Beyond the Ariel

[IslandPink]

Originally Posted by Lynn Olson

You have good taste! The SPU Royal N is exactly the model I'm interested in; the only question is which arm is best - the stock Technics arm, at a princely replacement cost of $65, the Jelco 750D at $500, or the famed Ortofon 212D at who-knows-what cost.

There are a few others Lynn, I had a long list of candidates once from James D who you may know and whose judgement I trust.
The others I can suggest are the SME 3012 II and the 3012 I ( heavier, better ) both a bit pricey on Ebay these days, but likely better than either of these is Thomas Schick's arm -

Schick tonearm

Which I have had rave reviews on, from another correspondent who has an all-valve system and used to use the Mk. 1 3012 . You may have seen his prototype at ETF in 2004 or 2005 . Ok, it's about E900 so not cheap ...

[AJMARS]

Originally Posted by Lynn Olson

A DAC is far beyond my capabilities; although if I did one, I'd probably use the ESS Sabre 9018 DAC in current mode (if the load is less than 10 ohms, it goes into current mode, and you can use a passive I/V converter). The Buffalo kit looks like a good starting point - just throw away the analog circuit, and use a Raven variant to derive the gain lost in the passive I/V converter.

Lynn... I have been informed about your opinions concerning the passive I/V conversion plus filter, by Jean Michel LeCleac'h, I've seen a part of your posts concerning these options, a few months ago, I started modifying my personal Dac using PCM1704 with passive I/V and filters. The first step was to use a WE 417A tube (5842) as gain device. I built it, and lived a little bit with, I am really happy with the result. Recently, I decided to modify the output stage of the DAC, and I will use a classical Pultec structure (in fact only the first stage of the Pultec PC10). The gain is higher compared with my WE 417A amplifier, and I will be able to go to smaller resistances (about 10 ohms).

Thank you for your ideas, they really helped and decided me to go to passive I/V converters.

Here, a pic of the DAC in progress, with the Pultec like power supply and a Pultec PC10 output amplifier

Regards
Andre

[audiophiliac]

Originally Posted by DrBoar

Regarding open baffles, how much is due to open and how much is due to wide baffles?

I went to a high end Hifi Show, and for some reason i like wide speakers? When my HiFi interest started in the 70 more than 90% of all speakers were wider than they were deep. "Averege" speaker had a 8" driver in a 12" baffle or a 10" +5" in a 15" baffle. Fast forward to past the millenium most speakers are deeper than wide 6-8" wide baffles for 5-6.5" drivers seem to be the norm. So basically the baffle step has moved an octave up from 250-300 Hz to 500-600 Hz and this is were a lot of music has it peak power.

Does this affect how the speakers energize the room in this octave?

I seem to prefer the old style cabinets, not allways but quite often. I do not know if I am attraced to the old time look or if there is some accoustic signature with the very narrow baffles of current speaker design that I do not like.

There are many exeptions I like Anthony Gallos Reference 3 that is anything but a wide box and I have heard many bad wide boxes. But still there is nagging perceptions about some common denominators for the two groups, something that is elusive and I have yet to define...

I might have a go at open baffles using vintage alnico driver, even the tweeters are alnico cone drivers so that it diploles all the way
Open, Wide, Vintage, Alnico, what can go wrong?

Dr Boar,
What you are taking about is very interesting, however I feel you would have much more feedback on it if you started a separate thread. Baffle step can be your friend, if you treat it right.

[kevinh]

Originally Posted by Lynn Olson

A DAC is far beyond my capabilities; although if I did one, I'd probably use the ESS Sabre 9018 DAC in current mode (if the load is less than 10 ohms, it goes into current mode, and you can use a passive I/V converter). The Buffalo kit looks like a good starting point - just throw away the analog circuit, and use a Raven variant to derive the gain lost in the passive I/V converter.

I like the variant you've done on the Raven - I'm curious, what did you find when you changed the circuit from the version on the Web page? (I never spent much time on the Raven, so I'm not an expert on refining the circuit and fine-tuning it.)

Well my design ability is far less than your, that is why I was hoping you'd do it , I agree about the Latest Buffalo Dac as a good platform in current mode.

[6L6]

Originally Posted by audiophiliac

Baffle step can be your friend, if you treat it right.

This is a fascinating statement, can you elaborate? Baffle step only seems like a detriment to me...

[soongsc]

I think you can tailor it to eliminate some room modes.

[badman]

Originally Posted by soongsc

I think you can tailor it to eliminate some room modes.

I could see it being used as an eq filter in a very specific setup, but seems like it'd be difficult to tune and only effective at certain angles, certainly not a panacea and perhaps no better than a bandaid with weak adhesive.

[soongsc]

True, it is not the best solution.

[Lynn Olson]

An apology to the forum is due - I was in kind of a sour mood when I made the don't-like-music-servers post (coming down with a cold, and now on the path to feeling better). I've been using computers for a while, going back using RSTS on DEC PDP machine at Tektronix in 1979, using the ARPANET at Tek on a DEC Vax 11/780 running BSD Unix in 1985, and assorted CP/M, DOS, Windows 3.11 through XP, and Mac OS 3.2 through 10.6.8.

The cute little Macbook Pro that will arrive tomorrow will be set up to dual-boot through Boot Camp into OS X (default) and a fresh new copy of Windows 7, and the audio interface will be the RME Babyface with 192/24 capability in and out. The RME Babyface looks pretty slick, with low latency, good USB drivers on both sides of the fence, and eminently suitable for both precision measurements as well as playing around with making my own recordings, should I choose to do so.

