Ayre wants $1.5K for DSD'ed QB-9

[tailspn]

But, even your technical explanation has flaws: the noise inherent in DSD is a technical flaw, for example, which you appear to dismiss.

 

I am not sure why anyone would want to start these types of compaprisons with a PCM source (the confusingly named "DXD") although, I suspect that "DXD" when done well is transparent anyway... but for the sake of keeping things fair for DSD, I would prefer a native DSD file be used for comparisons.

 

Sorry, I didn't intend to ignore the inherent noise in DSD. But it's a playback issue, not record, and we're talking about a analog to digital converter. The quantization noise at 64fs DSD in the audio band rises to about -140dB in a well designed ADC at 20KHz, and is inconsequential IMO. It rises rapidly above the audio band to about -55db at 100KHz, or about the level or broadband tape noise in a second generation tape. That shifted noise however can play havoc in some downstream components, like amplifiers, and is not inconsequential. But it is fixable by DSD upsampling to 128fs, or even better, 256fs. This has been shown to have no negative effects within the audio band, but does shift the noise an octave or two higher, where it can be filtered with gentler post DAC filters.

 

I'm also sorry if I was not clear on the source of my comparisons. The source was either 256fs DSD, or 64fs DSD, non post processed session files, downsampled (or not in the case of 64fs DSD) to both 64fs DSD, and 352.8KHz PCM (DXD). This was done on a Merging Technologies Horus with Pyramix 8.1.2 software. It's high end studio stuff. In my experience, DXD PCM was not transparent, and was easily distinguishable from its DSD sibling.

 

But I do agree that the end result is in the ears of the listener, and for those interested, they should do their own comparisons. That's after all, the fun of this interest!

[firedog]

Charles If that is how you feel about DSD then why are you upgrading a DAC to support it ? just wondering. But of course incoming revenue comes to mind.

 

 

Why do people keep throwing this back at Charles? He said he thinks DSD sounds good, even better than RB; just not better than well done hi-res PCM. So why not put it in his products? If his customers want it, and he doesn't think it's BAD sounding, why shouldn't it be there?

 

Part of his job as a manufacturer is to make products that his customers want. He isn't engaging in a hobby just for fun - He's trying to keep a business going. Today the DAC market is clearly in a place where DSD playback has become a feature many customers demand in a DAC - and they will buy competing product if a given product doesn't have it.

 

Under these circumstances Ayre would be foolish not to include it. The only exception to this would be if they felt that DSD was so inferior that including it harmed the reputation of their products. But this clearly isn't the case.

 

The sarcasm about "incoming revenue" is out of place and divorced from reality of the marketplace.

[barrows]

Tom: Well hmmm. I think, in a sense, we are talking both about ADC and DAC when we are talking about how something "sounds". After all, we cannot "listen" to the output of an ADC, and the only way to know how something sounds is by listening.

Your point that any (digital) conversion produces errors may be technically correct, but it is important to point out, that many of the errors, in many conversions, will be entirely inaudible in any case (but not every case) as they will be entirely masked in the dither or noise shaped up and away. For a specific example, ESS claims that all error products of D/A processing in the 9018 chip are below -140 dB, I think we can agree that no one is ever going to hear that.

Yes, DSD noise is a playback issue, and will be system dependent, and 2x and 4x DSD likely makes it entirely a non issue, but since you were talking from a mostly technical perspective, I thought it bears analysis.

What I disagree with is that DSD (and I am keeping the discussion to 1x rates here, as this is what is available almost entirley) is inherently superior sounding to well done PCM, and it is in that statement I tend to agree with Mr Hansen. I would not go so far as to say PCM is superior though, I do not have enough experience to make that call either way for myself.

Then we get back to the practical world. I find it very unlikely that DSD will be embraced by the recording industry at large, sure, we will have the Channel Classics, and the Analogue Productions of the world producing a few recordings for audiophiles, but the expense of the hardware, and the difficulty in editing makes widespread acceptance of DSD impractical. Of course, since we do have Channel Classics, etc, it is nice to be able to play back DSD on hardware which also makes the most of all the PCM recordings available.

[mav52]

Why do people keep throwing this back at Charles? He said he thinks DSD sounds good, even better than RB; just not better than well done hi-res PCM. So why not put it in his products? If his customers want it, and he doesn't think it's BAD sounding, why shouldn't it be there?

