As great as SACD and DSD sounded in the standard modes, the PDP 3000 HV has a trick up its sleeve that takes the sound quality to another level. The SACD and DSD filter options, mentioned earlier and described in detail in the sidebar, proved their worth. With Mode 1 engaged (what would be the only filter on other DSD DACs), DSD sound is as I’ve described—stunningly transparent, immediate, detailed, and present. Even if this setting represented the PDP 3000 HV’s DSD performance, I would have still thought it the best DSD I’ve heard. But switch to Mode 2 (a filter with a higher cutoff frequency and gentler slope) and all those qualities stepped up a notch. The sense of air and detail increased, and with it the timbral, dynamic, and spatial realism. The final step is to remove any filtering (Mode 4, or what T+A calls “True DSD”) and listen to the raw 1-bit datastream. The combination of no filter, T+A’s custom DSD DAC built from discrete components, and an analog output stage that has been optimized for DSD playback elevated the performance to new heights. The wider filters allowed more very fine detail to be resolved and stripped away that last scrim between you and the instruments. On the Alex de Grassi track with Mode 1, you can sometimes hear him breathe or shift in his chair, but it’s not very distinct. In Mode 2 these sounds are more clearly heard as what they are. But Mode 4 renders them so clearly that you get a goosebump-inducing impression of a person in front of you. In Mode 4, the illusion of hearing live music-making on the James Matheson album was absolutely startling. In fact, the third movement of the String Quartet was almost overwhelming in its intensity.
These filter modes are for decoding DSD files via the USB input. SACD playback also offers two filters, with the wider-bandwidth filter (Filter 2) offering a greater sense of palpability, fine detail, and sense of an instrument hanging in space surrounded by the recorded acoustic.
There’s one big caveat, however; the wider filters won’t work in all systems, and engaging those wider filters (or running with no filter, as in Mode 4) has the potential to damage tweeters. See the sidebar for more detail for why these filters are necessary and how the wider filters have the potential to damage your system.
The T+A PDP 3000 HV is an extremely sophisticated, versatile, and highly musical disc player. When playing CDs and decoding PCM files, the PDP 3000 HV was among the best DACs I’ve heard, particularly in its ability to convey the music’s dynamic expression and rhythmic flow. It offers the kind of presentation that draws attention to the music rather than to the sound.
Had this been all there was to the PDP 3000 HV, it would have earned an enthusiastic recommendation. But the machine also offers what is far and away the best SACD and DSD playback I’ve heard. The various SACD and DSD filter options, unique to T+A, vault the performance to new levels of transparency, resolution, and realism. The PDP 3000 HV is an SACD and DSD lover’s dream.
Although $22,500 is a lot of money for a disc player, the PDP 3000 HV nonetheless represents high value considering the performance, sophisticated technology, and battleship build-quality. Many companies charge this much or more for less. It took a long time for North American audiophiles (myself included) to discover this 38-year-old German company, but the PDP 3000 HV proves it was worth the wait.
Specs & Pricing
Type: CD/SACD player and DAC
Digital inputs: AES/EBU (x1), SPDIF on RCA (x3), SPDIF on BNC (x2), TosLink optical (x2), USB (x1)
Digital output: PCM on RCA jack (x1)
Formats supported: CD, SACD, DSD (up to DSD512), PCM up to 384/24
Analog outputs: Decoded PCM on RCA and XLR jacks, decoded DSD on RCA and XLR jacks
Conversion: Double-differential quadruple converter with four 32-bit sigma-delta DACs per channel (PCM); T+A True-1Bit DSD conveter (DSD)
Filtering and upsampling: Custom T+A PCM upsampling filter with four filter options; custom SACD filter with two filter options; custom DSD filter with three filter options
Disc mechanism: Custom T+A linear-tracking drive
Power: Dual IEC AC input jacks
Dimensions: 18" x 6.7" x 18"
Weight: 57.2 lbs.
T+A ELEKTROAKUSTIK GmbH & Co. KG
Planckstraße 9 – 11
D - 32052 Herford, Germany
Phone +49 (0)52 21 / 76 76 – 0
Loudspeakers: Magico Q7 Mk.II, EnigmAcoustics Sopranino self-biasing electrostatic super-tweeters
Preamplifier: Constellation Altair II
Power amplifiers: Constellation Hercules II and Berning 211/845
Digital sources: Aurender W20 music server, Berkeley Alpha USB USB-to-SPDIF converter
Support: Critical Mass Systems Maxxum equipment racks (x2), Maxxum amplifier stands (x2)
Loudspeaker cables: MIT Oracle MA-X SHD and MIT ACC 268
Interconnects: MIT MA-X SHD, AudioQuest WEL Signature and AudioQuest Wild
Digital interconnects: Audience Au24 USB, AudioQuest Wild Digital AES/EBU, AudioQuest BNC
AC: Four dedicated AC lines; Shunyata Denali conditioners, Shunyata Sigma power cords
Acoustics: ASC 16" Full-Round Tube Traps, ASC Tower Trap, Stillpoints Aperture Panels
Accessories: Shunyata cable lifters, Stillpoints UltraSS and Ultra6 isolation
DSD Primer and the PDP 3000 HV’s DSD Modes
Direct Stream Digital (DSD) is Sony’s tradename for an encoding technique in which the audio signal is sampled very quickly (2.8224 million times per second), but with an amplitude resolution of just one step per sample. This encoding couldn’t be simpler; if the signal is increasing in amplitude, a binary “1” is recorded. If the audio signal is decreasing in amplitude, a binary “0” is recorded at each of the 2,228,400 sample points per second.
