Current Equipment

(Note that some of the below sections are too long and probably should be broken out into separate reviews.)

August 2019 Update: I've replaced Foobar 2000 with JRiver Media Center since the latter has a much better user interface and sounds very significantly better, probably due to playing from a ring buffer in RAM and having very direct audio paths through Windows. It also runs under Linux and Mac.

I've also upgraded from the stock Sennheiser HD 800S cable, which was mechanically failing, to a 2 meter Wireworld Silver Eclipse 7, and it's definitely a very good sounding cable and much better than the stock. It's almost sonically invisible, very clean and accurate. I think it has some sound, but it's very hard to discern, which of course is a good thing.

Wireworld is somewhat unusual in designing cables with the aid of ABX double blind listening compared to a direct shunt. So they're basically comparing test cables to no cable in order to find cables that are the most transparent. This seems like a valid and logical approach, even if it's very different from the approach that essentially uses cable euphonics as tone controls. The former should lead to accuracy; the latter away from it. (That said, for power-level cables like speaker cables, design may have a larger effect than for signal level cables, and also greater matching/equipment compatibility effects.)

Wireworld does have some proprietary cable designs and insulating materials, and use very high purity metals. They claim the materials and designs are solidly based in science, to go along with the science-based listening tests. As a scientist, it's hard to fault either approach, and the designs seem scientifically sound. They claim the helical twist of the cables is more compatible with the unimpeded propagation of electromagnetic flux, the insulator has low triboelectric effects, etc. As an audiophile, the result seems very good.

The Silver 7 Eclipse is probably a third from their top of the line a couple years ago, so it's not state of the art, but it was also a good price due to being older and a review sample. Retail was $650, but they sold it for half price, being used, etc. I got it directly from their web site.

Listening further, sound through the cable has much better time accuracy (less time smearing) which results in much more precise localization in space of minimally miked recordings and of panned multitrack pop recordings. Timbral accuracy is also much better. It's like having a much cleaner window into the recording. This is a major upgrade for my system, and it shows how mediocre the factory cable is.

Surprisingly to me, the optical media drive I used to rip my disc collection made a very audible difference in sound. The most recent, highest resolution LG 4k Blu ray burner I got at the end of 2018 sounded best when ripped using Exact Audio Copy. It was very significantly cleaner sounding than older DVD drives, and slightly better than my ASUS non-4k Blu ray burner. Ripping at maximum speed produced the best sound, perhaps due the effect described next.

My theory is that the highest throughput media format drive (4k Blu ray) by definition must have the highest bandwidth and timing accuracy to go with it, where such increasaed accuracy benefits even when ripping CDs, though of course this disagrees with the "bits is bits" argument. But the results were very clearly audible to me: more realistic instrument sounds, better hall sound, deeper reverberation tails, etc., that seem indicative of higher resolution sound in general. This pattern has been repeated in going from lower resolution recordings such as CD to higher resolution DVD to 192 kHz 24-bit, etc. (In fairness, CD resolution can sound very good if well-engineered, but most CD sound is not well-engineered.)

Current audio rig: Sennheiser HD 800S plugged into a Benchmark DAC2 HGC, fed from my Pioneer Elite DV-58AV mentioned below or from Foobar 2000 on PC playing lossless FLACs over USB. Foobar 2000 is a somewhat minimalist Windows audio player that correctly handles FLAC, bit-accurate transit, long word lengths, etc. It's specifically recommended by Benchmark, and I would recommend it also. At the recommendation of a friend, I'm also using Exact Audio Copy for ripping. EAC does a very thorough and careful job of correctly getting the audio data off CD. For CD track information, I paid the small (7.99 USD) one time fee to use GD3 metadata with EAC. (Metadata are needed since CDs largely don't have track, performer, title, etc., information encoded onto the discs themselves. This was kind of a dumb oversight in the design of CDs.) GD3's coverage has been quite good, and the data are more consistent than free sources like FreeDB, and I did not see support for the large, competing metadata provider Gracenote in EAC. Some of the GD3 data are somewhat inconsistent, especially for classical music which is mostly what I listen to now, so some hand editing has been necessary in some cases, and this is somewhat disappointing. Here are my notes about the excellent Benchmark DAC1 PRE.