I also admit that the idea of spending hundreds of hours transferring all of my CD collection to a hard drive(s) - all in order to save a 1-minute walk to the CD shelf - seems a little odd. I like physical media; more than 2000 books, late-Sixties underground comix, 19th-century Japanese prints, about 300 original-issue 33-rpm records, and yes, about 600 CD's and the assorted DVD-A and SACD disc. I even have about 20 LaserDiscs and 20 78-rpm records.

For the more impermanent collection, iTunes is up to about 42 GB, selectively copied from my own CD's at full AIFF resolution. There are about four songs at the funky iTunes store resolution - I think of them as overpriced samplers of the real thing. They certainly sound low-fi; I don't see how people can stand the flat, dull, not-quite-stereo sound of compressed audio. (Lossy-compression algorithms squeeze out much of the reverb information, as well as losing most of the interchannel phase relationships. The more severe forms of compression are little more than steered mono.)

iTunes at full AIFF resolution sounds decent, and I'll probably be getting Pure Music as a way to play it over the big system. Pure Music and Amarra support higher resolutions, so when downloads start offering high-resolution versions of music I'd actually like to buy, well then, I'll probably get a little more serious about a high-quality music server. The suggestion to look into DIY USB-to-S/PDIF converters is interesting and appreciated, and would take advantage of the Burr-Brown based DAC I have now. I was surprised that most of my complaints about the flat dynamics and colorless sound of many modern DACs was simply the generic sound of delta-sigma conversion (as well as the built-in opamps that blight many voltage-output converters).

At the risk of steering this thread back to loudspeakers (what! sacrilege!), I partially agree with the wide-loudspeaker sentiments of the previous poster. This might seem odd coming from the designer of the ultra-narrow Ariels, but I had to use a fair amount of crossover balancing to offset the "sound" of the narrow cabinet and the MTM driver layout. What measured flat did not sound flat, and what sounds flat takes precedent over what measures flat. (Using pink-noise as a subjective reference, along with music.)

The impression of scale - which is what I think we may be discussing - does seem to be affected by the physical size of the driver in the lower-midrange and upper-bass region. The bigger drivers, well, sound bigger. Big instruments like pianos and tympani sound more real, more present, even at quiet volumes.

A wider cabinet is not quite the same thing; true, it has more bass fill, which always sounds nice, but if you really want the drums and piano to sound right-here right-now, 12" and 15" drivers do a better job. The catch is finding the good ones, but they're out there. For some funny reason, 10" drivers sit uneasily between the small-driver "audiophile" presentation and the traditional big-sound, big-speaker presentation. And front-horn-loaded Lowthers and Fostexes sound different again.

I suspect this is more than a matter of simple equalization and headroom, since the presentation quality is independent of listening level. The big speakers always sound big, no matter how quietly you listen, just as certain MC cartridge always have a distinctive tonal vividness, which is almost certainly a reflection of magnetic structure.

Line arrays of small drivers don't sound like large electrostatic panels; the spatial impression is different, and at all listening levels. They tend to sound "quick" (using audiophile terminology), but the sense of scale and dimension is different. And the systems that truly sound symphonic in scale and proportion frequently use 12 and 15-inch drivers, as well as the larger and more room-dominating bass horns.

But .... just to see things from the other side ... the Ariels (in a stereo pair, well out from the walls) sound far larger than they are. True, a single Ariel is almost unlistenable, sounding absurdly tiny and small; it's when both are lit up the room fills with sound. What's a little strange is the spatial impression is strongly amplifier-dependent; with good vacuum-tube amplifiers, the entire room is illuminated, while with even quite good transistor amps, the sound space shrinks to the area around and between the speakers. There's some kind of interrelation between a very low diffraction cabinet and the spatial location of subtle room-space impressions.

[Lynn Olson]

Listening to Eric Whiteacre's "Light and Gold" right now. Oooh, that's what I'm talking about. Very glad that nearly all of my iTunes content is uncompressed, so it sounds great on my dinky Cambridge Soundworks computer speakers, the Sennheiser HD580 on the iPod Touch, and of course, the big system.

Picking up on the last point before the posting gizmo timed-out, amplifiers differ substantially in their ability to render spatial cues - to me, this is the most immediately audible difference (along with tonality), something I hear right away. My best guess is some amplifier topologies - particularly those with the lowest distortion in the forward path - preserve very low-level reflections in the original recording, and the perceptual system of the listener translates those low-level reflections back to an impression of a performing space. If the low-level reflections are masked by distortion, or worse, program-dependent noise, then the impression of space disappears.

It should be kept in mind that so-called "noise shaping" in the digital domain isn't (uncorrelated) noise at all, but very complex program-dependent artifacts that average out to noise. It's a type of program-driven dithering, where the physical 5-bit converter uses very complex PWM patterns to approximate the smaller missing bits (the other 19 bits that aren't there in a delta-sigma converter). The extreme lengths that EES went to in their Sabre DAC made it apparent that a lot of assumptions in delta-sigma designs might not be correct.

I do know that small oversights in amplifier design can trash the spatial impression, and the same applies to loudspeakers. Audiophiles obsess over cables, but the real area that yields results is looking at the driver stage of the amplifier (how linear is it under dynamic conditions?) and hard-to-measure diffraction artifacts from the tweeter and loudspeaker cabinet. Tweeter capacitors, especially metallized-foil types, can be big offenders here, wiping out all the subtle details in the sound. (Quantum-level effects? Self-microphonics? Dunno.)