Prana-Bindu Posted November 12, 2001 Share Posted November 12, 2001 I know Spider is into this stuff, so if you see him anywhere, let him know about this thread. O.K.: I just got a Perpetual Technologies P-3A DAC. The plan is to get the P-1A digital correction unit when the digital floats up again as the weakest part of the system (after I upgrade my speakers). As the P-1A does not accept the Toslink output of my cd-player, I plan to get a transport at the same time. I want to minimize jitter problems, so I'd like to get a transport with a I2S output to feed into the P-1A's I2S input. In my search for compatible transports, I discovered that there's now I2S Enhanced, but I can't find any information that compares it to regular I2S. What gives? If there's anyone out there that knows this junk, please help me out. I'm concerned about compatibility issues down the road, as all new transports seem to only include I2S Enhanced. Does regular I2S include in its transmission the biphase-mark encoded signal? Is regular I2S capable of the as-of-this-time theoretical Level 1 performance (where the receiving unit's clock is transferred back to the sending unit to use as its master clock)? Is the only difference between the two formats that the Enhanced version has sturdier termination? Any help would be greatly appreciated. Spider, if you're out there, I know you're working on a way to get the above-mentioned Level 1 performance included into your system. Where are you coming from, SpiderMan? Nobody knows who you are.... ------------------ May the bridges we burn light our way.... Quote Link to comment Share on other sites More sharing options...
Ray Garrison Posted November 12, 2001 Share Posted November 12, 2001 Quoting from a Stereophile review of the Sonic Frontiers CD transport and D/A converters... Two Don't Always Tango One of the key elements of Sonic Frontiers' new Series 3 digital components is the inclusion of an UltraAnalog-designed I2S-Enhanced interface. The original I2S (Inter-IC-Sound) bus found in nearly all CD players provides discrete pathways for the master, word, and bit clocks, as well as audio data signals as they travel between processing sections within a player, and eliminates the need for the transmission and receiver circuitry mandated by the AES/EBU or S/PDIF transmission schemes to connect two or more digital components, which is prone to jitter due to the fact that the word clock is embedded within the data. The recent offerings of I2S interface options for two-box systems are derivations of this bus applied externally for inter-component, rather than inter-IC, communication. For example, the critical reference clocks and audio data each travel from a transport to a digital processor intact, on discrete lines, without suffering the inevitable degradation caused by bi-phase encoding, transmission, and subsequent decoding of the combined dock and data signals, as occurs with traditional interfaces. The result is a dramatic reduction in interface jitter. Good examples of external I2S inter-faces include the I2S-E and the original 13W3-I2S standard developed by Kevin Halverson of Muse, and now offered to other manufacturers by Digital Axiom Corporation as a ready-made solution in the form of drop-in modules. By using either scheme, the relative lack of interface jitter allows a designer to fully realize all the advantages of two-box construction. These include the isolation of critical decoding circuitry and the analog output stage from interaction with transport servos and unwanted power-supply coupling. For the past year and a half I've listened extensively to systems equipped with I2S interfaces using 13W3 hardware. These include two separate Muse systems, and now the Sonic Frontiers pair. In each case I've heard a significant enhancement of low-level detail and dynamic contrast, better focus and transient definition, and a notable increase in the perceived effortlessness of expression compared with any version of an AES/EBU or S/PDIF connection I listened to - so far, all great news. However, as mentioned in the body of the review, there's a compatibility issue prospective buyers must be aware of. UltraAnalog branched off from the original 13W3-I2S specification to design their own version, as was certainly their prerogative,. However, UAs employment of the same 13W3 cable and connectors (probably because they're ideally suited to the task) ensured the kind of consumer confusion and incompatibility between the 13W3-I2S and I2S-E schemes that I experienced. I don't have the space here to explore all the technical details distinguishing the two now-competing versions of 13W3-based I2S designs. Instead, I'll look at my experience with the Sonic Frontiers version, and a few practical differences of interest to the end user. I'll take a closer look at the technical merits of the 13W3-I2S interface in an upcoming review of the Muse Model 8 and 296 DVD/CD system. I went through a considerable "debugging" period with the I2S-E system in the Processor 3. In December '97 I was sent an SFT-1 transport equipped with the interface for use with the P-3. However, it