I have to say that the open baffle concept just goes against everything I would hold true from a bass reflex or sealed box. Talking about critical placement, that is the only way you can create any semblance of real bass from an open back. I would expect some of the worst comb filter effects from any open backed cabinet in the bass and mid-bass regions.
John's compression test got me to thinking about the energy in different musical registers. There are some graphs around the web, but I decided to DIY mine. You might find them interesting.
Spectrum of Musical Genres
Spectrum of Musical Genres
R2R DAC > Class A amp > BTA-like speakers
This post is mostly directed to Lynn Olson (the originator of this thread)...
I have gradually transitioned towards a system that falls in line with many of the recommendations provided by Lynn over the years.
At present, the backbone of my listening chain (I will intentionally just mention the core technologies employed and withold the brand names for the time being) is composed of :
1) A network player reading a mix of 16/44, 24/88, 24/96, 24/192 FLAC and 2.8M & 5.6M DSD files from a NAS drive.
This device uses a PCM1795 Delta-Sigma DAC chip which switches to a "direct DSD" mode when decoding DSD files.
2) An conventional single-ended analog solid state preamp and a push-pull MOSFET power amp that is biased heavily into class A (up to 45W per channel into 8 Ohms, beyond which it transitions to class AB and tops out at 100W per channel into 8 Ohms)
3) A "Beyond-the-Ariel" inspired loudspeaker system consisting of:
- a low-mechanical loss and medium-high efficiency 15" hybrid cone woofer,
housed in an overdamped bass-reflex box made of thick baltic birch multiply,
- a large-format compression driver with Beryllium diaphragm,
- a ~300Hz hypex radial horn, carved out of a solid block of baltic birch plywood and with a throat adapter that perfectly matches the geometry and exit angle of the driver's internal throat and smoothly transitions to the radial portion of the horn,
- a horn-loaded ring radiator supertweeter used to improve dispersion above 6kHz,
- a passive crossover that ensures almost perfect phase tracking between the woofer and compression driver over a wide range of frequencies around the first crossover point of 600Hz.
Given that while putting together this system, each step along the way has been an improvement, I have come to hold in higher and higher regard Lynn's opinions.
And having read his musings on the virtues of R2R multi-bit DACs, I am now intrigued by the idea of testing this out for myself, by inserting one such DAC in between 1) and 2) above (obviously, this would be for the reproduction of the PCM FLACs only; the network player's internal PCM1795 would still be called upon to play DSD files).
The only problem with this is that there are VERY FEW multi-bit DACs on the market today (MSB Technology, AMR, Aqua, ...?) and most of them are very expensive indeed.
Others are less expensive (Schiit, Audio G-D, Soekris, ...?), but somehow I am a little sceptical about them (the top-of-the-line Schiit unit, for example, was reviewed by Stereophile, and the measurements don't look like anything to write home about...).
Then, browsing the internet, I have also come across one recently-discontinued DAC that can still be found new in Japan: the RATOC Audio Lab RAL-24192DM1.
This is intriguing to me because it is a dual mono unit that uses two PCM1704 24-bit chips that are highly regarded by Lynn, it looks like it is very cleanly and neatly put together, and it does not cost the Earth.
It also received a rave review in a German Hi-Fi magazine (for what it's worth...)
I'm currently debating whether to pull the trigger or not...
And of course, Lynn: any further thoughts you might have to share would be greatly appreciated!
Cheers,
Marco
This post is mostly directed to Lynn Olson (the originator of this thread)...
I have gradually transitioned towards a system that falls in line with many of the recommendations provided by Lynn over the years.
At present, the backbone of my listening chain (I will intentionally just mention the core technologies employed and withold the brand names for the time being) is composed of :
1) A network player reading a mix of 16/44, 24/88, 24/96, 24/192 FLAC and 2.8M & 5.6M DSD files from a NAS drive.
This device uses a PCM1795 Delta-Sigma DAC chip which switches to a "direct DSD" mode when decoding DSD files.
