Thoughts on a bunch of DACs (and why delta-sigma kinda sucks, just to get you to think about stuff)

[Eee Pee]

Except almost everyone here quotes entire posts rendering the block feature nearly worthless.

 

[evillamer]

 

Rankings are my own based on my own preferences. Your rankings may be different. I believe I said something to that effect in the first post. Maybe I didn't.


I have heard DACs with minimum phase and linear phase filters. Some D-S DACs I have owned or borrowed for extended periods offered several filters. PWD2, Gamma2, Vega, LH, etc. (too many to recall.) In these cases, I've almost always preferred the linear phase filter. Smoother to my ears; however I still wouldn't say analog sounding or particularly smooth compared to R2R. The other ninjas in the evaluations preferred other filters. I'd say the filters are more a matter of "different" than better or analog.

Using your logic, the smoother "analog" sounding R2R DACs of yesteryear and Yggy would sound very digital because they have linear filter characteristics with both pre and post ringing. BTW, if you didn't know, most of the R2R DACs of yesterday also used cheap pre-canned filters. 

Finally, the Audiolab DAC Optimal Transient filters do have pre and post ringing. So there goes your theory. I'd be happy to provide measurements for you if you don't believe me. Don't get bamboozled by their marketing literature. It's impossible to not have any kind of ringing with filters.

Sorry to say it, but you don't know what you are talking about.

Sabre 9018 32-bit reproduces much more detail than the R2Rs, so it would sound glary with Standard Linear Phase filter with pre and post-ringing. With R2Rs, it wouldn't...assuming all other things are equal.

Are you comparing Standard Linear Phase filter vs. Min. Phase on SAME DAC unit implementation?

I'm referring to Audiolab's Optimal Transient/XD/DD Filters which has no ore and post ringing, not the Optimal Spectrum. I'll be happy to see evidence that they have pre and post ringings.

Yeah, you know what you're talking about, so we consumers that know nothing technical should just shut up and be blind bats to what you guys are advertising?

 
evidence:
http://www.hi-fiworld.co.uk/index.php/cd-dvd-blu-ray/62-cd-reviews/642-audiolab-q-dac-filters-.html?start=1

 

[Insidious Meme]

Sabre 9018 32-bit reproduces much more detail than the R2Rs, so it would sound glary with Standard Linear Phase filter with pre and post-ringing. With R2Rs, it wouldn't...assuming all other things are equal.

Are you comparing Standard Linear Phase filter vs. Min. Phase on SAME DAC unit implementation?

I'm referring to Audiolab's Optimal Transient/XD/DD Filters which has no ore and post ringing, not the Optimal Spectrum. I'll be happy to see evidence that they have pre and post ringings.

Yeah, you know what you're talking about, so we consumers that know nothing technical should just shut up and be blind bats to what you guys are advertising?

Ooh, so it's "you don't know what you're talking about" to "you know too much, you're hiding something". Can't wait for the next iteration. :rolleyes:

 

[purrin]

Sabre 9018 32-bit reproduces much more detail than the R2Rs, so it would sound glary with Standard Linear Phase filter with pre and post-ringing. With R2Rs, it wouldn't...assuming all other things are equal.

 
First of all, resolution and "digital sound" are exclusive. Increased resolution does not necessarily cause more digital glare or digital hashy crap. Again, I cite the Gen V and Yggy as examples of this. Extremely resolving DACs but with a very fluid sound with no glare or artificial grain.
 
Secondly, I would not argue against your assertion that Sabre 9018 is more resolving than PCM1704. (I am not a big fan of the PCM1704 BTW). However I would argue that the 20 year old PCM63 is just as resolving as the Sabre 9018. And the recently developed AD5791 (R2R / resistor string) is way more resolving than Sabre 9018. In the end, many other considerations such as power supply, clock/jitter, digital receivers, stacking D-A chips, are just as important for resolution.
 

Are you comparing Standard Linear Phase filter vs. Min. Phase on SAME DAC unit implementation?