wasn't yet ready for prime time - I couldn't get a clean signal lock. Sending the unit back to the factory failed to fix the problem, but by February '98 the bug had been fixed, and I installed a factory-supplied modification to the I2S-E circuitry within the P-3. Since then, the interface has worked great. In May, Murphy's law returned in the form of a bug unrelated to the I2S issue: The SFT-l's front-panel controls began to consistently lock up, forcing me to use another transport until the T-3 arrived. Since then, I've not had a lick of trouble with the T-3 or the P-3. As no digital rigs contain examples of the I2S-E and 13W3-I2S interfaces, is impossible to directly compare the audible differences, if any, between them. Despite any claims to the contrary, when both systems are properly implemented, using their respective highest levels (in which the master clock is located in the processor and the transport is the slave), there should be little if any theoretical or practical sonic differences resulting from interface issues. In any event, what really matters to the listener is the overall experience from a given complete system, as discussed in the body of the review. Curiously, the version of I2S found in the expensive SF gear uses the lower of two performance levels offered in the I2S-E specs; ie, with a transport-based clock and the processor in slave mode. Chris Johnson said they felt this was sufficient for the demands of CD, and that they would implement "Level One" - with the master clock in the processor - when the P-3 is updated for 24-bit/96kHz capability. Regardless, you can rest assured that both the I2S-E and 13W3-I2S schemes work extremely well with CDs. However, since the Muse system plays DVDs and AADs as well, the 13W3-I2S specs have been optimized to address a number of that format's performance and legal requirements. The current specs of the I2S-E interface will need updating to address these issues prior to use in DVD-based systems. Just keep in mind that a 13W3-I2S interface found in other digital components will not work with Sonic Frontiers' new gear. Here's hoping they'll find a way to include compatibility in future iterations of their 13W3-based I2S interface. - Shannon Dickson ------------------ Music is art Audio is engineering Ray's Music System Quote Link to comment Share on other sites More sharing options...
Prana-Bindu Posted November 12, 2001 Author Share Posted November 12, 2001 Ray, as always, you rock! It's helpful to know that the I2S-E design is still new. I wish Shannon had made it clear whether the standard I2S design supported Level 1 operation (using the transport as a slave to the DAC's master clock). However, I believe the P-1A has to reclock the signal anyway when it converts and interpolates into 24/96, so perhaps it's a non-issue for me. I'll do a search under 13W3 and update as it becomes relevant. Thank you Ray! ------------------ May the bridges we burn light our way.... Quote Link to comment Share on other sites More sharing options...
soundog Posted November 17, 2001 Share Posted November 17, 2001 I solved the problem with the lack of an optical input in the P1A by using an MSB Digital Director which provides optical to coaxial conversion as well as "advanced jittter reduction and digital data regeneration" (it is my understanding and experience that jitter reduction is cumulative). While the I2S interconnect system seems to offer superior performance, it seems too new and "rare" to be practical (and seems to be found on only very expensive transports). I had to make a choice: did I value ease of access more than the possible improved sonic qualities of the I2S (plus some doubt that the transport makes that much of a difference vs the quality of the DAC and the presence of jitter reduction when dealing with a digital stream}. Since I already had a number of Sony CD changers and I had the computerized Nirvis CD jukebox system and interface, did I want to go back to single play transports and give up the ease and fun of being able to search for and program CD material as well as download album covers, and CD contents from the internet and then be able to project them on my TV monitor? Since I had begun to augment my CD collection with copies of borrowed CDs (which are virtually indistinguishable from the originals), this capability takes on addded value. Never-the-less I am open to learning more and welcome any further enlightenment from any and all. I plan to purchase a SACD player this weekend and will be able to compare and report on the difference between digitally enhanced (24/96) regular CDs (vis-a-vis the Perpetual Tech devices) and the new SACDs (of course the connectors there are analog to prevent copying). To be honest, I'm not absolutly sure what that Stereophile article means. (Ray or anyone - can you translate?) As with life in general, it's all a great adventure and learning experience. It's sure great to share and learn from you guys. ------------------ Soundog's HT Systems This message has been edited by soundog on 11-17-2001 at 05:01 AM Quote Link to comment Share on other sites More sharing options...