2) An conventional single-ended analog solid state preamp and a push-pull MOSFET power amp that is biased heavily into class A (up to 45W per channel into 8 Ohms, beyond which it transitions to class AB and tops out at 100W per channel into 8 Ohms)
3) A "Beyond-the-Ariel" inspired loudspeaker system consisting of:
- a low-mechanical loss and medium-high efficiency 15" hybrid cone woofer,
housed in an overdamped bass-reflex box made of thick baltic birch multiply,
- a large-format compression driver with Beryllium diaphragm,
- a ~300Hz hypex radial horn, carved out of a solid block of baltic birch plywood and with a throat adapter that perfectly matches the geometry and exit angle of the driver's internal throat and smoothly transitions to the radial portion of the horn,
- a horn-loaded ring radiator supertweeter used to improve dispersion above 6kHz,
- a passive crossover that ensures almost perfect phase tracking between the woofer and compression driver over a wide range of frequencies around the first crossover point of 600Hz.
Given that while putting together this system, each step along the way has been an improvement, I have come to hold in higher and higher regard Lynn's opinions.
And having read his musings on the virtues of R2R multi-bit DACs, I am now intrigued by the idea of testing this out for myself, by inserting one such DAC in between 1) and 2) above (obviously, this would be for the reproduction of the PCM FLACs only; the network player's internal PCM1795 would still be called upon to play DSD files).
The only problem with this is that there are VERY FEW multi-bit DACs on the market today (MSB Technology, AMR, Aqua, ...?) and most of them are very expensive indeed.
Others are less expensive (Schiit, Audio G-D, Soekris, ...?), but somehow I am a little sceptical about them (the top-of-the-line Schiit unit, for example, was reviewed by Stereophile, and the measurements don't look like anything to write home about...).
Then, browsing the internet, I have also come across one recently-discontinued DAC that can still be found new in Japan: the RATOC Audio Lab RAL-24192DM1.
This is intriguing to me because it is a dual mono unit that uses two PCM1704 24-bit chips that are highly regarded by Lynn, it looks like it is very cleanly and neatly put together, and it does not cost the Earth.
It also received a rave review in a German Hi-Fi magazine (for what it's worth...)
I'm currently debating whether to pull the trigger or not...
And of course, Lynn: any further thoughts you might have to share would be greatly appreciated!
Cheers,
Marco
Last edited:
In particular, what do you make of these quotes?
"RAL-24192DM1 doesn't use DF1704 8 x Oversampling and Digital Filter chip. It's Internal firmware control PCM1704 directly and bring out pure and natural sound from the DAC"
and
"RAL-24192DM1 doesn't use DIR(Digital Interface Receiver). RAL-24192DM1 includes software S/PDIF decoder, End Point(memory buffer) for 192 Frames and Timing/Control signal generator for Multi-bit DAC(PCM1704).
The timing/Control signal generator circuit is fully driven by dedicated two independent X-Tal Oscillator modules (24.576MHz and 22.5792MHz). This circuit provides low jitter Bit-Clock and it drives D/A converter accurately. "
Thank you in advance, and sorry everyone for the slightly OT tangent.
Marco
Well, the Ratoc DAC looks interesting; my "other" R2R Dac is from Audio-GD in Hong Kong and features the PCM-1704. My primary DAC is a 15-year-old Monarchy that uses the long-discontinued PCM-63K and a discrete all-tube analog signal path.
The other DAC I've auditioned in my system that sounded pretty impressive, maybe even better than the Monarchy, was Gary Dahl's Canadian-made exaSound DAC. For reasons that aren't clear to me, the exaSound doesn't sound like other DACs that use the ESS converters, so they must be doing something different. At any rate, it directly drove the Karna amplifiers in balanced mode with no trouble.
A few pages back, I mentioned that the crossovers for the Ariels feature sets of DPDT switches that select between a Marantz home-theater amplifier (MA8003) and the pair of Karna amplifiers. The grounds, as well as the signal, are switched, so there is never any ground connection between the entirely separate HT system and the all-triode, zero-feedback 2-channel system. The quick switchover between the Marantz HT and the exotic triode system keeps things honest; the specs of the THX-certified HT system are pretty decent, and Marantz has dialled-in the subjective sound of their electronics so they don't sound as shrill and edgy as the majority of HT equipment. I could describe them as "smooth" and easy-to-listen-to, and not a bad choice for folks who don't care for the hassles of vacuum tubes, but don't care for the edge and grain of many transistor amps.
