Taking the “The Sound Floor”

Prelude:

I’ve been away from writing reviews on the site because my system was undergoing major sonic shifts affecting my LP’s relative sound rankings. The major change was the use of low ultra-low noise LT1028’s for my ancient LF353N integrated circuits in my crossover which I’ll detail later.  I wrote the rough draft of this in November and just as I was about to finish my subwoofer amplifier decided that it needed the same attention as its crossover.  It did this by failing without any visual evidence.  I’ll get into the details later, but the amp has now been through two stages of changes with more to come.  These involve the storied Black Gate capacitors.  The sound is stable and much improved now so I’ve decided to get back to this article.

“The Sound Floor”:

The great Arthur Salvatore of Supreme Recordings fame is the father of this mythological audiophile concept.  See the Salvatore Sound Floor.  It is the dominant sonic attribute in determining an LP’s greatness.  It is only mythological in that the typical Electrical Engineering precepts and measurements do not corroborate the findings.  It is audio science (another mythological concept).  By that I mean those with experienced ears can discern the effects on sound and convey those findings.  High End Audio engineering is both art and science, but only because the conventionally not measureable art has not been properly explored.  What follows is an initial delving into the subject.

Salvatore states that the Sound Floor is a related, but very different concept from signal to noise ratio and noise floor.  Salvatore attributes sound floor to shortness of signal path combined with good noise floor performance.  My own recent experiences in this area have led me to a more conservative approach.  In essence, I believe noise floor and signal to noise ratio do well explain much of this phenomena.  One also needs to look at bass control as a major factor.  Dynamics and timbral accuracy also are of great importance.  I would submit that these four, sound floor, bass control, dynamics, and timbral accuracy are largely independent sonic attributes.

I own quite a few of the top Supreme Recordings from the divinity and demi-god lists.  Though always excellent records, I did not think they were great.  My tune has changed recently due to an unexpected improvement in my system from a recent modification.  My system has a separate active crossover for its bass/subwoofer section, the integrated circuits of which I just upgraded.  The final switch from LF353N’s to LT1028’s (adapted) made a gigantic improvement in my noise floor and suddenly the Salvatore Gods came alive. (It blows my mind that this solid state component was the weak link in an all tube system outside the subwoofers, but it was an additional gain stage.)

The Gods that came alive were sound/noise floor performance.   One can hear with most of these Salvatore discs their blatant superiority in this area.  Backgrounds are black, treble has the finest detail, and you can just hear that the sound is more alive.  Alive, not in an immediate, or dynamic sense per se, but just in a “Sound Floor” way which is all about the blackness of background allowing subtle detail and micro dynamics come to the fore.  Sound floor is a very addictive sonic attribute to be sure, especially since it is hard to put your finger on exactly what it is doing, let alone describe it.  It is all encompassing and Mr. Salvatore’s list of Supreme Recordings is principally driven by it and is an impressive codification of what is available (Salvatore might be the large criosphinx above while this reviewer's thoughts might be the eagle in his presence).

A Brief Interlude:

As mentioned in the prelude above, I also have been forced to finally address my venerable Electron Kinetics Eagle 2A amplifier (a great subwoofer amp).  After 25 years of flawless operation, one of the main 80,000uF filter supply caps failed, taking out the bridge rectifier.  I had been planning to get the recap kit for the amp for some time, but since I’d taken the trouble to make a crude schematic many years ago I decided to roll my own capacitors.  The failed cap was the priciest and I was lucky that my good friend Dr. T had a free pair of Hitachi 140,000uF (yes!) and a dremel tool to assist my hack saw in cutting down the posts and screws to fit in the Eagle chassis. 

The bigger change was the IXYS hexfred bridge rectifier.  I believe this was a significant upgrade and was responsible for the bulk of the change in sound.  The change was further improved noise floor performance.  The refreshed larger caps I suspect accounted for a little tighter bass, but it also might have been a bit slow.  So, the quest for replacement and upgrade of the remaining capacitors began.