For ripping with EAC, I've found that Burst Mode gives the best sound. Comparing all four modes available in EAC (Under EAC -> Drive Options -> Extraction Method), Burst Mode sounds distinctly clearer and more realistic to me. Burst Mode is also the fastest and highest data rate, and with CRC checking both against test reads on the disc itself, and against networked checksum databases, the bits will be correct. Being the highest data rate, the bandwidth will be the highest and the timining accuracy needs to have the tighest margins. Whatever the reason, Burst Mode made the best sounding rips for me. As a nice side effect, a CD can be ripped in a few minutes in Burst Mode on a recent Blu-ray 4k drive.

(For some damaged discs, I did need to rip a few tracks in Fast Mode or Paranoid Mode in order to get a good rip with the checksums correct. For some discs, Paranoid Mode worked better and suprisingly faster than Secure Mode. I did monitor the drives for stalling and overheating on difficult rips.)

Previously was using (and still have) a Headroom headphone amplifier module (picked up at the factory in Bozeman, Montana) fed from Assemblage DAC-3 de-jittered and sometimes upsampled and interpolated with an Assemblage D2D-1. Assemblage are exceptionally high-quality, high-value kits and assembled products made by Sonic Frontiers. No other relatively affordable audiophile gear uses the top-quality digital, analog and passive components of the Assemblages. Their general design concepts are outstanding also. See my initial DAC comments below and my listening notes about the Assemblage D2D-1/DAC-3 combo.

The Sennheiser HD 800S headphones are unbelievably good, but be aware that they require at least 200 hours of break in before they start sounding open and uncolored. They also get the most out of very high-quality electronics behind them, such as the Assemblage DACs and Headroom headphone amplifiers. They have enough resolution to show the flaws in mediocre electronics.

The Cardas replacement cable for my previous HD600 dynamic Sennheisers is outstanding! Mine is starting to bloom after breaking in for about a day and it clearly has vastly more resolution and clarity than either of the stock Sennheiser cables. The Cardas is about $200, which seems expensive compared to the headphones, but it represents an excellent value as a hand-crafted, high-end audiophile product compared to the scientifically mass-produced, high-performing headphones. (Sennheiser has two compatible stock cables. The thin Kevlar reinforced one is ok, but the wires break at the earpiece plugs easily. The thick one sounded terrible when new and only got slightly better after months of break in. The thick cable does have much larger earpiece plugs which seem comparatively indestructible.) Do not use the stock Sennheiser cables! Sennheiser using these cables is like Ferrari shipping their cars with minivan tires.

To anyone who thinks cables don't make any difference, consider that they are a complex, physically-extended, capacitor, inductor and resistor network with microphonics, dielectric effects, electromagnetic, electrostatic and mechanical modulation, etc. They need to have a signal running through them for many hours to form the dielectric. The same thing happens in capacitors, but it happens much more quickly there since the power levels in a typical capacitor application are much higher.

Have not upgraded the HD 800S cable yet, but will probably consider Cardans and Wireworld Platinum Eclipse eventually.

My current digital source is a Pioneer Elite DV-58AV bought for $350 from Vann's. It retails for $500. Manufacture date is September 2007 in Thailand. It's gotten rave reviews and has excellent components in its analog chain, such as the superb Burr Brown PCM1796 DACs. Unlike the several years older consumer DV-563A, the 58AV will send unencrypted DVD-Audio to the S/PDIF digital outputs to an external DAC, receiver or surround processor. Neither player will send SACD out their S/PDIF outputs, but the newer player will send both SACD and DVD-Audio out its combined digital video and audio HDMI output. (An HDMI receiver would need to be able to handle encrypted signals for encrypted DVD-Audio and SACD recordings. Such receivers may be uncommon, though Pioneer lists some in the owner's manual.) The earlier player does not have an HDMI output; its only digital audio outputs are S/PDIF over Toslink and RCA. It's main downside is that it doesn't play Blu-ray, and none of Pioneer's Elite Blu-ray players have the same quality of analog electronics. Maybe future models will. Given the specific SACD and DVD-Audio format support and high quality DAC chips, the DV-58AV is clearly specialized as a high-definition audio disc player. As such it's a somewhat unusual product, especially for a mainstream consumer brand like Pioneer. It seems Pioneer has its some audio fans making things happen.

