How Telarc put their 50 kHz recordings onto the CD-layer of SACDs

Rev 9/20/21

While looking into whether Telarc's early 50 kHz digital recordings have been digitally converted to CD-grade digital by using an asynchronous sample-rate converter (which became available in about 2006), I ran across a 2004 review of Telarc hybrid SACD-60634 (Saint-Saëns, Symphony No. 3 “Organ”, Eugene Ormandy, Philadelphia Orchestra) which was made by converting the 50 kHz master to DSD using a dCS 972 digital-format converter, then to analog, and then to CD-grade PCM using a custom Telarc ADC. The article indicates that they had tried using a sample-rate converter to convert directly from 50 kHz to 44.1 kHz, but that the results didn't sound very good due to problems with sample-rate converters at that time. The Wikipedia article on Soundstream provides detailed information on the digital recorder which Telarc used for making their 50 kHz recordings, which could be released as 50 kHz FLACs to minimize the number of sample-rate conversions which it would undergo in the process of being converted to analog.

I've also learned that someone developed a way to convert 48 kHz recordings to 44.1 kHz very early in the digital era, and that Decca released a lot of its 48 kHz recordings on a CD-collection known as The Decca Sound, which according to The Decca Sound by S. Andrea Sundaram, doesn't sound so great, although it's not clear exactly why.

So, there are probably other examples of early digital recordings, with sampling rates other than 44.1 kHz, which were transferred to CDs early in the CD-era. Such recordings were released as LPs, so there had to be decent DACs to convert them to analog, which could have been digitized with something like a JVC VP-900, an oversampling ADC with a 16/44.1 output, apparently introduced in 1982. So, it would have had 16-bit linearity and a linear phase characteristic, so that its recordings would have good low-level detail, good high-end detail, and good imaging, but it was expensive. In September of 1985, Apogee introduced its linear-phase aftermarket input filters for the typical early digital recorder, and recording engineers adopted them in droves as quickly as possible, so that most digital recorders soon had a clean high end and good imaging. But before then, a lot of digital recordings had poor detail and imaging.

However, mass-market CD players in general were lousy until at least 2010, due to the sound quality of low-cost audio-DAC chips. According to Benchmark's app note entitled A Look Inside the New ES9028PRO Converter Chip and the New DAC3 [November 14, 2016]:

"It has been a little over 7 years since ESS Technology introduced the revolutionary ES9018 audio D/A converter chip. This converter delivered a major improvement in audio conversion and, for 7 years, it has held its position as the highest performing audio D/A converter chip. But a new D/A chip has now claimed this top position. Curiously the successor did not come from a competing company; it came from ESS. On October 19, 2016, ESS Technology announced the all-new ES9028PRO 32-bit audio D/A converter. In our opinion, ESS is now two steps ahead of the competition!"

So, I gather that the 9018 Sabre DAC introduced in 2010 was the first really good audio-DAC chip, and it would have been too expensive at that time to put in mass-market players. But now there are many good inexpensive audio DAC-chips (although Benchmark is still partial to Sabre DACs), and Sabre DACs are appearing in low-cost players and DACs. I have a $100 2017 Nobsound Bluetooth 4.2 Lossless Player with a 9018 Sabre DAC, and it's amazing, although it's crude compared to Benchmark's DACs.

But due to piracy fears, the best versions of some albums are reserved for high-res streaming and LPs. According to various audio experts, including high-res expert Mark Waldrep, PhD (a.k.a. Dr. AIX), whose website is RealHD-audio.com, high-res recordings and LPs sound better than CDs because they're mixed and mastered better, and not because of the recording format. So, CDs supposedly could sound as good as high-res or LPs, if they were mixed and mastered as well, and the low-level detail could be kept out of the dither-region, where it is mixed with noise which is intended to mask the severe low-level distortion of 16-bit digital. There was a period in CD-history known as the "loudness wars," when CDs were recorded at the highest possible level because they would sell better, perhaps because high recording levels kept the low-level details out of the dither-region. Unfortunately, this approach required excessive compression and might have led to clipping.

LXmini speaker system

As much as I'd like an affordable version of the B&W Nautilus speakers, which cost between $60K and $90K, someone probably would have started producing such a speaker by now if they were ever going to. They aren't patented, since when B&W tried to patent the principle, they found that someone had patented it about fifty years earlier (see Musical 'mollusc' is fifty years late https://www.newscientist.com/article/mg14920123-400-musical-mollusc-is-fifty-years-late/). It seems to me that the tapered coiled transmission line for the woofer could be made inexpensively by molding it out of special plastic in the form of left and right halves, lining them with special foam, and gluing them together.

Another attractive but probably also quite expensive line of speakers is made by PMC of the UK. They've refined the traditional transmission line concept, of which their Advanced Transmission Line web-page (https://pmc-speakers.com/technology/atl) has one of the most succinct descriptions I've seen .