 

Part of his job as a manufacturer is to make products that his customers want. He isn't engaging in a hobby just for fun - He's trying to keep a business going. Today the DAC market is clearly in a place where DSD playback has become a feature many customers demand in a DAC - and they will buy competing product if a given product doesn't have it.

 

Under these circumstances Ayre would be foolish not to include it. The only exception to this would be if they felt that DSD was so inferior that including it harmed the reputation of their products. But this clearly isn't the case.

 

The sarcasm about "incoming revenue" is out of place and divorced from reality of the marketplace.

 

It appears there was so confusion in the post before mine, get a grip, it's a forum, not a town hall meeting. I happen to be interested in Ayres's new DAC since I recently sold my Mytek..

[vortecjr]

Jesus: Sorry for any confusions, I was responding to Tom.

And, Tom: sorry for my confusion, I re-read your post and realized you were comparing native DXD to native DSD, but it was too late for me to edit my previous post.

 

And, why wouldn't Ayre include DSD playback (and record on the QA-9) capabilities on the QB-9? It is easy to do so, and Mr Hansen himself agrees that there are some titles where DSD is the best sounding version available. The fact is that adding DSD capability adds no significant cost to the QB-9 (as long as you consider that the upgrade improves performance across the board, and is not just to add DSD capability) so of course it makes sense to allow for DSD playback capability.

I suggest that anyone who is curious to learn more about what Mr Hansen is saying, and the design approach of his company, read the Stereophile review of the QA-9 ADC he has linked below-there are many details there which might illuminate how PCM can achieve improved performance.

 

That's funny!

 

Jesus R

[Hiro]

Sorry, I didn't intend to ignore the inherent noise in DSD. But it's a playback issue

 

I would say, it's a playback non-issue, as every 64fs DSD recording can be upsampled to 128fs or 256fs DSD in the computer environment.

 

And then, you also have the newly enabled double-rate and quad DSD recording capabilities.

 

The future for the DSD format has never looked brighter.....

 

[tailspn]

I would say, it's a playback non-issue, as every 64fs DSD recording can be upsampled to 128fs or 256fs DSD in the computer environment.

 

And then, you also have the newly enabled double-rate and quad DSD recording capabilities.

 

The future for the DSD format has never looked brighter.....

 

I completely agree.

 

The 128fs and 256fs DSD record rates however do not have the performance increase you may be anticipating. The faster rates will significantly aid sound quality in the post production process, where mixing and layering combine the noise and quantization errors from each channel or track. This is a big advantage to studios using those techniques (like all), and opens the avenue to more pop music studio projects using DSD. For field recording of classical music, especially where the balance and mix are done in analog at the session, it's the quality of the ADC, and not its sampling rate that determine sound quality. In fact, from a recording fidelity point of view, there's evidence to suggest that because of sampling settling time issues in the A/D, that it's better to use a very high quality ADC at 64fs, and upsample digitally either at playback, or in post production.

 

With the introduction of hardware for faster DSD record sampling rates, and the eventual introduction of post processing tools in the multi-bit Delta-Sigma Modulator formats like the Sonoma, I believe DSD like (I wish there was an acronym for mulit-bit data streams) will overtake PCM as the studio production format. Please remember, DSD is just the delivery format, not a production format.

[barrows]

I would say, it's a playback non-issue, as every 64fs DSD recording can be upsampled to 128fs or 256fs DSD in the computer environment.

 

And then, you also have the newly enabled double-rate and quad DSD recording capabilities.

 

The future for the DSD format has never looked brighter.....

 

[ATTACH=CONFIG]5751[/ATTACH]

 

Hiro: I am curious, do you honestly believe that DSD will become a standard recording format rather than just a niche format used by a few small audiophile labels? Personally, while I certainly enjoy the small handful of DSD titles available, I listen to many types of music, and almost all of that music is recorded in PCM (rock, world, jazz, american roots/country, etc). I am certainly not going to choose music to listen by virtue of its format (PCM or DSD). You really think we are going to see a time where DSD becomes a standard recording format for these genres?

 

To be clear, as far as playback is concerned, I do not actually consider the HF noise of single rate DSD a real world problem, as with good hardware, it will be a non-issue from a listening perspective. I brought up this noise, as Tom was speaking from a technical ("accuracy" rather than listening) perspective and from that perspective the HF noise of DSD is a problem.

