Why DSD is a terrible idea?

Anybody has an opinion on this Linn story on DSD:


"DSD was invented to solve problems that no longer exist today.

A good idea in 1999, DSD was conceived to improve the quality of music at home over the prevailing 16-bit CD format. By bypassing the down-sampling and up-sampling filters in the CD audio chain, DSD sought to improve performance by shortening the audio signal path:

However, DSD has been overtaken by modern technology; A-D and D-A converters have moved away from 1-bit, to far superior multi-bit processes, and the down-sampling and up-sampling filters that DSD sought to bypass have been rendered transparent by the use of higher sample rates and modern algorithms. In fact, DSD is now the quality bottleneck in the modern recording and playback chain.

The arrival of 192kHz 24-bit capable Linn DS players in 2007 signalled the end of DSD. There is no longer the need to convert the original Studio Master recording, which is usually in PCM format, to a DSD stream, because the Studio Master can now be decoded natively on the Linn DS player. Hence, a needless and lossy conversion from PCM to DSD has been eliminated.

DSD has, like so many audio formats, come to the end of it’s natural life. It’s time to move on. "

1 Like

If I understand correctly he’s against converting existing hi-rez PCM records to DSD. In this case he’s right. It can be done on the fly by PSA DS, dCS, the latest Sony and possibly some others. However, what about those who do not have such DAC’s/players? :wink:

Speaking of the play back devices, the proof is in the pudding. It would be interesting to compare them side by side.

What the article completely misses is nearly all modern A/D converters are basically 1-bit sigma delta modulators (DSD) and the 24 bit 192 DAC Linn’s touting uses a similar architecture at their outputs as well, although instead of 1-bit they use a few more in an attempt to get the same performance as a DSD DAC.

Maybe Ted can jump in here but this smacks of a manufacturer not having a DSD based DAC explaining why their lacking approach is better.

It’s BS.


I posted a similar article from Berkeley Audio Designs a few months ago to stir the DSD pot a bit myself. I’ve said it before and I’ll repeat. There is more than one way to skin the PMC to analog conversion cat. Paul’s / Ted’s approach (as is EMM Labs and Playback Design) is to convert to DSD. They apparently managed to build a very fine sounding DAC using this approach which may be a breakthrough in price performance. This does not prove the inherent superiority of using DSD conversion in the PCM to analog conversion chain, (nor the inherent superiority of the native DSD format over PCM, which is an entirely different discussion) .

Linn takes another approach (i.e. architecture) - even if with that approach they manage to build a better sounding DAC than the DirectStream for less money (which I of course highly doubt), this would not prove the inherent inferiority of DSD conversion either - this is all just technology leapfrogging.

1 Like

And then we have this one (same discussion):


Paul McGowan said: What the article completely misses is nearly all modern A/D converters are basically 1-bit sigma delta modulators (DSD) . . .

Only the 1980's delta-sigma DACs employed a true 1-bit architecture. Modern DSD converters are four to six bit. The advantage is much greater S/N.

My experience and knowledge is that neither PCM or DSD is inherently "better." As with many things, it comes down to implementation.

It is a fun topic to watch, especially as DSD proponents argue their chosen format is superior - not appreciating how the majority of DSD spends time as PCM and that DSD stream is almost always decoded as multi-bit. For example, many tout 2L as a source for DSD files, apparently unaware the music was originally recorded in DXD (24/352.8 PCM).

My conclusion is that manufacturers which argue one format is inherently superior are merely marketing a product which - coincidentally I am certain - happens to transcode to/decode that format.

1 Like

I guess that means 4 years from now we’ll be buying Paul’s new decoder that plays 16X DSD files but decodes “hyper” (or whatever) PCM. That is why I would try and discourage anyone spending five figures for digital conversion, unless money is no object (and I don’t know anyone like that)

20 grand for a turntable, sure that makes perfect sense, if you can afford it. But twenty G on digital, that just seems crazy to me.

Also all this DAC talk makes me think back to the Wadia DAC I bought used around 1988 when I got my first real job, to upgrade the sound from a phillips CD880. Remembering back, it actually didn’t improve the sound that much.

I feel your ideals in this are just wrong. audio improvements continue. Now your choice to not be in this is yours. But for me as our lives are very much a short life span I intend to hear all I can. The next ideal is I am not sure there is all that much left to retrieve from current recordings anyway . So I think an investment now is a very worth while idea.


Fortunately we don’t need to spend $20000 on digital. The DirectStream lists for $6000 and will probably be available on Audiogon for $4500 a few months after it is released. I suspect that it will compare favorably with if not be equal to any DAC at any price.

As Al said, buy and listen to whatever you want. As for me, $20000 for ANY front end (or pre or amp or speakers) seems crazy but if money were not a significant issue then I would still be tempted.


I spend $19K on a DAC (MSRP), hoping for it to be (almost) SOTA (in my pricerange) for 2-3 years and then take a big depreciation hit. If the DirectStream lives up to expectations, 2-3 years will have turned out to be 1 year. Economics in digital audio are not too good indeed.

@edorr: what current DAC do you have now?

MSB Signature Plus - Lined up for shootout against DirectStream when it gets here. Both will be running into ARC Ref 5SE for double blind source switching.

That’s is going to be very interesting indeed! What speakers do you use?

Evolution Acoustics MM3

Wow, that’s an impressive speaker! Compliments.

I also heard good stories about the Trinity DAC (10K) and the Jeff Rownland Aeris DAC (60K) but I have the feeling that below 10K the DirectStream will be the champ.

The Trinity is actually $50K. Don’t know about the Jeff Rowland, so may be you have them flipped. According to Paul the DirectStream beat a respected current model $15K DAC in a direct comparison, so I am hopeful it can punch further above its weight than in the <$10K category. We’ll find out about this soon I’m sure.

Sorry, yes I flipped the prices.

Man I wish I had the Eff You money to spend on some of this stuff… $50 Grand would buy me a whole lot of concert tickets, all new Anderson windows, new siding, all kinds of REAL stuff :slight_smile:

And they would provide much better memories and climate control for you hahaha.

In the end it’s for enjoyment that we most likely do not share , but it’s much better than gambling or drugging .


$50K… See what happened with the old $30-40K dCS stack. Fortunately I was smart enough not to buy it when the price dropped to $8K ( I went the “network” way instead :wink: )