After much research, I realized that the quantities and prices for the Super Gold ARSA Silmic II caps I was looking for on eBay were adding up, so I decided to investigate Black Gate Capacitors which I had ruled out.  For those unaware, Black Gate has been out of production for nearly ten years.  They were true audiophile electrolytic capacitors using graphite particles based on a patented design. 

I researched the heated debates tinged with religious fervor (Dr. T electrical engineering types versus the “Audio Fools”).  With the Black Gate promise of dramatically lower noise floor, I was sold given the strides I’d been making in this area.  My Eagle is unlikely to ever be replaced in my current speaker system, so I saw no reason not to spend a few more bucks on such a deserving amplifier.

At this point, Black Gate C Series Capacitors have replaced four 2.2uf ($10 each) and two 47uf 50v ($25 each) caps.  The next stage will be eight 22uF PK Black Gates ($5 each) and two 1000uF FK’s (very hard to find, used only $10 each!).  The final stage (which will involve some engineering help) will be to replace the single channel LF351N’s with LT1028 (TO5 cans on hand now).  The LT1028’s will need bypass caps from plus to minus and I’ve acquired some more unobtainium in the form of a pair of 0.1uF Black Gate NX Hi-Q capacitors ($15 each).  I’ve also got a pair of the 22uF PK for a position in my active subwoofer crossover.  I’d like to get eight filter supply caps for the crossover if all this goes well, but that is stupid money at $125 each and over five times the already controversial original retail prices.

It has taken some time for just the Black Gate C Series Capacitors to break in (and it has been as arduous a process as described elsewhere).  They have delivered in big ways.  Always in the past cap changes have been subtle and nothing to get overly excited about especially given that fresh caps do improve the sound some.  Well, Err, ah, Black Gates are a whole new ball game.  The esteemed Martin Colloms claims they are better than most polypropylene film capacitors.  All I can say is that they completely obliterate the typical garden variety electrolytic (and I suspect handily trounce the likes of Oscon, Silmic, Cerafine, etc.). The noise/sound floor is hugely improved (likely some synergy here with the IXYS bridge and LT1028s) and everything else is like a dream.  My reservations about the LT1028 bass being too lean have been obliterated.  The perceived slowness of the massive Hitachi filter caps; obliterated.  Not only is noise floor performance greatly enhanced, but a tube rightness has been restored to many of the vintage LP’s in my collection.  It is as if the Eagle amplifier, while improving greatly in bass and treble dynamics along with control and definition, also has become more tube like (bridge rectifiers and Black Gates).  (Pardon me, but it is old hat with these speakers, but I’ve consistently found that changes in how a system projects bass affect the entire perceived sound including the midrange and the treble.)  I am excited to continue recapping as the Black Gates have been game changers.  If all goes well with the continued changes I will be able to recap large parts of my line and phono stage power supplies with black gates for original retail price (five times or more the price of a typical electrolytic, but peanuts compared to other capacitor upgrades like V-Cap Teflon.)

And now back to Sound Floor:

Where might I demur from the Salvatore mythos?  Well, I also got a big boost on the digital front (for which I care little) from these changes.  I recently acquired a pioneer Blu-ray player for its superb DAC’s.   These Wolfson DAC’s sport a 117 dB signal to noise ratio (never mind the budget ICs behind them and extensive circuitry).  Hardly high end, but I could easily hear a nice difference on this with some of the Silvestri CD’s from the EMI artist set reviewed on this site.  True, not as good as the Salvatore Gods, but one wonders with some pricier, lower noise ICs backing the Wolfson what might happen (of course the master tapes, would be 68 dB at best).  Here we have a very long signal path after the Wolfson with far too many stages and yet the noise floor numbers do translate.  (As an aside I'll mention that the gap between vinyl and CD has widened with all of these improvements, especially the black gates.)  In summary, I would say that the Salvatore mythos misses out on identifying this truly as only a noise floor phenomenon while blending in some other principals of component selection (shortness of signal path).  We need to also discuss bass control, dynamics, and timbral accuracy.