If I ever upgraded the 5-channel system, I'd pick the Bryston multichannel amp, which sounded really good a few years back at the RMAF show. At any rate, I have two completely different systems to drive the Ariels, and both sound pretty decent. One thing I have discovered is that as efficiency goes up, it acts like a magnifying lens on the rest of the signal chain, whether solid-state or vacuum-tube.
The other DAC I've auditioned in my system that sounded pretty impressive, maybe even better than the Monarchy, was Gary Dahl's Canadian-made exaSound DAC. For reasons that aren't clear to me, the exaSound doesn't sound like other DACs that use the ESS converters, so they must be doing something different. At any rate, it directly drove the Karna amplifiers in balanced mode with no trouble.
A few pages back, I mentioned that the crossovers for the Ariels feature sets of DPDT switches that select between a Marantz home-theater amplifier (MA8003) and the pair of Karna amplifiers. The grounds, as well as the signal, are switched, so there is never any ground connection between the entirely separate HT system and the all-triode, zero-feedback 2-channel system. The quick switchover between the Marantz HT and the exotic triode system keeps things honest; the specs of the THX-certified HT system are pretty decent, and Marantz has dialled-in the subjective sound of their electronics so they don't sound as shrill and edgy as the majority of HT equipment. I could describe them as "smooth" and easy-to-listen-to, and not a bad choice for folks who don't care for the hassles of vacuum tubes, but don't care for the edge and grain of many transistor amps.
If I ever upgraded the 5-channel system, I'd pick the Bryston multichannel amp, which sounded really good a few years back at the RMAF show. At any rate, I have two completely different systems to drive the Ariels, and both sound pretty decent. One thing I have discovered is that as efficiency goes up, it acts like a magnifying lens on the rest of the signal chain, whether solid-state or vacuum-tube.
Last edited:
I owned one, sold it and bought another one
Hard to describe the sound of it but I'll try- it was quite transparent with a bit of "something" that I can't quite explain that made me enjoy classical music more and want to explore more performances than simply think about how good the digital I was listening to was. I also noticed myself listening to historical recordings that sound harsh from 78 rpm surface noise at a comfortable volume, whereas before I would turn down the volume. I'd encourage anyone interested to have a listen to it or the multibit Gungnir, which I have not heard.
I sold it because I initially thought it was overpriced and I really didn't like that the manufacturer recommends leaving it on all the time due to the DACs used since I experienced occasional power outages and was worried about it getting damaged. It was afterwards that I realized I missed the sound of it.
In my experience the correlation is there but people insist on measuring the wrong thing. THD is the obvious one, great for hiding the nasties of GNFB, the poor handling of DAC RF etc, the list goes on...
I doubt Stereophile used any relevant measurements, which IMO is the reason commercial HiFi has ended up clinging to the lifeboat of video surround sound.
Sticking with irrelevant measurements over the decades combined with the ruthless destruction of quality in today's recorded music has resulted in todays HiFi store's rather expensive, poor sounding gear having in my experience a small 'polite' sound slightly below the quality that a cheap Panasonic system generated in the 1970s.
I guess the reviewer connected it up with expensive mains cable, cryogenic beeswax filled super fuses, pricey HiFi USB and optical cables , but may have put the nice sound down to his valuable speaker cables on Mpingo cradles
DAC development itself seems largely driven by cost these days, perhaps DIY DACs with ladders of resistors, cermet pots driven by MOSFETs or TTL chips are not too far off!
Being a owner of both Soekris and a homemade PCM1704 DAC (see details below), I must say that I prefer the PCM over the Soekris. PCM is less "electronic" in character. But the difference is far less than the strange high-order distortions suggest to be; this HOD varies with level, so I suppose to be related about various shift registers internal rds(on) mismatch. So is a non-monothonic behaviour.