[gumby]

I woke up on Saturday morning and paid a visit to the LINNForums for the first time in about month. I found a recent post by LINN staff taking the position that DSD is aterrible idea in 2013. I then found my wayto this forum, where a representative of another audio equipment manufacturer, Ayre’sCharles Hansen was spending a great deal of time explaining this in moredetail. Mr. Hansen did a great job of carefullydefining terminology and clarifying a bunch of technical issues that were neverentirely clear to me and I really appreciated that. Moreimportantly, to his credit, he took on the broader issues behind the post onLINN, which is the added cost to the industry, and the consumer, of carryingtwo formats that for all practical purposes can produce a perceptively equalproduct.

Charles also argues that more of the major record labels arelikely to start providing all of us higher quality music files if the industry couldstandardize on one format, which he argues would need to be PCM because iseasier to manipulate. I presume thebasis of that assertion must be related to the costs of providing the music inboth formats and/or having to choose one type of packaging only to find it islater not supported by the playback devices and/or because the cost of getting theperformance right at the recording session is too high and/or the PCM files aremore likely to also be sold by iTunes or an Amazon, increasing the availablemarket beyond the audiophile community.

Well, it seems that Ayre have decided that the cost ofproviding playback of either format can be supported by the market, otherwisewhy have they included DSD playback in an upgrade to their DAC after over tenyears of debate on this subject? Charlesmade it clear the cost to incorporate 64fs DSD was an inconsequential portionof the cost of the upgrade Ayre was providing. Adding DSD playback option was the initial impetus for the upgrade, butAyre took the opportunity to incorporate other benefits. He also decided that another $300, on top of thatcost, was too steep a price for those who want 128fs DSD, but had purchased a designvariant of his product that required more work to upgrade. Did I get that right?

If I have invested 10’s of thousands of dollars in my musicsystem, and thousands of dollars in music, another $300 doesn’t seem too bad tome. People have more money invested incables. I believe people were buying upgrades to their flagship LINN DS’s formore than $1,000 to get improvements in PCM playback from them.

That leaves us with the cost of providing quality recordings. I don’t know, but my experience is thatquality recordings sort of went out the window about 20 years ago, at least forpopular music and I am having a hard time thinking adopting PCM as the industrystandard is going to bring it back anytime soon.

There ought to be more emphasis on the actual quality of themusical information in digital formats. There are bizarre high levels ofdynamic compression and the huge amounts of digital clipping in some music thatI have purchased. Paul Simon’s latestoffering, “So Beautiful So What” is a case in point. I love this music but just cannot listen tothis and it is 24 bit/96kHz. I play my Redbookrip of Graceland and turn the volume up until my wife threatens to divorce meand it just gets better. And don’t getme started on “remastered” products. Whathappened?

I think “Boris75” said it best in post #314 on this thread:

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/dsd-or-not-dsd-16093/index13.html.

“My conjecture is that producing a true DSD recording forcesa very short processing chain from the mikes to the master, hence stronglyencouraging minimal production and resulting in a very natural soundingrecording. This suggests that the same could be achieved with 24/96PCM but inmost cases is not because PCM allows a more lazy approach where initial faultscan be "corrected" with more intense production.”

I am not suggesting that the people who care, and who areproducing their products in PCM, will get too lazy if there is no competitionin recording quality. But at the sametime, I am not convinced that the ability to go into any venue, gather up somesound in a hurry, and then screw around with later is going to result in aplethora of quality PCM downloads for me to buy if the industry standardizes onPCM exclusively.

DSD has some advantages. If you are going to take advantage of them, you need to know what youare doing. You need to be able to getcarefully set up and work with the musicians and capture the performancecorrectly and then leave it alone to the maximum extent possible. This discipline is much more likely to resultin more quality music for me to purchase, if for no other reason than themusicians will be there to hear what the recording engineer is doing to theirperformance.

I played classical guitar for a while as a young adult (late70’s into mid-80’s). I collect this typeof music. I have purchased ChannelClassic back catalog of guitar music by the Katona Twins available only inRedbook CD. I listen to it fairlyfrequently lately. I also listen torecordings I have of John Williams and Julian Bream playing some of the samemusic, recording over 30 years ago. Interpretations are different, but these Channel recordings are equal toor better than the older ones that I have. Quality recording requires a realdedication to the art, which if absent, no technology can fix.

I just don’t see why we need to decide which of theseformats is technically superior. But ifsomebody can provide an actual economic analysis as to what the future cost to theconsumer will be, I am ready to be convinced.

Besides, the “Fourth National Scandinavian Nose Whistler'sOrchestra” is one of my favorite orchestras and if I can only get their musicin 128fs DSD, then we need to keep DSD alive for this alone.

[Hiro]

Hiro: I am curious, do you honestly believe that DSD will become a standard recording format rather than just a niche format used by a few small audiophile labels? Personally, while I certainly enjoy the small handful of DSD titles available, I listen to many types of music, and almost all of that music is recorded in PCM (rock, world, jazz, american roots/country, etc). I am certainly not going to choose music to listen by virtue of its format (PCM or DSD). You really think we are going to see a time where DSD becomes a standard recording format for these genres?

 

Delta sigma has already become the standard recording technology, as 99.9% of A/D converters are front ended with sigma delta modulators. It's now only a matter of getting rid of the needless PCM downsampling and bringing the native end-to-end SDM audio to our homes.

[barrows]

Delta sigma has already become the standard recording technology, as 99.9% of A/D converters are front ended with sigma delta modulators. It's now only a matter of getting rid of the needless PCM downsampling and bringing the native end-to-end SDM audio to our homes.

 

OK, so now you are advocating another new format, say 6 bit ~2.8 MHz SDM? Why, exactly? Sounds like a storage problem, and I really see even less chance of record companies shifting to this format?

[labjr]

I woke up on Saturday morning and paid a visit to the LINNForums for the first time in about month. I found a recent post by LINN staff taking the position that DSD is aterrible idea in 2013. I then found my wayto this forum, where a representative of another audio equipment manufacturer, Ayre’sCharles Hansen was spending a great deal of time explaining this in moredetail. Mr. Hansen did a great job of carefullydefining terminology and clarifying a bunch of technical issues that were neverentirely clear to me and I really appreciated that. Moreimportantly, to his credit, he took on the broader issues behind the post onLINN, which is the added cost to the industry, and the consumer, of carryingtwo formats that for all practical purposes can produce a perceptively equalproduct.

Charles also argues that more of the major record labels arelikely to start providing all of us higher quality music files if the industry couldstandardize on one format, which he argues would need to be PCM because iseasier to manipulate. I presume thebasis of that assertion must be related to the costs of providing the music inboth formats and/or having to choose one type of packaging only to find it islater not supported by the playback devices and/or because the cost of getting theperformance right at the recording session is too high and/or the PCM files aremore likely to also be sold by iTunes or an Amazon, increasing the availablemarket beyond the audiophile community.

Well, it seems that Ayre have decided that the cost ofproviding playback of either format can be supported by the market, otherwisewhy have they included DSD playback in an upgrade to their DAC after over tenyears of debate on this subject? Charlesmade it clear the cost to incorporate 64fs DSD was an inconsequential portionof the cost of the upgrade Ayre was providing. Adding DSD playback option was the initial impetus for the upgrade, butAyre took the opportunity to incorporate other benefits. He also decided that another $300, on top of thatcost, was too steep a price for those who want 128fs DSD, but had purchased a designvariant of his product that required more work to upgrade. Did I get that right?

If I have invested 10’s of thousands of dollars in my musicsystem, and thousands of dollars in music, another $300 doesn’t seem too bad tome. People have more money invested incables. I believe people were buying upgrades to their flagship LINN DS’s formore than $1,000 to get improvements in PCM playback from them.

That leaves us with the cost of providing quality recordings. I don’t know, but my experience is thatquality recordings sort of went out the window about 20 years ago, at least forpopular music and I am having a hard time thinking adopting PCM as the industrystandard is going to bring it back anytime soon.

There ought to be more emphasis on the actual quality of themusical information in digital formats. There are bizarre high levels ofdynamic compression and the huge amounts of digital clipping in some music thatI have purchased. Paul Simon’s latestoffering, “So Beautiful So What” is a case in point. I love this music but just cannot listen tothis and it is 24 bit/96kHz. I play my Redbookrip of Graceland and turn the volume up until my wife threatens to divorce meand it just gets better. And don’t getme started on “remastered” products. Whathappened?

I think “Boris75” said it best in post #314 on this thread:

http://www.computeraudiophile.com/f6-dac-digital-analog-conversion/dsd-or-not-dsd-16093/index13.html.