For me, bass control is probably the second most critical component sonic attribute.  If the sound chain (both recording and playback) does not define the bass well, instruments are not fleshed out.  Without a proper bass foundation, perceived treble clarity and extension also will suffer (an acoustic phenomenon). With a tubey bass or ill-defined bass, one can achieve some interesting sound effects particularly in the midrange and often this is the golden glow we hear with tubes and early stereo tube recordings.  (EMI by far has the tightest bass of the golden age labels).  It can be very nice and one may optimize a system for these recordings, but the ill-defined bass will undermine more modern recordings.  An example would the original Vanguard Plow that Broke the Plains or the Analog Production reissue (review forthcoming).  Despite some great things going one with this LP when the massed strings come in at the end of this piece, one is left with an ill-defined blob of strings.  Nothing can rescue this as it is on the tape, but fortunately most of the LP sounds quite magical with this tubey signature.  I suspect dynamics might even be enhanced by this coloration, but this may be an advantage of perhaps higher voltage tube circuitry in the signal path which would aid dynamics.  The great Salvatore’s inclusion of the Vanguard Plow makes me wonder if his reference system is exemplar in low bass and definition.  My preliminary listens to many of the EMI on his list indicates they are under ranked (yeah EMI!)

Timbral accuracy fleshes out the overtones of instruments and is a fine detail often lost due to imperfect turntable motor performance and other vibrations throughout the sound chain.  Here, Salvatore’s shorter signal paths (tubes based circuits) come into play with less opportunity for vibrational mayhem.  Remember, any AC signal puts out a field.  Vibrations can interact with these fields and the AC signal will be affected.  Many theories abound on Capacitors and other components and I will not detail something this well known.  Inherently, the length of solid state chains even with good noise floor performance tends to have negative impacts in the timbral area from a host of issues.  Fortunately, solid bass is key to instrumental definition, so often solid state can achieve adequate timbral performance.  This is a most confusing area as tube recordings with ill-defined bass sometimes have a heightened sense of harmonics that while often a beautiful coloration, does not match anything heard live in my experience.

Dynamics are also important.  To be sure macro dynamics are well known and of importance (Salvatore).  This is one of the hardest things for a playback chain to achieve.  However, often overlooked are micro-dynamics.  It has been a while since I’ve thought much of it, but generally here is an area where tubes often have advantages.  At best a solid state device can deliver double its rated power at for a dynamic peak.  For tube units, ten times the rated power might be possible.  This has much to do with the liveliness of tube based chains.  Generally, the Salvatore Gods may have some micro dynamic weaknesses due to their generally solid state recording chains.  Things like ½ speed mastering can help.  (Unfortunately, more power always means worse noise floor performance.  Higher sensitivity speakers also project more noise.  Balance is needed.)  The Black Gate capacitors are strong in both types of dynamics and this might be attributed to their patented design’s alleged speed of power delivery.

I have much listening and comparisons to do in the upcoming months.  Unfortunately, the arduous Black Gate break in period for the upcoming changes is going to make this a trying period for review.  I am still leery that there may be a point with Black Gates where one reaches the point of too much of a good thing, hence the need for a more plodding methodical approach.  I may do a general post on my current thinking on Lyrita, EMI, Decca, RCA, etc. rather than detailed reviews for which I would like to have a more stable reference system.  I will also have to update some previous postings and review as these awesome sound floor improvements have been great for some LPs while others have experienced much less improvement (the reviewed Nonesuch stands out.)

UPDATE (1/31/2014):

I was reading up on 300B tubes (need a new set) and I found discussion of the signal to noise ratio of these amps.  Apparently, the 300B and any directly heat triode drivers have almost unmeasurable noise.  The downfall is the other tube stages in these amps which brings them into the measurable range.  The 300B is a noise floor thoroughbred.  This explains why I am hearing big noise floor improvements in my solid state sub-woofer section from the changes discussed above.  It is clearly the weak link versus the rest of the amplification which is all directly heated triode.   By the way, the final cap upgrade is now in and already has made a very strong sonic impact on the noise floor with amazing sound despite needing more break in.

UPDATE (3/3/2014):

The second wave of Black Gate Capacitor upgrades broke in quickly as some were used, so last week I moved on with the Black Gates in the crossover and also replaced the diodes with Schottky diodes ($2 total). With the Black Gates through the worst of their break in period now, it is clear how powerful the impact of the Schottky has been on the noise/sound floor. Immediately on first listen it was clear the Schottky had completely cleaned up some hash in the treble with better soundstaging, but their was also a hint of dryness to the sound and other affects that were not musical. The improvement was more dramatic than I anticipated, but in hindsight the crossover being an earlier amplification stage should be a greater determinate of the sound. After settling in, the changes have made all of my tube based recordings improve nicely despite their limits in noise floor performance as a solid state signature has been removed from the system (i.e. crappy diodes and crappy electrolytic capacitors).

The Schottky impact has been to not only reduce noise floor, which it did some, but it has also clarified the bass region (not really tighter or more impact, but clarify). In the words of psuedoscience, the lack of any ring or overshoot with the Schottky seems to have done more than just improve the sound floor. It has stopped a ringing in the bass that has allowed more clarity in their definition. This has translated into finer detail and texture all the way into the treble. Typically this is what you hear with noise floor improvements, but there is more improvement here and in the bass by a magnitude at least than what one would expect just from noise floor. I wrote in one of the Lyrita reviews that it made nimbus pressings sound like the more defined EMI pressings. I proclaim this the Schottky effect which in short is a leap forward in clarity due to the abscence of ring and overshoot. (My good friend Dr. T, the electrical engineering PHD who "pities the fool who tweaks" need not see this, but does get dragged into design decisions so he'll hear it eventually.)

How does this impact are four sonic attributes and our understanding of the elusive Salvatore "Sound" floor? First, I think the lack of solid state diode distortion and stellar good old electrical engineering noise floor is critical to the sound floor. Maestro Salvatore would disagree sharply in that he also favors shortness of signal path which I have no doubt is true, but the fact remains all of the monumental gains in my system have been from addressing noise floor issues with solid state components with no shortening of signal path. Clearly noise floor is a huge component of the sound floor.

Bass definition continues to be another huge factor as time and time again recordings with bloby ill-defined bass are unable to project clean and pristine high frequencies. The same can be said of systems with poor bass control. Add in the bass clarity of the Schottky effect and we see the importance of low frequency foundation.

Dynamics and Timbre also appear to be very interrelated with the sound floor and removal of distortions like the Schottky effect. Do not be fooled by false timbral accuracy that some tube chains produce as an affect of poor bass definition. There is something very real here with the distortion characteristics and microdynamics of tubes. I reference an article of interest for tube and sound floor lovers that explains their superiority; The Hidden Harmonics behind THD

The Fifth Parameter:

This is kind of like the Fifth Element (watch the movie).  I have been scratching my head over how the Quadophile has been getting such amazing midrange performance and I believe this parameter accounts for much of the Quad advantage. The parameter is the sound of the room, a big part of which is smoothness of bass response. The planar Quad has amazingly smooth in room bass response. I measured plus or minus a db at most test frequencies. I recently went through a measurement process with Acoustic Fields and found that my response was fairly flat in this area for a dynamic speaker, but it still pales compared to the Quads. I am tip toeing into the area of room treatments and may use some of the cheap plans Acoustic Fields provides to build their $1000 plus treatments (225 pound bass trap is one and a quadratic diffuser another.) This will be down the line behind some non-Audio projects I'll be doing, but already some preliminary experiments with room acoustics have yielded promising results.

Getting these room parameters right can do big things. A weird one I experienced recently was closing the doors at the back of my room.  By the measurements I thought that having the doors open would yield more 20 Hz reinforcement while garnering less reinforcement around 30-40 Hz. With orchestral LP's doors open (opened up a closet too) did not sound improved.  By closing all the doors I got much better hall presence and bass slam.  That is a very big change. I'll note that for home theater style bass, often having all the doors open gives an airier/nicer sound. Acoustics is complicated!

I suspect that getting the low bass response under control (and other acoustical nasties) may have similar impact to the almighty Sound Floor. It makes sense as accoustical irregularities in the bass and elsewhere will obscure detail. Dennis Foley of Acoustic Fields has a very entertaining series of videos for those who are not acoustic engineers. 

Note: In no shape or form do we recommend products, but free advice this writer likes. Caveat Emptor!