[leeperry]

 

I don't know, some people love colored sound...Tubes.
 
Even some manufacturers harp on low jitter and purity of signal or their Dacs and still use a tube.  Sounds hypocritical to be on both sides of the fence but I guess its a manufacturers prerogative.

hehe, yeah...running a tube stage off a NOS DAC output

 
When I heard OPA-Moon, I was seriously shocked by how distorted it sounded..and when I tried a tube headamp(Firestone LC3), it was even worse. the word is not "colored", it's "utterly distorted".
 
OPA-Moon sure begged to differ in those tests..."tubey" they said, hah: http://audio.an-pan-man.com/files/rmaa/earth_vs_moon_vs_sunv2_vs_lt1469.htm

 

[regal]

Quote:

 

hehe, yeah...running a tube stage off a NOS DAC output

 
When I heard OPA-Moon, I was seriously shocked by how distorted it sounded..and when I tried a tube headamp(Firestone LC3), it was even worse. the word is not "colored", it's "utterly distorted".
 

The LC3,  boy you went all out to find a good tube amp didn't you,  hey it even has a couple burr brown opamps in it.  Maybe it was all that NFB that sounded bad.

 

[Wood]

Here is a quote from the front page of the Audio-Gd website.
 
"[size=small]Most people know the global feedback design can offer better specs in test measurements, and non-feedback can't do well in test measurements but can offer better sound for the human's ears. Here is a conflict of the classic circuits."[/size]
 
In the early 80's a professor in my department sold off all his extensive and immaculate LP collection, because like the rest of us, he was seduced by the marketing, based on impressive specs of 'perfect sound for ever' promised by digital.
 
That was nearly 30 years ago when a home computer was a Sinclair ZX81. Despite the, almost unbelievable, progress in chip technology elsewhere, this debate over dac design still drags on, unabated. If there was this much debate, and difference of opinion over which was the superior tech in the computer world, ZX81 vs Intel i7, it would suggest that little or no progress had been made in computing. Its understandable why there is difference of opinion over amplifier design, tube vs ss, because these operate in the analogue domain. But to have so little progress in the digital domain after all these years is a shameful reflection on how badly the digital music industry has served the consumer. Perhaps its not the industries fault, perhaps digital music reproduction in its current state, is fundamentally flawed.
 
I really hope the delta sigma (or something else) route will eventually pay of, at least before I go deaf. But I am certainly not going to allow myself to be misled by the latest and greatest spec sheets, and/or the aggressive marketing that goes with them. I decide what I like and what I buy. I have no allegiance to any particular technology, I just, selfishly, want audio reproduction that at the very least communicates, and for now I find have to compromise with NOS R2R with all its imperfections. 

 

[Parafeed]

Quote:

Don't make me say what I didn't say...good measurements don't prove anything, but bad ones mean that the sound will be colored.
 

One persons coloured is another persons neutral!

I know from experience that I couldn't tell the difference between -3db down at 20k, so I wouldn't get too hung up on a frequency response graph until I'd heard the item in question. On the other hand, 10% THD - well, I wouldn't bother to use my ears. Specs and measurements are an indicator, not the final arbiter of SQ.

 

[Parafeed]

Quote:

Parafeed,  I agree with Lee on the oversampling,   it has always proven to sound better to me than NOS but I'm open minded about it.  Which NOS config are you referring to,  I've been pondering another go at building an NOS DAC but there are so many compromises unless you have the money for double crown TDA1541's?

Wasn't referring to a specific NOS config. Personally, I'll take the AD1865 over the TDA1541, double crown or not. I've built NOS DAC's with the AD1865, PCM63, TDA1543, TDA1541, but keep coming back to the AD1865. Not surprising really as I have a soft spot for the Audio Note DAC's.
 
OS, NOS, ASRC...... I've heard good and bad implementations. You'll probably laugh, but DAC's in my system at the moment - PMC1704 8x OS, AD1865 NOS, Sabre ASRC. I'm not in the camp that would damn one "topology" while hyping another.

 

[ciphercomplete]

OK folks do me a favor and tell me the difference between R2R and Sigma Delta Dacs and how to tell when you have one or the other.  The only thing I know for sure is that the Sabre is sigma delta and NOS designs are R2R.  Beyond that I am lost.

 

[ninjikiran]

Quote:

OK folks do me a favor and tell me the difference between R2R and Sigma Delta Dacs and how to tell when you have one or the other.  The only thing I know for sure is that the Sabre is sigma delta and NOS designs are R2R.  Beyond that I am lost.

Wish I could but I don't think i have ever heard an R2R dac... BUT I can help a little.  You can tell in terms of specs if a dac chip is Sigma-Delta or R-2R(whoaaa I remember making an R-2R circuit in one of the basic electronics courses I did) is if you look at the specs of a dac chip.  There are more than just the sabre, the sigma-delta design after reading this thread and doing ab it of research seems to be quite old.
 
They all seem to have "Delta-Sigma Modulators" in their circuit design.  It seems most dac chips these days are delta-sigma based.

 

[oqvist]

Quote:

OK folks do me a favor and tell me the difference between R2R and Sigma Delta Dacs and how to tell when you have one or the other.  The only thing I know for sure is that the Sabre is sigma delta and NOS designs are R2R.  Beyond that I am lost.

the differences is very small judging from my oppo bd83se and audio GD DAC 19 DSP. It comes down to the implementation in the end I suppose

 

[Currawong]

The whole issue is interesting, but mostly above us due to the complexities of the technology involved. Lately, for me, what has been interesting has been owning the Cambridge 840c CD player.  It has a special DSP feeding AD1955 DA chips.  Not only that, but the filtering is done entirely in the DSP, not in the DA chips (maybe i'm not expressing this correctly, so if I've made an error here, forgive me).  The result is pretty good all things considered.  It's certainly good enough for a headphone rig.  The unit itself has a a bit of glare in the treble, but only noticeable in comparison with other gear.  What makes that interesting, however, is the importance of the digital receiver in a DAC.  It would be interesting to compare it to one of the new Sabre 32 DACs to see if the overall sound was similar in this regard.
 
As for "neutral", I don't get this idea about neutral being a point of view.  The neutrality of something, or rather, the amount of distortion of a device, is measurable. The trouble with DACs is, however, that by the very nature of conversion you have to introduce noise and filters to prevent certain kinds of distortion that come as a result of the limitations of conversion.

 

[middachten]

I'm just a few months back in the audio scene after being very busy with family, career and other hobbies

.
During the eighties and nineties I've worked in design, manufacturing and sales of high-end audio equipment. Most early CD players sounded outright awful. Harsh, flat soundstage and lifeless dynamics. Somewhere in the late eighties/early nineties they started to become acceptable. The CD players I liked most at that time used either the TDA1541(A, S1, S2) or the PCM64. And later the Ultra Analog D20400. All R-2R dacs. Then a new type of converter was introduced they called it Bitstream, MASH, delta/sigma. They all used a form of delta/sigma modulator to reconstruct the original waveform.
 
The technical difference is that a R-2R dac immediately switches between amplitude levels that represent the signal shape. A delta-sigma modulator produces a string of pulses with a certain density that 'add up'  to the amplitude level that represents the signal shape. The reason for this development was the fact that its extremely difficult to produce a R-2R DAC with a true 16bit or higher resolution. This requires ultra precise calibration of The 'R' along the R-2R network. A delta/sigma modulator uses just one calibrated switch (or a few, like the Sabre does).
 
Thats the technical difference. What does this mean soundwise?
The best example for me has always been the development of the ARCAM products. One of my favorite players was the Delta 70.2, which used the TDA1541. It was succeeded by the 70.3 that used the SAA7350, the first bitstream DAC made by Philips (who also produced the TDA 1541). The 70.2 was really shining when it came to dynamics, impact and tonal balance. When the 70.3 came it showed better detail, refinement and depth of soundstage. But it was missing the impact and dynamics. For many years this difference has remained the same. But halfway the nineties chip designers and designers of audio equipment made some big advancements in their delta/sigma based designs. One of the most important ones was the fact that they realized that these dacs are much more sensitive to jitter and the quality of the output buffer/filter.
 
I doubt if I will ever find a DAC that can compete with the amazing Stax X1t which used the D20400. (I worked at the Dutch distributor of Stax at the time this DAC was introduced. I still remember every listening session we had with our demo unit!). But there are plenty CD players out there that will outperform a Marantz CD12/DA12 in terms of soundstage, low level detail and HF smoothness. And that will perform at leas on par when it comes to dynamics and impact.
 
A well implemented delta/sigma dac will outperform a mediocre R-2R design. How the best R-2R design will perform agains the same level of R-2R dac will be known soon...... As I assume that a large part of the head-fi Dedicated Source Components topics will be centering around the currently released Sabre based DACs from Audio-GD

 

[Parafeed]

Quote:

I doubt if I will ever find a DAC that can compete with the amazing Stax X1t which used the D20400. (I worked at the Dutch distributor of Stax at the time this DAC was introduced. I still remember every listening session we had with our demo unit!).

I doubt you will either. The Ultra Analogue D20400 (designed by Dan Lavry, so I've been told) still ranks as one of the best. It's a shame that not many products used it and consequently not many people have heard a product built around it.. That I sold my Sonic Frontiers DAC (which used it) is one of my biggest regrets. What was I thinking!

 

[middachten]

Its not so strange that its not being used more often. The UA dac module alone cost 2-3 times more than some recent dac/amp designs using the Sabre chip (which is one of the more expensive dac chips around these days).

 

[vrln]

I find it odd this whole PCM1704UK vs Sabre32 "war" has become such a big topic. It seems it really did take Audio-gd to start making Sabre32 DAC´s until the general opinion moved more into their direction. Honestly though, Buffalo32 has been here for quite a while already. So has the Eastern Electric and the Wyred 4 Sound DAC range, yet they haven´t become popular here. Isn´t really all that complicated. It isn´t really a brand new chip anymore, and I´d say there´s plenty of material around the web about both. I´d personally sum up the whole situation like this:
 
- The Sabre32 seems to be the first mainstream Delta-Sigma DAC chip that has gotten really, really good reviews, and handles the jitter problem reasonably well. Some say it´s the best, others still prefer the PCM1704. Everyone agrees it´s a good chip. 
 
- The chip alone does not make up the entire sound. There are other variables: analog stages, power supply, filter choices (with PCM1704UK DAC´s) and so on. Both chips are good choices, and if implemented in a high end way like Kingwa does in the REF/NFB series, the difference between the two is not that big. The person who´s most qualified to compare the two ranges says there are only slight differences (more "mature vocals" in PCM1704, better impact/noisefloor on Sabre32, more "analog" sound in PCM1704 vs. a more "modern sound"). As said though, these differences are not that big.
 
So why pick a Sabre32 one:
 
+ Fantastic price/performance
+ Theoretically less transport reliable due to the massive built-in oversampling
+ It´s the latest high end chip out there, new has its benefits
+ Heat levels have less of an effect on sound quality
+ Easier to implement, a kind of "complete package"
 
- A chip tech under development, there will be better Sabre´s eventually. But this isn´t really a minus, especially due to the low cost. And technology never stops, neither should it! Less of a novelty factor compared to R2R DAC´s though.
- In theoretical discussion, R2R vs Delta-Sigma, some claim R2R are fundamentally always better. This is obviously up to debate. Who knows? Not 99.999999% people posting here, including me!
- Not that much to tweak compared to PCM1704UK (filters etc), more of a "complete package"
 
What about the PCM1704UK?
 
+ Proven track record in high end hifi, used in most top line products from traditional vendors such as Naim/Wadia etc. It´s hard to deny this chip is a legend. This doesn´t mean it´s necessarily any better, but it´s proven to be a very worthy contender.
+ High novelty factor, owning one of these can be considered special as they are no longer commonly available. You´ll be able to buy a Delta-Sigma DAC any day, but there aren´t that many PCM1704 users out there anymore. Especially cheap ones ares non-existant. This makes the REF series unique: affordable high end R2R. But also more expensive, so do you want to pay for the novelty? 
+ In the "R2R vs Delta-Sigma" battle, this is the last great R2R chip so if you belong to the R2R camp, you won´t be upgrading any time soon.
+ Versatile as it leaves a lot of choices to the DAC designer.
 
- Expensive, why pay if Sabre32 is close, equal or even a little better and easier/more cost effective to implement?
- The novelty factor can also be seen as being "outdated" or "vintage"
- Heat levels have an effect on sound quality
- Hard to implement as it leaves a lot of choices to the DAC maker, such as the filter etc. 
 
So, IMHO, there is no straight winner. Sound arguments can be made to support both. Both have a lot going for them and are fine choices. And as for the theoretical wars, it´s important to understand the our knowledge of how the brain exactly works (and thus how sound is "created") is not that sophisticated. Measurements can tell certain things (if gear is neutral in the monitor-way), but it does not say if a sound is "good". We are all different. We prefer different sound signatures. Different kinds of music from different periods. In the end the only way to know what you like is to try to get a chance to listen to different gear and compare them yourself. Your own opinion is the most important one when it comes to audio. If you go on the forum research quest, you´ll always find polar opposite views on almost anything when it comes to audio. 

 

[leeperry]

 

Originally Posted by vrln /img/forum/go_quote.gif
 
+ Contemporary music is produced to be heard on Delta-Sigma
[..] 
+ Many classics from the 80´s/90´s were mastered to be heard on R2R.

I think those 2 points are not clear to everyone...most mastering studios since the early 90's have been mastering CD's using Sonic Solutions/Pro Tools(I guess I would know as I used to work as a mastering sound engineer). These obviously use dithering algorithms to render the 16/44.1 final mix. Who/How/Why would anyone tell what kind of DAC they've been optimized for? It's a well known fact that recording/mastering studios don't really give a damn about all the audiophool cable/DAC stuff, and I can only confirm it. Maybe that's the reason why the MFSL remasterings are so successful.
 
are SBM and UV22HR supposed to sound better on one technology than on the other? any hard proof to this?
 
Hope I'm not OT, I honestly don't think I am..

 

[ciphercomplete]

Nice Post, vlrn
 
I recently discovered that my Monarchy NM24 sports dual PCM1704UKs, fancy that.  The funny thing is that before I even knew about all this R2R stuff depending on what tubes I place in the Monarchy I can get it to sound damn near identical to my W4S Dac 1.  Like you said they are both good and I think it is a very good idea to own both.  I prefer the W4S on newer recordings and the NM24 on older ones.  Can't go wrong