And the R-2R network in Soekris needs a PERFECT voltage source to operate correctly... is not a thing easy to make. Various pages about this reference supply are in vogue some time ago.
***My homemade PCM DAC have:
1 - a CS8416 S-PDIF receiver. Is better than USB, because every A/V equipment have it; not only PC, and one can use optical for insulating dirty devices like TV or BlyRay players;
2 - a homemade secondary PLL with Tent Labs VCXO;
3 - a DF1704 digital filter;
4 - only one PCM1704 for channel;
5 - nanocrystalline core I/V transformer (1:10 voltage ratio). Nominal load seen by PCM is 10R, including a supplemental 75R direct at PCM I-out pin, for not having open load for PCM at some HF frequency;
6 - Aikido with E88CC for final gain.
7 - all stages supplies are shunt regulated.
Sounds really good. I expected that my new experiences with Soekris etc will be more musical sounding, but the converse is true... the PCM DAC still stands in my no.1 system.
But interesting is that the sound descriptions are more or less like Linn talked about: less electronic, more natural etc etc...
Sorry about this far high OT level
but since I have both DACs...(by the way, the PCM measures far better in high order distortion than Soekris, and my Soekris have 0.027THD with HOD vs 0.008% with almost none HOD from my PCM DAC)
Last edited:
Interesting, thanks!
In the mean time, I have gone ahead and ordered the RATOC dac.
BTW, the RATOC reportedly "doesn't use DF1704 8 x Oversampling and Digital Filter chip" but instead uses internal firmware to control the PCM1704 directly.
Not sure how (if) this may affect the overall results. Time will tell, and I will report in due course.
Marco
Interestingly, the RATOC dac "doesn't use DIR (Digital Interface Receiver)" either, instead relying on a "software S/PDIF decoder".
I wonder about the pros and cons of these choices.
Marco
That's a telling endorsement.
If that was referring to my comment I can see that
Given how much these DACs cost I would encourage anyone hear it for themself before buying, there is a lot of hype around them and I was naturally pretty skeptical of them as well.
With the discussion of PCM1704 and DIY I have a friend that knows digital design very well and he says the 1704 is very difficult to implement and doesn't think that a simple IV like resistor or transformer is the way to go. YMMV as he is very engineering minded. According to him one of the best designed PCM1704 DACs he had seen was the Assemblage DAC3.1 which was sold and made by PartsConnexion. That schematic might be on the net.
This guy in Japan seems to have done something of this sort, but using a classic Altec sectoral horn on top of the 604:
YouTube
Marco
Not sure the free-air impedance measurement of the Radian is that informative. The driver is not designed to be used in free air, without a horn. I'm guessing the minor peak around 2.1 kHz is the result of a tube resonance (between the opening of the driver and the back of the phase-plug). A more useful measurement would be into a plane-wave tube, which are designed to avoid pipe modes and horn standing waves.
I'm curious about the 80x50 constant-directivity horn. How big is it? What is the cutoff frequency? That will have a strong effect on the impedance measurement.
Last edited:
I generally pefer to have an impedance curve of the driver without any horn assembled because of one reason: if someday I want to change the diapragm I have this curve as reference. Together with CLIO pocket and the driver positioned on the table is very easy to do an impedance measurement and in my opinion the only way to exactly align the diaphragm. I do not know the inner parts of this driver and how the diapragm will to some degree self align but I have this curve now.
For my Altec similar FC compression drivers aligning the diaphragm is really a hard job. These do not self align and changes for a fraction of an mm are relevant! Hearing a clean sinus does not mean the the impedance curve is perfect...
Btw, the peak you mentioned around 2 kHz is nearly the same as radian themselve measurent for their recent datasheet on the specified horn. So, it must be related to the driver inner resonaces which can be damped with a horn attached.
The CD horn I used is an OEM version but I assume that it is similar to the Beyma TD385 which has an cut-off around 800 Hz.
I hopefully will do some other measurements in the next months if I receive an tractrix test horn but cannot promise this.
Last edited: