Perpetual Technologies P-1A D/D & P-3A D/A processors Page 2
The operating instructions are in the process of being revised, and I must say they need it. I have some experience with the type of operational logic that Audio Alchemy used, and it's similar to that of the Perpetual Technologies products; even so, a couple of times I ended up selecting non-optimal sample rates and/or bit densities. It was the drop in sound quality that told me something was not quite right.
For anyone using only the P-3A, things are not too complicated. You connect the appropriate digital link (I2S most preferred, followed by AES/EBU, S/PDIF, and TosLink, in that order), the RCA analog outputs, and finally the DC power (from the standard power supply or, preferably, the Monolithic Sound P3 Perpetual Power Plant. Momentarily pressing and holding the Input button cycles through the various inputs, and pressing and holding the Program button selects normal/inverted absolute phase. A program-display LED indicates the sampling rate according to a code: red for 44.1kHz, green for 96kHz, off for 192kHz.
The plot thickens when the P-3A is used in combination with the P-1A. The P-1A has three inputs (I2S, AES/EBU, S/PDIF) and three outputs in the same formats. The selected input is present on all outputs. Regardless of the input, the recommended connection to the P-3A is via the I2S. (A short I2S cable is included.) Pressing the Program button cycles through four options: 1) Bypass, 2) CD Resolution Enhancement, 3) Speaker or Speaker/Room Correction (when they become available), and 4) both Speaker or Speaker/Room Correction and Resolution Enhancement. "Bypass" is not true bypass, just bypass of the Resolution Enhancement (Output Bit Density) function controlled by the Program button; the P-1A's jitter-reduction circuitry and upsampling (if selected) are still engaged.
Now the real fun begins. The name of the game is upsampling/resolution enhancement, but there are two ways of achieving upsampling in combination with the P-3A. The P-1A can be set to output 44.1kHz, which will leave it to the P-3A's circuitry to upsample the data to 96kHz. Alternatively, the P-1A can be set to output 96kHz (or 192kHz, when it becomes available). In that case, and assuming that the P-1A/P-3A connection is via the recommended I2S, the P-3A has to be set to the I2S Direct mode (by pressing and holding the P-3A's Program button when I2S has been selected), which bypasses the P-3A's CS8420 input receiver. I tried both arrangements, and preferred the I2S Direct mode, with the P-1A's output having been set to 96kHz. I won't describe exactly how the P-1A's Sample Rate is set—suffice it to say that it involves pressing two buttons in sequence until the right LED lights up.
The sample rate having been taken care of, the next task is to set the Output Bit Density (16, 18, 20, or 24 bits). This controls the P-1A's interpolation algorithm, and the setting applies to all inputs. The normal setting is 24 bits (which, for some reason, the instruction manual fails to point out). You set this in the third control layer, which you access by pressing, holding, and letting buttons go in just the right order. Then, finally, the P-1A/P-3A combo is ready to do its job. At least the settings are retained in nonvolatile memory; power interruption does not erase them.
The P-1A and P-3A are very small boxes (even smaller than the Bel Canto DAC 1), so placement is not a problem, but you have to decide whether you want them vertical or horizontal. Each unit comes with a matching metal stand that allows for vertical orientation, but this arrangement wasn't all that stable and didn't seem ideal for vibration control. I ended up placing them horizontally, side by side, using the stick-on rubber feet provided.
Linking the P-1A/P-3A was the Mystic Reference I2S cable from Magic Audio; the power supply used for both processors was the upgraded Monolithic Sound P3 unit (see "Upgrades" sidebar). Perpetual Technologies recommends use of an AC mains power-regenerating device and upgraded power cord; I used the PS Audio P-300 (with MultiWave option set at "SS5") and TARA Labs Decade power cord. The overall transparency was improved by placing a VPI DB-5 "magic brick" atop each processor and a Shakti stone atop the P3 power supply. The P-1A and P-3A were among the more slow-to-break-in products of my experience; sonic improvements continued to emerge even after four or five weeks of use.
Most of my listening was to the P-1A in combination with the P-3A, which I'm told is the configuration used by most Perpetual Technologies customers. However, I also spent some time listening to the P-3A by itself, using a direct AES/EBU connection from the PS Audio Lambda II transport.
On its own, without an assist from the P-1A (but with the upgraded Monolithic Sound P3 power supply), the P-3A gave a very good account of itself, its sound quality generally competitive with the Bel Canto DAC 1 and the MSB Link III with upsampling board, Full Nelson upgrade and P1000 power supply (see, respectively, the December and September 2000 issues of Stereophile). (For the P-3A and the Link III, the inputs were AES/EBU; for the DAC 1, which has no AES/EBU input, the input was coaxial S/PDIF.)