Both used as digital transports, the unmodified Pioneer Elite DV-58AV sounds vastly cleaner and more resolving than my previous unmodified Pioneer DV-563A. Comparing S/PDIF digital outputs from either player fed into the BNC input of the Assemblage digital gear described above, the DV-58AV is quieter (less noisy/hashy/fuzzy), more realistic, and overall much higher quality. Both artificial reverberation and natural spaces are much more clearly rendered. Instrument sounds and harmonics in general are much more apparent and distinctive. Good recordings like Steely Dan's Gaucho sound phenomenally good through this setup. This is a recording frequently cited as a sonic reference, and the reasons become much clearer when heard through better equipment. Chord changes in multipart harmonies are much clearer. Massed instruments and overdubbed voices sound separate and distinct, and the skill of the playing, singing, arranging, composing and recording are much more apparent than before. This is indeed a great recording on both sonic and artistic grounds.

I have not looked inside or gotten the DV-58AV service manual yet, but Pioneer presumably spent greater effort at producing a cleaner digital output chain in the Elite electronics. The 563A is low-end consumer electronics. The Elite is more carefully and expensively done, call it upper middle-end consumer electronics. In particular, the choice of DACs in the 58AV is surprising since they're four dollar parts typically found in units costing far more. (Four dollars may not sound like much money, but in consumer electronics selling for less than four hundred dollars, it's notably extravagant. It's a rather incredibly expensive, high-quality part in this market segment.)

Starting with such good-to-great basic parts such as the remarkable DAC, modifiers have turned the DV-58AV into high end players. Naturally they add lots of dedicated, low impedance power supplies, replace capacitors, and especially update or replace the analog output stages. In other words they do the same kinds of mods we've been doing since the early Philips/Magnavox CD players.

Previous digital source was a Pioneer DV-563A, which plays DVD-Video, DVD-Audio, SACD, CD. Manufacture date is September 2003. It was a cheap multi-format player picked up as a demo unit from Circuit City, like my previous multi-format Pioneer disc players having unencrypted, high-resolution digital audio outputs. As a digital source it's ok, but the analog section is expectably mediocre. It was chosen for having 96 kHz x 24-bit digital outputs.

Player before that was an October 1998 manufacture Pioneer DV-414 which was also chosen for having high-bandwidth, PCM digital outputs.

Player before that was a heavily-modified Magnavox 560 which was internally a Philips player. Mods came from the New Jersey Audio Society, Ben Duncan's AMP-01 power supply, plus some of my own. NJAS had a nice DC servo modification with LF411 opamps instead of the original electrolytic output coupling caps. The AMP-01 had nice, low-impedance, heavily-bypassed, three-terminal regulator-based power supply. Regulator boards were right next to the circuits they powered. I gave dual-mono, split-rail supplies to the output opamps, plus four more supplies for the digital chips on the output board. Filtered DC for the regulator sections came from a nicely-designed external supply. Ben's AMP-01 preamplifier and its power supply design was very careful and well thought out. I also changed the DAC bit-current capacitors to silver mica, and replaced the good, stock Signetics 5532 output/reconstruction filter opamps with very much nicer Analog Devices AD712s. Internal power wiring was Kimber Teflon wires unwoven from their 8TC speaker cable. Does anyone have copies of my original Usenet postings about these mods?

Got an pair of Ultimate Ears Super.Fi5 earphones, mostly for listening to music while travelling. These are in-ear-canal earphones that compete with Shure and Etymotic. Ultimate Ears are apparently widely used by musicians, well at least ones who use amplified sound. The Super.Fi5 has an extremely detailed and apparently accurate midrange, but is rolled off in the high and low frequencies. It's possible that the dual-armature Super.Fi5Pro performs a bit better at the frequency extremes. The result is pleasing for casual listening, allowing much of the emotion of solo instruments and voice through, but the lack of low and high frequency information is disconcerting and unnatural in a more objective sense.

The choice of earpieces is very important. Use ones that seal properly and allow the deepest placement in the ear canal. I found that armature up, with the cables over the ear sounded better. Silicone earpieces sounded better than the foam, but the foam isolated external sounds much better for plane travel. May want to try Etymotics.

The top Ultimate Ear models, the UE-10 and UE-11, are custom molded to an individual's ears, have 3 and 4 armatures respectively (bass, midrange, treble), and cost around $1,000. If they have the kind of extreme midrange resolution that the Super.Fi5 has but over a wide frequency range, then they could be outstanding.

See links on the left for more thoughts about digital audio equipment, amps, preamps and possible future systems.