“My conjecture is that producing a true DSD recording forcesa very short processing chain from the mikes to the master, hence stronglyencouraging minimal production and resulting in a very natural soundingrecording. This suggests that the same could be achieved with 24/96PCM but inmost cases is not because PCM allows a more lazy approach where initial faultscan be "corrected" with more intense production.”

I am not suggesting that the people who care, and who areproducing their products in PCM, will get too lazy if there is no competitionin recording quality. But at the sametime, I am not convinced that the ability to go into any venue, gather up somesound in a hurry, and then screw around with later is going to result in aplethora of quality PCM downloads for me to buy if the industry standardizes onPCM exclusively.

DSD has some advantages. If you are going to take advantage of them, you need to know what youare doing. You need to be able to getcarefully set up and work with the musicians and capture the performancecorrectly and then leave it alone to the maximum extent possible. This discipline is much more likely to resultin more quality music for me to purchase, if for no other reason than themusicians will be there to hear what the recording engineer is doing to theirperformance.

I played classical guitar for a while as a young adult (late70’s into mid-80’s). I collect this typeof music. I have purchased ChannelClassic back catalog of guitar music by the Katona Twins available only inRedbook CD. I listen to it fairlyfrequently lately. I also listen torecordings I have of John Williams and Julian Bream playing some of the samemusic, recording over 30 years ago. Interpretations are different, but these Channel recordings are equal toor better than the older ones that I have. Quality recording requires a realdedication to the art, which if absent, no technology can fix.

I just don’t see why we need to decide which of theseformats is technically superior. But ifsomebody can provide an actual economic analysis as to what the future cost to theconsumer will be, I am ready to be convinced.

Besides, the “Fourth National Scandinavian Nose Whistler'sOrchestra” is one of my favorite orchestras and if I can only get their musicin 128fs DSD, then we need to keep DSD alive for this alone.

 

 

Interesting. I would argue that your spacebar is working intermittently.

[Hiro]

With the introduction of hardware for faster DSD record sampling rates, and the eventual introduction of post processing tools in the multi-bit Delta-Sigma Modulator formats like the Sonoma, I believe DSD like (I wish there was an acronym for mulit-bit data streams) will overtake PCM as the studio production format.

 

The dominance of SDM/PDM/PWM in today's audio world is unquestionable, whether you look at A/D converters, USB DACs, or digital amplifiers. However, at this point all of these devices still involve PCM conversions somewhere along the way. With PCM out of the way, we will finally be able to unleash the full resolution of S-D technology.

 

With the recent advancements on the DSD front, we're closer to this reality than ever before.

[Hiro]

OK, so now you are advocating another new format, say 6 bit ~2.8 MHz SDM? Why, exactly? Sounds like a storage problem, and I really see even less chance of record companies shifting to this format?

 

No storage problems whatsoever, as multi-level 2.8MHz/ 5.6MHz / 11.2 MHz SDM can be easily converted to SDM-based DSD format. Tom's already addressed this earlier.

[gumby]

Interesting. I would argue that your spacebar is working intermittently.

 

I cut and pasted this from WORD and the carriage returns were not handled properly. I then started trying to fix it and got timed out. First post here....screwed it up. Asked admin for permission to fix and have not heard back.

[labjr]

I cut and pasted this from WORD and the carriage returns were not handled properly. I then started trying to fix it and got timed out. First post here....screwed it up. Asked admin for permission to fix and have not heard back.

 

I was just trying to be funny. I can read it just fine.

 

I agree with a lot of it. I don't see a problem with having multiple formats. It's downloaded. Not like they have to distribute physical discs in multiple formats. All this hi-rez stuff is a niche market anyway. It's not a format war. If Cookie Marenco wants to record in analog and then distribute DSD files, it's no big deal. It isn't going to make or break her business. If she changes her mind next week and wants to distribute PCM files, she can do that too. That's the good thing about downloads.

 

Personally, I'd rather have physical discs which I then rip the files to my computer. Everything doesn't have to be DSD or PCM for hi-rez downloads to succeeed or fail. The market is tiny.

[Hiro]

If Cookie Marenco wants to record in analog and then distribute DSD files, it's no big deal. It isn't going to make or break her business. If she changes her mind next week and wants to distribute PCM files, she can do that too.

 

I think that the Blue Coast recordings are already distributed in DSD and PCM. So, if someone finds PCM to be "good enough" they don't have to bother with 5.6448MHz DSD files.

[One and a half]

 

I agree with a lot of it. I don't see a problem with having multiple formats.