This technique produces a signal with a very wide bandwidth but a very high level of quantization noise. In fact, the signal-to-noise ratio of raw DSD encoding is just 6dB. You may recall that a digital audio system’s signal-to-noise ratio is determined by the number of bits in each sample. A 16-bit system gives us a signal-to-noise ratio of about 98dB, or 6.01dB per bit.
So how can a digital encoding system with such high noise work? A technique called “noise shaping” shifts the noise out of the audioband to a higher frequency. With standard-rate DSD encoding as found on SACD (called “DSD64” because the 2.8224MHz sampling frequency is 64 times that of CD’s 44.1kHz rate) the noise begins rising just above 20kHz, with most of its energy between 40kHz and 100kHz. All SACD players and DSD DACs incorporate a low-pass filter (typically at 50kHz) to prevent this noise from appearing at the product’s analog output jacks. If left unfiltered, this ultrasonic noise could cause your power amplifier to become unstable or oscillate, potentially overdriving your tweeters. The amplifier can also create intermodulation products by this combination of high noise level and the audioband musical signal. Even if the amplifier is well behaved when driven by a high level of ultrasonic noise, that noise puts a burden on tweeters. You don’t hear the noise, but the tweeter’s voice coil must still dissipate that energy. The wider the amplifier’s bandwidth, and the more stable and linear it is at those ultrasonic frequencies, the less problem DSD’s ultrasonic noise poses. Moreover, today’s tweeters are better at handling higher power levels that tweeters of twenty years ago.
As the DSD rate is doubled (“double-DSD” or “DSD128”), the noise energy can be spread out over a wider bandwidth than with standard-rate DSD. The filter requirements are thus relaxed for DSD128 compared with DSD64. Quad-rate DSD (DSD256) allows the noise to be spread out over an even wider frequency band, again changing the optimal filter characteristics.
Almost every other DSD playback device I’m aware of employs a single filter and noise-shaping algorithm for all DSD rates, compromising the performance of higher-rate DSD files. But the PDP 3000 HV allows you to select between two different SACD filters and noise-shaping algorithms when playing SACD, and four different DSD “Modes” when decoding DSD files via the USB input. These Modes correspond to three different DSD filter characteristics.
The SACD 1 mode is a standard filter just as you’d find in any SACD player. The owner’s manual doesn’t specify the cutoff frequency, but my guess is it’s around 50kHz. SACD 2 has a higher cutoff frequency and a more gentle slope. T+A recommends this filter for systems with amplifiers that can handle a higher level of ultrasonic noise.
Moving to the DSD DAC Modes, when playing files through the USB input, Mode 1 is a standard low-pass filter that is suitable for DSD64 and any playback system. It removes all the ultrasonic noise. With Mode 1 engaged, the PDP 3000 HV sounds as I described it: the best DSD64 playback I’ve heard. But switch to Mode 2 (a higher cutoff frequency and gentler filter slope) and the sound takes on an increased transparency, resolution, and timbral realism.
Mode 4 is what T+A calls “True DSD.” This mode imposes no filter whatsoever, letting the DSD DAC run wide open. T+A recommends this mode only for higher-speed DSD, and if your amplifier has a wide bandwidth with very low transient intermodulation distortion. As described in the review, Mode 4 is absolutely startling in its transparency. Very fine detail, such as the sound of fingers moving on strings on the Alex de Grassi track, for example, took on a greater vividness.
Mode 3 isn’t a fourth filter option. Rather, Mode 3 automatically engages the Mode 2 filter for DSD64 files, and switches to Mode 4 (no filter) for DSD128 and higher sources.
The PDP 3000 HV’s manual cautions against using the True DSD mode (no filter) unless your amplifier and loudspeaker can handle the high levels of ultrasonic noise. I listened in Mode 4 with DSD64 files without a problem, although I kept the volume level moderate. As a safety measure, you must allow the PDP 3000 HV to engage True DSD in the basic system configuration menu in additional to selecting it from the DSD Mode menu.