But there is an inexpensive speaker known as the LXmini (https://www.linkwitzlab.com/LXmini/Introduction.htm), which is based on transmission-line principles, and which according to many reviews, provides amazing sound quality for its price. This isn't surprising, considering that it was designed by the late great Siegfried Linkwitz, an electronics genius with a passion for designing the ideal loudspeaker. If you peruse his website, you'll find abundant evidence of his genius and passion for speaker-design. He had designed speakers even before the ones shown on his site, so he had quite a bit of experience, and the LXmini was his latest.

These are the best LXmini reviews I've found:

A) Stereophile article on an audio show where Linkwitz speakers were demonstrated (https://www.stereophile.com/content/rmaf-2014-reichert-sunday)

B) LXmini review from home theater reviewer (https://www.hometheatershack.com/threads/linkwitz-lab-lxmini-kit-speaker-performance-review.137434/)

C) LXC [LXmini knock-off] page (https://sites.google.com/site/cdenneler/home/lxc)

For more bass, there's the LXmini+2 system, which includes two dipole woofers, two more power amps, and a miniDSP 4x10 crossover instead of a 4x4HD. An advantage of dipole woofers is that they don't pressurize the room, at least as much, and therefore don't excite room resonances as much as typical subwoofers.

The 4x4HD crossover used with the LXmini has two digital inputs and an analog input, which is internally converted to digital. It also has a volume control which is controlled with an optional remote. Because I prefer to minimize conversions between the analog and digital realms, I had to find a replacement for my analog-output lossless player, and after considering many alternatives, decided to use one of my PCs as a music-server. This would allow me to use my collection of music files on 4GB USB flash drives, which I like because they streamline the process of de-cluttering, reorganizing, and defragmenting, and if one of them dies, I don't lose much, considering that each one costs just a few bucks.

For analog purists, there's also an LXmini analog electronic crossover, designed by Nelson Pass, who is famous for his high end audio electronics designs. It's available as a kit for about $200, which is a steal, considering its performance.

The PC has a Toslink output which I'd run to the Toslink input on the miniDSP crossover. To listen through headphones, I'd connect an FX-Audio DAC-amp to the PC. My TV would go to the miniDSP's USB input, and my old analog receiver's preamp outputs would go to the miniDSP's analog inputs, for listening to FM. But whenever I listen to FM, I'm reminded that it's become a wasteland, now that all the good music has gone to Sirius XM.

Top expert concludes that CD-grade digital is as good as we need

Mark Waldrep, PhD, a.k.a. Dr. AIX, is one of the top experts on high-res audio, if not THE top expert, as a look at his website, Real-HD Audio, should convince anyone. After decades of believing that high-res consumer recordings sound better than CDs, he has concluded that the reason that they have a reputation for superior sound quality is due to a variety of reasons, but not due to their higher resolution (see The Truth About High-Resolution Audio: Facts, Fiction and Findings). This is also the conclusion of other experts, such as Sean Olive, PhD, former president of the AES and now a director of Harman International's research division, and Goldmund Labs, where cost is no object and they would love an excuse to sell even more expensive digital gear than they already do. Dr. Waldrep conducted a survey, which might be ongoing, by releasing some of the high-res recordings which he made, and knows to be truly high-res, in both high-res and CD-grade, and having participants identify which is which by just listening to them, and fill out an on-line questionnaire. The aforementioned article contains a link to this survey.

However, the fact is that some high-res recordings, such as those available via lossless streaming, do sound better than the corresponding CD-grade recordings, although not necessarily because of the resolution. I surmise that these better-sounding recordings are streamed in high-res so that the music industry can attribute the higher quality to the higher resolution (to avoid having to explain why they're not released on CD), while preventing them from being copied digitally. The LP-version might sound better yet, because the record companies can put their best recordings on LPs without having to worry much about anyone being able to make an exact copy, because so few people can afford laser turntables, and those who do have laser TTs can be watched for indications of piracy.

I've devised a piracy-proof pay-per-play system (described in a previous post) which would allow us to download and store encrypted music files which could be decrypted only by the player with the corresponding decryption key, to avoid having to download lots of data every time the music is played (which might not be possible under all circumstances in which you would want to hear the music), while generating long-term income for the industry. The player would erase the key if any tampering were detected or if the internal batteries were allowed to drain excessively. If the key were erased, it could be replaced by a newly-generated decryption key at the factory under tight security (to avoid the risk of having to store the decryption keys anywhere besides the players), but all of the music files would have to be re-downloaded, which would act as a deterrent to tampering or letting the batteries discharge excessively. But if CD-grade files are sufficient, at least the downloads wouldn't be huge. Some people object to pay-per-play, but they could continue to use the existing inferior systems. I see it as a cost-effective alternative, because I'd be able to obtain the best recordings in digital form and avoid the expenses associated with the existing systems.