So, two questions:
How does the ācascadeā improve the network?
What if I donāt hear any difference?
A network is in part a series of hardware modules and firmware that convert the data stream into various physical forms in order to facilitate its movement from one location to another. Networks are built on various sets of standards, so called layers.
Thereās the application layer, presentation layer, session layer, transport layer, network layer, data link layer and the physical layer. To be transported, a data file must be transformed into electromagnetic signals. The specifications that govern these transformations are encompassed in the physical layer. The Physical layer is the first layer in the OSI (Open Systems Interconnection) model, which is a conceptual framework that standardises the functions of a communication system. The Physical layer deals with the physical transmission of data bits over a communication channel and defines the electrical, mechanical, and procedural aspects of transmitting raw data between network devices. It primarily focuses on converting digital data into a format suitable for transmission and ensuring reliable and efficient communication at the physical level.The physical layer is basically the underlying physical structure of the bit stream. Itās the voltages, the optical pulse forms, the radio wave frequencies and power, timing, noise levels etc. etc. What I found is that the more perfect (closer to the ideal spec) the physical layer that reaches the DAC is, the better the resulting music sounds. From an IT perspective, as long as the physical layer meets a minimum requirement for data transmission and integrity, everything is fine. Its only when that bitstream is converted to an analog music signal, converted to sound pressure waves by speakers, converted to fluid pressure waves and nerve impulses by the ears and interpreted by the brain that the absolute quality of the physical layer becomes pertinent and for an audiophile, important.
So how can the physical layer be improved? Each module of the network handles and/or transforms the data stream and each module performs to a certain physical layer quality level; its noise level, degree of filtering, its ability to generate āperfectā voltage polarity switching, its ability to accurately time bits and transmissions, the amount of EMI/RFI it produces or picks up, the degree to which vibration and resonance affect its output, its ability to isolate from external noise etc. etc. So each module on the internet produces an output with a certain physical performance spec. Typically āaudiophileā devices designed specifically for audio will produce an output with a far more refined physical specification. For example, a regular IT consumer-grade SMPS will not pay much attention to noise or the ability to produce an accurate square wave polarity switch, beyond complying with the minimum physical layer requirements, whereas a top grade LPS is designed to produce DC with exactly those abilities/refinements. Similarly an audiophile switch will isolate, filter and accurately time the bit stream to a much higher degree than the ādesigned to be as good as necessary and as cheap as possibleā consumer grade switch. So each module on the network is producing an output with a certain physical specification. The higher quality the module the better its output. But the network comprises a series of modules, so if the LPS on your modem produces a super-clean output that then encounters a noisy SMPS on your bridge, the LPS will have little effect, as the noise it removed is re-added at the bridge. It is therefore critically important that the bitstream encounters increasingly better physical specs as it travels along a network. Again as an example, thereās no point producing a bit stream from a switch with 10ppm accurate timing if you then send it into a server with 100ppm timing accuracy. The bit stream is only as good as the last module on the network.
The other discovery I made is that better in = better out, so the better the input to ANY module on the network, the better its output. If you arrange your network as a series of physical layer improvements, the output from say your router will be improved at every following module e.g an improvement at the modem will compound as it travels along the network, producing an improved input and output at all following modules.
So hereās the thing. If you make an āimprovementā to something on your network and you hear no difference, it either wasnāt an improvement ie didnāt produce an output with a more accurate/superior physical layer or it was in the wrong place on your network i.e was followed by a module with inferior physical layer specification that ācancelledā the improvement you made. For an improvement to be ājaw droppingā, its has to result in a bit stream with a considerably more refined physical layer reaching the DAC.
I hope the above clearly answers the 2 questions.