 

Yes. For each of the DACs mentioned (and many more), they have adjustable filters. Not comparing across different DACs but on the same DAC. For example, the PWD has five filters, the Gamma 2, three filters, the Vega has a quite a few, etc.
 

I'm referring to Audiolab's Optimal Transient/XD/DD Filters which has no ore and post ringing, not the Optimal Spectrum. I'll be happy to see evidence that they have pre and post ringings.

 

Audiolab Optimal transient:

 
Audiolab Optimal transient XD:

 
Audiolab Optimal transient DD:

 
Audiolab linear phase slow roll off (least amount of ringing of them all - even less than the "Optimal"s)
 

Yeah, you know what you're talking about, so we consumers that know nothing technical should just shut up and be blind bats to what you guys are advertising?

 


There's nothing wrong with being a consumer with less technical knowledge. There is something wrong at arriving at wrong conclusions based on coincidence and lack of exposure to different DACs. There is also something wrong with false technical knowledge.
 
Anyways, I am curious what PCM1704 DACs you have heard?

 

[snip3r77]

Have you tried Terradak's PCM63 x 4 or any of your friends?

 

[rawrster]

I've never tried any R2R dac's and was thinking of getting a new dac down the line anyway. It sounds pretty interesting but not sure if it's possible for me. Is there anything in current production that doesn't take up that much space? I was looking at the Auralic Vega as a potential candiate for my next dac.

 

[agooh]

I have one a big ? mark : why did I see many guys attack or didn't like the sound of R2R ? , I read 2 reviews about it all score 10/10,9.5/10 it wa s an old product but sound very good ,
if still compatible and can be developed why did they stop the production ? what's the story ?
 why there are not many companies to produce R2R dac ??????
really strange !! if someone can answer or put a link .
Thanks

 

[mowglycdb]

It's more expensive, that's why, implementing Sabre 9018 is cheaper.

 

[agooh]

mmmm that was a simple answer, could you name one product ?
Thanks

 

[diamondears]

First of all, resolution and "digital sound" are exclusive. Increased resolution does not necessarily cause more digital glare or digital hashy crap. Again, I cite the Gen V and Yggy as examples of this. Extremely resolving DACs but with a very fluid sound with no glare or artificial grain.

Secondly, I would not argue against your assertion that Sabre 9018 is more resolving than PCM1704. (I am not a big fan of the PCM1704 BTW). However I would argue that the 20 year old PCM63 is just as resolving as the Sabre 9018. And the recently developed AD5791 (R2R / resistor string) is way more resolving than Sabre 9018. In the end, many other considerations such as power supply, clock/jitter, digital receivers, stacking D-A chips, are just as important for resolution.

Yes. For each of the DACs mentioned (and many more), they have adjustable filters. Not comparing across different DACs but on the same DAC. For example, the PWD has five filters, the Gamma 2, three filters, the Vega has a quite a few, etc.

Audiolab Optimal transient:



Audiolab Optimal transient XD:
[COLOR=FF4400]
[/COLOR]

Audiolab Optimal transient DD:



Audiolab linear phase slow roll off (least amount of ringing of them all - even less than the "Optimal"s)




There's nothing wrong with being a consumer with less technical knowledge. There is something wrong at arriving at wrong conclusions based on coincidence and lack of exposure to different DACs. There is also something wrong with false technical knowledge.

Anyways, I am curious what PCM1704 DACs you have heard?

Ok. When I refer to the Standard digital filter, I refer to the Optimal Spectrum filter of Audiolab, not the slow filters. Standard Digital Filter=fast brick wall filter, the standard filter that measures best objectively, the most common digital filter for so long found in most DACs and CD players. The one that has no timing phase distortions nor frequency issues, except pre and post ringings. What should we call this filter?

And even assuming my pet Optimal Transient filters still have pre and post ringings, I'm pretty sure they're minimal or substantially less compared to the Standard digital filter.

But we're getting out of my main point---that you cannot fault the glary-ness on the D-S DAC chip when you're hearing it using the Standard digital filter that has perfect phase and frequency response but has substantial amount of pre and post-ringing not only objectively but also subjectively.