 

Personally, I'd rather have physical discs which I then rip the files to my computer. Everything doesn't have to be DSD or PCM for hi-rez downloads to succeeed or fail. The market is tiny.

 

Because the market is tiny, you need to have diversity of formats. If the DAC manufacturer can upgrade or make new a DAC perhaps a new format entirely (which is yet to be invented) if it were a breakthrough, then all the better, it increases that loyalty to existing customers, as well as attracting new ones. It's those manufacturers that whine, complain and obfuscate about a format decision that has the exact opposite reaction. In this day of software development changes can occur weekly, however to materialise a product to hold in your hand to play music has long lead times for R&D and production.

 

For downloaded material, especially DSD, DXD, even 192/24, the file sizes are large and require some attention to management and backup. I would also be very supportive of a supplier of HD files to be able to download again in case a file is corrupted or destroyed by the storage medium. Maybe that's another thread.

Physical media provides a sense of ownership, but you have to file/store it with the same management as downloads. Comme ci, comme ca.

[Hiro]

I woke up on Saturday morning and paid a visit to the LINNForums for the first time in about month. I found a recent post by LINN staff taking the position that DSD is aterrible idea in 2013.

 

Coming soon on Linn Records...

 

http://www.computeraudiophile.com/f15-music-general/worlds-highest-resolution-recording-16011/

[tailspn]

Tom, I have been looking for this kind of information for my DSD database. Can you post a link or a quote from Channel Classics on this?

 

Stay tuned, Jared will answer this himself. Meantime you can view a film of the recording of the Beethoven Symphony 5, which includes much of the Channel Classics DSD recording technique here:

 

Making a SACD with the Budapest Festival Orchestra on Vimeo

 

You can also view the custom analog mixers and mic preamps used at the sessions on the Channel Classics site > General Information > About Recording Equipment.

[Charles Hansen]

Why does it have such a huge amount of harmonic and intermodulation distortion?

 

(especially the repeating IMD images every 20 kHz look nasty, peaking at 60 kHz)

 

Hello Miska,

 

Thanks for your interest. I assume by "it" that you are referring to the measurements of the Ayre QA-9 as published on the Stereophile website:

 

Ayre Acoustics QA-9 USB A/D converter Measurements | Stereophile.com

 

Particularly Figure 7.

 

This question is the perfect example to understand the difference between a "digital engineer" and an "analog engineer". An analog engineer would glance at this graph and see at once what is happening. But the digital engineer looks at it and cannot understand it and completely mis-interprets what he sees.

 

First of all remember what the input signal is -- equal amounts of a 19 kHz and 20 kHz sine waves input simultaneously. Each one is at a level of -6 dB (half of the full scale voltage). Then periodically the signals will reach their peak at the same time and add together and reach a full-scale level of 0 dBFS. This is the MAXIMUM possible level for this test. It is considered to be a "pathological" test -- in other words, nothing that would be seen in real life, but rather a "torture test" to see what happens to the circuit when it is pushed to the utter extreme.

 

This test is one type of test for intermodulation distortion (IM). In the "old" days, the standard for this test was that the two signals were 60 Hz and 7 kHz (non-harmonically related, just as in the case of 19 and 20 kHz), but to simulate a more realistic signal that was closer to a real music signal, the low frequency would be a higher level than the high frequency, specifically in a 4:1 ratio.

 

Any non-linearity in the DUT (device under test) will result in all kinds of "spikes" when viewed as an FFT as in Figure 7 in the linked article. The main "spike" of concern is the "difference signal", which in this case will be 20,000 Hz - 19,000 Hz = 1,000 Hz. And if you look at the graph, this spike is at, Mr. Atkinson writes " The 1kHz difference tone lies at just –94dB (0.002%) in the left channel, –106dB (0.0005%) in the right (both ref. the peak amplitude of the input signal)."

 

So to the "digital engineer" these are "huge amounts of distortion! But to an analog engineer, the analog circuit itself is probably responsible for most of this distortion. Now to achieve a measured IMD of the "torture test" variety (rather than the older, easier standard of 60 Hz and 7 kHz in a 4:1 ratio) of between 0.002% and 0.0005% is INCREDIBLY GOOD performance! Especially since this analog circuit is a ZERO-FEEDBACK design!

