AudioCircle
Other Stuff => Archived Manufacturer Circles => Aspen Amplifiers => Topic started by: bhobba on 8 Aug 2010, 02:56 am
-
Hi Hugh and All
A DAC shootout is going to be occurring on September 4th on the Gold Coast at Lenehan Audio. People like Joe Rasmussen of JLTI fame are coming and it will have the WFS, the Tranquility, Joes JLTI modded Oppo, Mikes new DAC, and a $6K NOS DAC made here in Aus. It will be done blind and people will be giving rankings on each of the DAC's. If anyone wants to join us drop Mike a line and he will add you to the list. Hope to see you there.
And Hugh if you can have the NAKSA ready we may be able to arrange a listen of that as well.
Thanks
Bill
-
Thats VERY interesting. Last two years i tried a lot of DAC's and build all kind of DAC's to hear there main differences on 16/44.1. Heard the Sabre, Wolffson, PCM63, TDA1543 and TDA1541 and others in several configurations with different output (I/V) circuits. My main question was; which configuration is able to give the most musical performance. So my ears and brains where the only judge. It was an interesting journey and in the end i came to some conclusions. Non oversampling DAC's are winner in the musical area. On one way or the other none of the OS DAC's could reach there level, even not the high-tech SABRE 32 with Lundahl trafo's. Another conclusion was that the noisefloor of the powersupply of the DAC chip is VERY important. In my experience no main Power Supply could beat batterie feeding. Even a single voltage regulator between batterie and DAC had audible effects with some DAC's. There are some more issues. My world changed when i met John Brown from ECdesigns. He became the Hugh Dean in DAC's for me. He invented a non-conventional circuit for the NOS DAC chip TDA1541. Last year i followed all his tweaks and optimisations and i dare to say that this is proberly to most musical DAC ever build. It would be nice to take this DAC in the contest, but then someone should have to buy it at your place. It cost 474 euro, thats something else then 6K :o
http://www.ecdesigns.nl/modules/dm1541a
-
Thats VERY interesting. Last two years i tried a lot of DAC's and build all kind of DAC's to hear there main differences on 16/44.1. Heard the Sabre, Wolffson, PCM63, TDA1543 and TDA1541 and others in several configurations with different output (I/V) circuits. My main question was; which configuration is able to give the most musical performance. So my ears and brains where the only judge. It was an interesting journey and in the end i came to some conclusions. Non oversampling DAC's are winner in the musical area. On one way or the other none of the OS DAC's could reach there level, even not the high-tech SABRE 32 with Lundahl trafo's. Another conclusion was that the noisefloor of the powersupply of the DAC chip is VERY important. In my experience no main Power Supply could beat batterie feeding. Even a single voltage regulator between batterie and DAC had audible effects with some DAC's. There are some more issues. My world changed when i met John Brown from ECdesigns. He became the Hugh Dean in DAC's for me. He invented a non-conventional circuit for the NOS DAC chip TDA1541. Last year i followed all his tweaks and optimisations and i dare to say that this is proberly to most musical DAC ever build. It would be nice to take this DAC in the contest, but then someone should have to buy it at your place. It cost 474 euro, thats something else then 6K :o
http://www.ecdesigns.nl/modules/dm1541a
First both myself and dB Audio have reached exactly the same conclusion. The older DAC's are more fluid, liquid and musical but the newer DAC's such as the ESS have greater detail and slam. In designing the Tranquility dBAudio had to work hard to get it to have the detail and slam of the ESS. They say they got close but the ESS still pips it however in musicality etc the Tranquility is better. Of course there is some personal preference here but the designers of the Tranquility built an ESS dac and preferred the design they settled on. dB Audio will be releasing a high resolution DAC soon and it had the exact opposite problem - detail slam etc was no problem - they had to work hard to get the musicality etc. They believe they have succeeded but did insist if I wanted one they would send me one to compare with the Tranquility. Getting both seems a bit like the holy grail.
I would dearly love this DAC in the mix but I already have forked out dosh getting both the WFS and Tranquility in. If someone else would like to get it in I am sure it would be a worthwhile addition.
Thanks
Bill
-
Bill,
The issue of slam and impact v. musicality and 'roundness' really caught my attention. I too face this issue in amp design, in short, tubes v. SS.
It has always seemed to me that the slam and impact of SS goes very nicely with the musicality and roundness of tubes. It was largely this which prompted me to spend a few years designing amps and preamps which combined these distinguishing hallmarks.
Ain't life strange, Bill?
Still doing logos, sadly.......
Hugh
-
Based on the recent review in 6moons, you might also want to see if you can bring in a Weiss Minerva/DAC2. Tho I figure we will be reading Srajan's impressions of how it compares to the W4S fairly soon.
-
Hello
The Tranquility dac are a non-oversampling design, if I remember correctly it use a PCM1704 dac chip.
NOS designs generally do best with simpler music (Jazz, ballads, strings quartets), but tend to obscure inner details in complex (orchestral, heavy rock) music.