You didn't even mention what filter you used you're basing your ranking on. I don't like bashing D-S DACs, or any DAC chips for that matter, and I strongly suggest account for or review and rank the digital filters used in the market today. I'll be all ears on that.

I suggest further that you check for yourself the Audiolab DACs and see and say for yourself how it measures (objectively) and sounds (subjectively) using the Optimal Transient digital filters, which by the way are slow filters tweaked by the designer that maintains no timing phase distortions.

I'm not sure why digital filters aren't talked about much, maybe you can enlighten us on this, but I suspect it's their very proprietary nature (aka expensive). What do you think?

There's a reason why R2R DAC chips are being phased out, and it's because not only are they expensive but also they reproduce no more details/resolution than the best D-S (and Advance Segment) DAC chips around. Again, unless convinced otherwise, IMHO the culprit on the digital glare being unjustifiably inquisitioned to the D-S DAC chips is the digital filter used that has maximum pre and post-ringing usually used or associated with such D-S DAC chips. You don't think so?

 

[Tachikoma]

There's a reason why R2R DAC chips are being phased out, and it's because not only are they expensive but also they reproduce no more details/resolution than the best D-S (and Advance Segment) DAC chips around.

 
And Analog Devices/Wolfson/AKM care about ultimate detail reproduction because....? High-end audio has no bearing on their business whatsoever.
 
R2R DAC chips have not been fully phased out of business, just the highest spec'ed ones, whereas the cheaper ones have all survived the cull.

 

[diamondears]

There's a reason why R2R DAC chips are being phased out, and it's because not only are they expensive but also they reproduce no more details/resolution than the best D-S (and Advance Segment) DAC chips around.

And Analog Devices/Wolfson/AKM care about ultimate detail reproduction because....? High-end audio has no bearing on their business whatsoever.

R2R DAC chips have not been fully phased out of business, just the highest spec'ed ones, whereas the cheaper ones have all survived the cull.

To me, that's because there's no reason to buy the expensive DAC chip, which means there's not enough SQ improvement/detail resolution to justify the price. Otherwise audio manufacturers would still buy it for their added detail resolution and SQ improvement and just recover the added cost thru pricing, which in turn would be justified to consumers because of higher SQ.

Even assuming, for the same sake of argument, that the R2R chips present better detail resolution and prevents digital glare, the audio manufacturers would have grabbed them and recover the cost to higher pricing. But they didn't because there's no reason to justify the price. That alleged added detail resolution and digital glare avoidance is zip, nil, non-existent.

How much is each R2R DAC chip? Let's say $250. If in fact they'd present better detail resolution and prevent digital glare, you think audio manufacturers and designers wouldn't budge to get them? Cmon now. And we're not talking here about definition yet (such as bass articulacy) and noise elimination/reduction that over sampling digital filters' main goal is for.

Again, IMHO, the culprit on the digital glare accusation is the Standard digital filter with lots of pre and post ringings, not the D-S DAC chip.

I may be getting the Yggy just for kicks, and keep it if my current opinions gets kicked by it. What digital filters does it have btw?

 

[evillamer]

Why would Mike Moffat go all the way to taking a painfully(vs cheap d-s chips) expensive and hard to implement miltary spec chip and try to carve out a dac out of it? Is he nuts or it is because he truly cares getting the best sound of his products? He could have slapped on 4 x AKM AK4495SEQ/4 x ES9018AQ2M/2 x ESS9018S and produce a gungnir gen 2 at the current price of ygg(thus taking more profits due to lower production costs) and his customers would still have flock to buy it right?

 

[eddypoon]

 

[purrin]

 
Quote:

There's a reason why R2R DAC chips are being phased out, and it's because not only are they expensive but also they reproduce no more details/resolution than the best D-S (and Advance Segment) DAC chips around.

 
I won't disagree with you. The remaining R2R chips that exist today, namely the PCM1704, don't resolve as well as the current D-S chips. The PCM1704 also has other issues, syrupy bass, overly forgiving, etc. If you had bothered reading some of the posts instead of jumping in with a closed mind, you would have better understood where I came from.
 

