while HQPlayer often finds the ASIO device busy (bars keep sliding from right to left and left to right) or doesn’t output sound and in any case can’t load HE AAC v 2 files and other compressed sources.
Top Bottom