The Ecdesigns dac use a linear interpolation by shifting the data of four TDA1541 giving a 4x non-digital oversampling.
OS dac need a very low jitter clock and source and the best filtered power supply, If you lower the jitter in OS dac by changing the on-board xtal near the dac chip by a well filtered low jitter xtal clock chip and moded the power supply for lower noises, you will gain lot of musicality and have the precision of the OS dac.
I have done thoses mods to my Adcom CGD-600 cd player and got an excellent result, that Adcom use a TDA1541A in a OS configuration.
I was tempted by the Twisted Pear Buffalo dac using ESS dac chip but I still hesitate. I will try the PCM1794 (I have some of that chip) in a diy dac and see if it worth the try.
Bye
Gaetan
-
www.dbaudiolabs.com/
The Tranquility dac are a non-oversampling design, if I remember correctly it use a PCM1704 dac chip.
How did you know it uses PCM1704, since which chip is used is never revealed in their literature? Did you actually look inside to confirm?
-
Hello
I've read it in a post on AudioAsylum, but anyway, there is not lot of dac chip who are use for NOS dac, only R2R type can be use, like the TDA1541, TDA1543, PCM1704, AD1865
All newer dac chips are delta-sigma type, except the PCM 1794 who are delta-sigma/R2R hybrid type.
Bye
Gaetan
-
The Tranquility dac are a non-oversampling design, if I remember correctly it use a PCM1704 dac chip.
The exact DAC the Trainquility uses has never been made public. However it has been revealed it is not 24 bit so it is not a 1704.
NOS designs generally do best with simpler music (Jazz, ballads, strings quartets), but tend to obscure inner details in complex (orchestral, heavy rock) music.
That's my experience and the experience of others I have spoken to and corresponded with. The Tranquility guys claim they have largely overcome this using techniques such as short signal paths and litter reduction but is still not as good as the latest chips such as the ESS. Evidently it is now close, but the ESS has the edge. That said the claim is when combined with the strengths of NOS overall it beats the ESS. We will find out.
Thanks
Bill
-
The issue of slam and impact v. musicality and 'roundness' really caught my attention. I too face this issue in amp design, in short, tubes v. SS. It has always seemed to me that the slam and impact of SS goes very nicely with the musicality and roundness of tubes. It was largely this which prompted me to spend a few years designing amps and preamps which combined these distinguishing hallmarks.
Actually I think it is of relevance here. The Tranquility guys claim one way they overcame the weakness of NOS was using individually matched FET output transistors and not valves which many guys like Mike seem to prefer.
Ain't life strange, Bill? Still doing logos, sadly.......
Take your time Hugh. Have other amps I can use in the interim.
Thanks
Bill
-
The exact DAC the Trainquility uses has never been made public. However it has been revealed it is not 24 bit so it is not a 1704.
That's my experience and the experience of others I have spoken to and corresponded with. The Tranquility guys claim they have largely overcome this using techniques such as short signal paths and litter reduction but is still not as good as the latest chips such as the ESS. Evidently it is now close, but the ESS has the edge. That said the claim is when combined with the strengths of NOS overall it beats the ESS. We will find out.
Thanks
Bill
Hello Bill
About the PCM1704, I readed that in AudioAsylum. But there is only few dac chips who can be use for NOS dac.
A NOS dac will have a lot of HF noise at the sampling frequency, and above. This noise will create IM distortion at audible frequencies. It's the main reason of the lack of precision in complex (orchestral, heavy rock) music.
That mosly why NOS dac need an excellent low pass output filter to cut that hf noise, but that filter will bring phase shift. So maby the Tranquility dac guys did find a another way to get rid of that hf noise.
Bye
Gaetan
-
Gaetan, I have the Tranquility DAC, and it definately does not use the 1704. I know what is used but out of respect for dB Audio I will let them reveal it when they feel it necessary. I also know the 24 bit chip used in the upcoming hi rez DAC, and I think they made a wise choice. Also ECdesigns abandoned the multiple DAC chip linear interpolation design long ago. Now John (ECdesigns) is using a single chip using passive i/v conversion using a custom mobius wound resistor. After years of working on linear interpolation and multiple types of i/v conversion using tubes and opamps, he abandoned them for the current offering claiming a single chip with simple i/v conversion using SD sound card for storage sounds the best. I think he tends to exaggerate the sound of his DACs based on his posts, but I will probably get his DAC since I have a Double Crown 1541A chip.
-
It seems like a lot of audiophiles get hung up on the "32 bit" sigma delta DAC chips (ESS, AKM, Wolfson, etc). Problem is these chips including the highly touted ESS chips do not provide more analogue bit resolution than the TDA1541 multibit chip from the 80s. At least this is my understanding. Here is a link to a very well written response from a different forum:
http://www.audioasylum.com/audio/digital/messages/15/153466.html (http://www.audioasylum.com/audio/digital/messages/15/153466.html)
Well, I''ll just post the god damn thing:
"Posted by Thorsten (M) on July 31, 2010 at 05:38:47
In Reply to: RE: Sabre DAC not 32-bit? posted by sheppard on July 27, 2010 at 13:31:52
Hi,
Well, I think we need to define what we mean with "32 Bit DAC" if we want to see if the claims are true or not.