Again, unless convinced otherwise, IMHO the culprit on the digital glare being unjustifiably inquisitioned to the D-S DAC chips is the digital filter used that has maximum pre and post-ringing usually used or associated with such D-S DAC chips. You don't think so?

 
I don't think so. I found no strong correlation to digital filters and "analog" type sound. My PWD2 sounded raspy, even with the filter that minimized pre and post ringing the most which I believe was #4. (I took measurements on the filters and they may be in a post somewhere. I don't remember,) The filters change the sound slightly, but the PWD2 sounds very much digital to my ears regardless of filter. The Wolfson in the PWD2 chips very much have a sound. Same with the Bricasti M1 which has like 9 linear phase and 9 minimum phase filters. It's all a matter of trade offs with the M1's filters, but in the end, the M1 sounds very digital. The AD1955s in the M1 simply have a certain sound. Same with the Vega and the filters it offered. More a matter of different sound then any one being analog sounding.
 
The only exception has been the AKM DACs which were always somewhat more bearable and had a sweeter sound than the other chips. But even then, there is a little bit of glare. I liked the sound of the AKM chips before Schiit came out with the Gungnir. The fact that they chose AKM was had nothing to do with me. Evidently, they thought it sounded the best among the D-S chips.
 
Also, take the R2R Theta DACs of yesteryear. Look at the impulse response of the DSPro Basic:
Looks like quite a bit of pre and post ringing to me, yet all R2R DACs back then have a smoother more fluid analog sound than today's D-S DACs.

 
I suspect what it comes down to is the noise shaping / quantization error that the shallower bit-depths of D-S DACs were operating under which give D-S DACs their digital sounding characteristics, not the pre-post ringing of the filters.
 
In the end, my point is that the state of DAC development went one step forward (more resolution), but two steps back (more digital artifacts).
 
In the context of that, the Yggy was nice surprise because it utilized a modern DAC chip that was more accuracy than anything else before it, D-S or R2R. But evidently, you already know everything including how the great DACs of the 1990s sounded like. We are talking about the Theta Gen V, SFD-2, Spectral SDR2000, STAX DACs, etc; not the craptastic R2R DACs of today, which do include some big names / high ticket cost items.
 
I keep asking you this? Again, what R2R DACs have your heard? What PCM1704 DACs have your heard? Ultimately, you need to cite a good implementation.

 

Even assuming, for the same sake of argument, that the R2R chips present better detail resolution and prevents digital glare, the audio manufacturers would have grabbed them and recover the cost to higher pricing. 

 
You have to get it through your head that no one here has said R2R chips in general resolve better. Some R2R chips resolve better than some D-S chips and vice versa. Practically all of today's D-S chips resolve better than the PCM1704 or whatever industrial DAC chip they use in the Metrum.
 
The only twist is that the AD5791 that the Yggy uses seems to be huge step up from any of the other R2R or D-S chips. But then again, no one has been crazy enough to use a milspec DAC chip with 1ppm accuracy at 20 bits for an audio application.
 

I may be getting the Yggy just for kicks, and keep it if my current opinions gets kicked by it. What digital filters does it have btw?

 

Don't bother with Yggy. The Yggy has the kind of digital filters you won't like with pre and post ringing. Using your preconceptions, it's going to sound very nasty and digital.
 

 

  Why would Mike Moffat go all the way to taking a painfully(vs cheap d-s chips) expensive and hard to implement miltary spec chip and try to carve out a dac out of it? Is he nuts or it is because he truly cares getting the best sound of his products? He could have slapped on 4 x AKM AK4495SEQ/4 x ES9018AQ2M/2 x ESS9018S and produce a gungnir gen 2 at the current price of ygg(thus taking more profits due to lower production costs) and his customers would still have flock to buy it right?

 
Because obviously Mike Moffat, the father of the standalone DAC, who has been doing this for 35-40 years is stupid, and knows far less than armchair DAC designer consumers who seem to who know it all.