 

Unlike 99.99% of all analog circuits that try to somehow correct an error after it has been made (which can fool a meter, but not the human ear!) no Ayre product uses any feedback anywhere in the signal path. So an "analog engineer" would be INCREDIBLY IMPRESSED by this measurement.

 

As far as the "especially the repeating IMD images every 20 kHz look nasty, peaking at 60 kHz" again, the "digital engineer" can make no sense of this, but an "analog engineer" instantly recognizes this as simply the 2nd and 3rd harmonics of the input signal, with the 2nd harmonic distortion being about 0.0003% and the 3rd harmonic being at about 0.001%.

 

This is why in the editor's comments Mr. Atkinson refers to the performance of the unit as "Summing up these test results is easy: Ayre Acoustics' QA-9 A/D converter offers superb measured performance that correlates with its equally superb sound quality."

 

There is more to life than imaginary digital representations of real physical phenomena.

 

Best regards,

[Charles Hansen]

Most format die, yes. But not all. Vinyl is the most prominent example. Vinyl is a good product – easy to understand, easy to use and playable on an easy to understand hardware (and on an open standard…).

 

Hello Freann,

 

You chose an interesting example. When all the record companies stopped producing LP's, the sales of the formats were:

 

CD = 33%

Cassette = 33%

LP = 33%

 

So why do you think they chose to discontinue LP, even though there was an obvious demand for it?

 

Most people don't know about these sales numbers. They believe that people stopped buying LP, but that is not true. Instead the truth is that in 1989 they stopped the production of LP for one reason -- they could make 10x more money by selling CD's instead.

 

At that time the production cost of an LP (not including the recording cost, but just to make the product that you purchase in the store) was $1. In the US the retail price was $8. But at that time a CD cost only $0.50 to manufacture (it has since dropped to $0.25) but they could sell if for $16!!

 

Plus with LP there was a "reject" rate (scratched, warped, skipping, etc) for LP of 10%. Every month the record stores would ship the defective product back to the distributor for replacement. With CD, the rejection rate at this time was virtually zero. So the profit was massively higher.

 

The record companies said, "Fuck the customer. We want their money. Who cares what they want. We can do whatever we want to and they can't do anything about it."

 

Now the computer has made it easy to "steal" music. So now they are looking for any way that they can to make money. If they can sell an LP in a "limited" edition for $30 they will do it. Because the demand for LP has always been there -- it never went away. People like the large cover with the beautiful artwork. People like the superb sound quality. CD was originally designed to replace the cassette. But when they marketed "digital is perfect" then they were able to get enough people to believe this lie that people would pay twice as much money for "digital" music than the "old-fashioned" LP....

 

Best,

[Charles Hansen]

Analog stopped being analog long time ago, just like PCM converters stopped being PCM.

 

Freann,

 

Hiro is referring to the fact that some LP cutting lathes have a signal that is delayed by conversion to digital before being sent to the lathe. The lathe can adjust the width between the grooves so that loud passages have more room available, while quiet passages don't need them.

 

But this is far from a universal practice. When the original master is digital it makes a lot of sense. But when the original master was made by going to all of the trouble to use a full analog signal path, no mastering engineer is stupid enough to do that. Instead there is an extra playback head installed on the tape recorder called a "preview head". The time delay for the tape to travel from the preview head to the main playback head accomplishes the same thing without a conversion to digital.

 

All high quality albums made today use the analog preview head, regardless of what propaganda Hiro is trying to spoon-feed you...

 

Best,

[Charles Hansen]

Regardless of the output format chosen, or derived, it can not be a more accurate representation of the original analog input than the AT1201's Delta-Sigma Modulators 6-bit representation. There are losses in every conversion.

 

I disagree.

 

The only loss that occurs with a properly implemented conversion to PCM is to filter out all of the out-of-band (OOB) noise that was "pushed" up above 200 kHz (in the case of the AT1201).

 

So let's consider the conversion into both PCM and DSD. PCM is a 2's complement binary word based system of discrete level values of n bits, which operates at slower sampling rates than the 6-bit Delta-Sigma Modulator. Much slower! Taken to the extreme for this QA-9 ADC, 11.29MHz (256fs) translated to 192KHz is a ratio of 58.8 times slower. The Nyquist frequencies of both sampling rates are the same ratio, meaning that a decimation filter, including Charles' proprietary (?) digital filter, has to filter to below -160dB the Delta-Sigma Modulator's 6-bit stream of all the frequencies above one half of 192KHz, or 96KHz. That requires an aggressive filter. And this is before the math conversion can take place, with its attendant round-off errors.

