Antipodes

[IanG-UK]

This isn't a plug for Antipodes, which I use (the DX), but a means of illustrating how one manufacturer tries to explain technical mysteries and their correlation to sound quality.

Background: 

The DX has been Antipodes single high end unit for some time. It is a single box combining server, renderer, power supply, storage (user installable) and ripper. Whilst maintaining the DX as a single box product, Antipodes has now also put some of these functions into separate boxes which together are about the same shape as the DX but put together are around 10% wider and 33% taller. 

So there is the CX (which can be a server and renderer but acts optimally as a server), the EX (which can be a server and renderer but acts optimally as a renderer), the ultra-slim P1 (which is the ripper) and the ultra-slim P2 (which is the reclocker and yet to be released). Storage on the CX or EX is 0TB - 8TB in 2TB steps, SSD or HDD. 

Here's what the manufacturer said about sound quality:

Audio Performance Differences

We are often asked which model sounds best and how the sound differs. We have tried here to describe the hierarchy. However in blind tests, the results are unanimous in only one thing - the superiority of the CX+EX solution. While the majority put the hierarchy for the other units as CX, DX, EX, each of them has its proponents. For example, the EX is favoured by those that like musical life and verve. The EX may be the least expensive of the trio, but its design means it does some things better than the other two. The CX is favoured by those that like refinement and fullness. The DX is favoured by those that like a sense of ease. These are not night and day differences, but you asked.

The main reasons for the sound quality differences are these. Using a single computing device to perform both server and renderer functions has economic benefits in that you only need to make one high quality device, and this will sound better than using two separate low-quality devices (one for the server and one for the renderer). But to push beyond that performance level, it becomes worthwhile to begin insulating the running of the renderer app from the server app, and this can be done in a number of ways. This concept of providing isolation to the renderer function drives the differences in the audio performance levels of the models. The size of the performance level differences is arguably small, but in higher resolving systems the differences become very obvious and significant. A deliberate benefit of our approach is that you can start with say an EX or CX and upgrade later to the CX+EX solution without needing to sell anything.

• The EX is designed to the same specification level as the rest of the X-Series except that it provides the least separation between the running of the server app and the renderer app.
  
  
• Note A.1 - The DX improves on the EX because it provides separation of the power supplies running the two functions, but it uses the older ODAPS1 power supply technology.
  
  
• The CX improves on the DX and EX by providing a lot more server app headroom with its high power V4H platform. But note that it does not use the power supply separation used in the DX.
  
  
• The CX+EX solution is a dramatic improvement on the other models because it completely separates the running of the server app and the renderer app. In this context, because it is running just the renderer app, the EX audio performance is unhindered and there is no need to use a CX or DX instead of the EX. In fact the EX has the best renderer performance of the three. So using the CX as the server and the EX as the renderer is the best sounding solution (needs a good quality short Ethernet cable to connect them).
  

Even I can understand it.