Prana-Bindu Posted November 18, 2001 Author Share Posted November 18, 2001 I can definitely understand the power of the convenience and options that come with running multi-disc players. In my search for digital upgrades, I started by looking at multi-disc players, but I wanted to hear cd's like I heard them through some nice single-disc players (Arcam stuff, e.g.). Though my original purpose was to reach a higher level of detail, dynamic range and darkness behind the music, I couldn't swallow giving up the ability to play six discs on random. I can only imagine the features your digital system must have, Soundog.... However, I see my musical interests placing a higher value on high fidelity and 2-channel lovelies like imaging, soundstage, "airiness", etc. the more I get into this crap. Such musical qualities vary from digital system to digital system (I suspect) in very subtle ways. My tastes are going in a direction that makes those differences very valuable. For example, the things the P-3A is doing to stuff like Sarah McLaughlin (spelling?), Cowboy Junkies, and (believe it or not) Guided by Voices are so wonderful, I can hardly wait to add the P-1A's influence; however, the things it's done to Jane's Addiction, Soundgarden, Butthole Surfers and Guided by Voices (heh, heh -- their harder stuff, I mean) I don't believe complement the music as well as they do the other music mentioned. Of course, the problem may be that the P-3A has brought such dynamic music to the limits of my RF-3's and VTL integrated amp (the horns seem more dynamic than the woofers, and the amp may not have the juice to keep up). I still don't understand why an s/pdif cable like Toslink or coaxial wire HAS to carry a clocked signal. Why can't the output be modified to carry only the disc data bit-by-bit for the DAC/P-1A/whatever to clock first? All that I read states that the reduction of jitter resulting from avoiding the marriage between the clock signal and the data stream inherent in s/pdif makes a big difference (relatively, of course). "Why clock it all all in the transport?" says I.... Nevertheless, I'm encouraged to get a transport with I2S output: I want to maximize my CD's content with the stuff added by the P-1A, the P-1A has a I2S input, I value highly the benefits of reduced jitter, s/pdif encoding is identified a major source of jitter. I sure hope SACD's end up victorious in the pending battle. Of course, my hope is not coupled with the financial commitment to start buying SACD's and players (rainy day supporter, I guess).... I've read that SACD is better suited to reach 192khz sampling than DVD-A. And isn't DVD-A made for multi-channel music anyway? The only-apparent problems associated with getting multi-channel music right are enough to hope the idea never gets anywhere near off the ground. It looks like Shannon Dickson is discussing only two versions of the I2S Enhanced interface: the one invented by Sonic Frontiers and the original 13W3 cable version (such cable has been used extensively by the computer industry). The Perpetual Technologies doesn't use either of these two; it uses a 5-pin mini-DIN interface, whereas the 13W3-I2S interface uses a thin, wide termination with like 10 pins (I think). The difference appears to be that both I2S-E and 13W3-I2S allow the use of the DAC clock as the master clock (forcing the transport to clock to be the one fed into it by the DAC), but the PTech mini-DIN doesn't allow for the DAC clock to be sent back to the transport. I don't think there are any transports that take advantage of the DAC's clock (Level 1 performance) at this point. The only possibility for jitter that I see in I2S Level II performance (DAC's clock is slave to transport's clock), is in the fact that the DAC would have to synchronize it's clock to the transport's clock; the idea being that this synchronization would end up being imperfect, since the clock cycles that we're talking about are as close as you can get to infinitesimal. Even so, however, what difference would it make to have the transport clock be slave to the DAC clock if the same synchronization has to occur at the transport clock (with some kind of phase lock loop, I think). I know Spider knows quite a bit about this clock issue. Where is he? Nobody knows who you are.... ------------------ May the bridges we burn light our way.... Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.