 

By contrast, the conversion to DSD (2.8224MHz 64fs) requires decimation filtering of a ratio of 4X, and no math translation.

 

In other words, there's just allot more accuracy degradation when converting to any sampling rate PCM, then when re-modulating to DSD from any Delta-Sigma Modulator front ended Analog to Digital Converter.

 

I can see why you would say this. But this is the perfect example of "knowing just enough to get you into trouble".

 

Conversion to PCM requires only one thing -- a low pass digital filter. That's it. Nothing else.

 

On the other hand, to convert to DSD is FAR, FAR more complex!

 

The "DSD" output on the AT1201 is not really DSD. It is actually DSD-256. The engineers at Arda said that to get the best quality single-bit output requires that the 6-bit output of the Delta-Siigma Modulator be sent through a 7th-order modulator:

 

 

This is so much more complicated that it required the entire resources of the latest generation Xilinx Spartan 6 FPGA that we use. There was no room for anything else. So when you flip the switch to "DSD", it actually has to reprogram the FPGA and load a new program where over 90% of the processing power is dedicated to this extremely complex algorithm. Then when you flip back to PCM, it clears out that big mass of processing so that it has enough room to implement no fewer than SIX different low pass PCM filters. And even with six different filters, there is plenty of room left over for other functions. But the DSD algorithm uses all the available power of the FPGA...

 

And not only is this conversion much more complex than the conversion to PCM, it also adds significant amounts of OOB energy! So it's no a loss, but it is a degradation of the signal!

 

To your suggestion though, I have not compared the QA-9 at 24/192 and DSD. I am however familiar with the excellent AT1201 ADC chip as used in the Merging Technologies Horus, and am very impressed with its sound quality. But using that hardware, and comparing its DXD PCM output (352.8KHz) as derived by the Pyramix software and its unrestricted computer operated filters, and the subsampled DSD, leaves the DXD PCM wanting. As I reported earlier, the DSD has a clarity and ease, a heightened sense of spaciousness and instrument individuality that is reduced in the DXD PCM. Just my humble observation, and may apply to me only. But to me, it's very noticeable.

Tom

 

But Barrow's point was the correct one. Just because you have heard one implementation doesn't meant anything. It means that on the Horus, you prefer their conversion to DSD than their conversion to PCM when played back on your hardware. There are so many variables in that statement that any conclusion that you draw cannot be generalized to other equipment, regardless of the converter chip used inside.

 

Best,

[Miska]

This question is the perfect example to understand the difference between a "digital engineer" and an "analog engineer". An analog engineer would glance at this graph and see at once what is happening. But the digital engineer looks at it and cannot understand it and completely mis-interprets what he sees.

 

I'm as much analog engineer as digital engineer, I've been making quite a lot of test and measurement systems during my career. And plenty of signal conversion and processing systems for passive sonar where it is typical to have analog gains from 60 to 120 dB and any distortion would be extremely harmful for analysis results.

 

level for this test. It is considered to be a "pathological" test -- in other words, nothing that would be seen in real life, but rather a "torture test" to see what happens to the circuit when it is pushed to the utter extreme.

 

Kind of yes, because it should have something like couple of thousand closely spaced peaks extending to 100 kHz at similar levels to better represent things like signals from percussions.

 

As far as the "especially the repeating IMD images every 20 kHz look nasty, peaking at 60 kHz" again, the "digital engineer" can make no sense of this, but an "analog engineer" instantly recognizes this as simply the 2nd and 3rd harmonics of the input signal, with the 2nd harmonic distortion being about 0.0003% and the 3rd harmonic being at about 0.001%.

 

For high quality conversion I assume all IMD and harmonic products being at lest below -110 dB, preferably below -120 dB.

 

Higher order IMD products around 19 and 20 tones at 18 and 21 are at -80 dB level which is quite high. These are typical problems for many converters.

 

But you answered to my question by telling it's without feedback, I was suspecting something like that. Would be interesting to know how much of the distortion comes from your digital stages and how much from the analog side.

 

Of course for fun you can compare Stereophile's figures 3 and 4 to loop-back (D/A -> A/D) performance of 159€ E-MU 0404 USB:

 

Of course some people like sound spiced with harmonic and intermodulation distortion. To me, the performance of QA-9 is similar to a typical tube amp.