We can interpret it in many ways:
1) A 32-Bit DAC is one that shows 32-Bits equivalent analogue resolution, that is 183dB dynamic range, measured the traditional analogue way. It goes without saying, not only can we not measure such a DAC should it actually exist, but of course no such thing exists nor is it possible, outside a laboratory system suspended in liquid nitrogen and maybe not even then.
2) A 32-Bit DAC is one that uses 32 individual binary weighted bit switches, so it is theoretically capable of producing 2^32 discrete steps, though it's analogue dynamic range is less than the postulated 183dB (or 192dB as some may say). No such thing exists either (yet), but it is at least theoretically possible to make such a device.
3) A 32-Bit DAC is a DAC that accepts a 32-Bit wide data word and outputs whatever real resolution it is capable of, in other words it is marketing number without any appreciable meaning.
It should be added that recent specifications for computer based audio (Intel/M$ HD Audio) call for systems that are able to handle 32 Bit words, simply because this is how computers like to work, they want 8/16/32 Bit words to work with, not 24 Bit. This is the reason for the 32-Bit DAC's now becoming more common.
This has no meaning other than a DAC should accept a 32 Bit word, for compatibility, not that it that actually does anything meaningful with the whole 32 Bit's. In fact, several "32-Bit" DAC's simply take the 32-Bit Data and dither it down to 24-Bit which is then applied to manufacturers 24-bit DAC Core.
However, as it is very easy for marketing departments to point out that "32-Bits are better than 24-Bits (even though under my definition in 1) there is no such thing as a 24 Bit DAC either) just as 24-Bits are better than 16-Bits" and so on, the 32-Bit part has become a major marketing macguffin for DAC (Chip) manufacturers and their customers.
Okay, on to the Sabre (and for that in principle almost all modern DAC's).
In the year of our lord 2010 all but one DAC targeted at audio use a concept that used to be called "hybrid DAC". This means these DAC's combine several bits worth of multibit core with a delta sigma modulator (aka one bit DAC).
The combination is used to achieve the total resolution by using a process called noise shaping from the "real" resolution of the DAC (that is the number of levels that the DAC can directly represent in the analog domain and the additional resolution attained using noise shaping. It is a little difficult to understand and even for those who understand to explain where this extra resolution comes from, but we do not worry ourselves here about details.
The bottom line is that the DAC will have a number analog levels that can be represented directly by 1-Bit and Multibit conversion. The rest has to be produced using noise shaping. The views on noiseshaping vary, my own is negative, where sonics are concerned, compared to having enough real resolution.
Back to the Sabre.
For the "6-Bit DAC" for the Sabre, this is both correct and incorrect.
The Sabre has 2^6 or 64 so called "unitary weighted" or "thermometer code" bit switches. These are able to represent directly 2^6 or 64 individual levels.
Further, the Sabre uses asynchronous sample rate conversion on ALL Input data and converts into a clock rate of 40MHz. If we assume for ease of calculation a 50KHz data sample rate (close enough to to the 44.1KHz used on CD) we can represent as many as 40MHz/50KHz or 800 individual levels using classic pulse width or pulse density modulation.
For ease of calulation I will round up to 1024 levels, which is equivalent to 10 Bit resolution.
This means that the raw resolution build into the ESS DAC is around 16 Bit for single speed (44.1/48KHz) Data, 15 Bit for double speed (88.2/96KHz) and 14 Bit for quad speed (176.4/192KHz) data.
I have to say that this is appreciably more real, raw resolution than most DAC's in the market offer. The ESS Sabre DAC's can actually represent CD Data in the analogue domain with no or very little noiseshaping, which may explain the fact that many find it superior to many other DAC's.
For reference, a highly regarded (by some anyway) 32-Bit DAC by another manufacturer uses a 32 Level (5 Bit) multibit section and 128 Times oversampling at all datarates (7 Bits), thus meaning the actual core of the DAC is able to provide only 12 Bit real resolution without noiseshaping.
Certain others are even more miserly on real resolution, because real resolution costs real money and why bother if you can fake it in the measured performance by agressive use of noise shaping?
It means the ESS DAC relies appreciably less on noise shaping to represent the full needed resolution than most (or at this time perhaps all?) others using the same principle, though it is less than what is attainable using a true multibit DAC. In fact, it is barely able to match the mid 1980's TDA1541 in terms of real (non-noiseshaped) resolution.
For reference, if we combine analogue resolution (24 Binary weighted bits) and the possibility to run at 8 Times oversampling (3 Bits) the Burr Brown PCM1704 (the last true multibit Audio DAC in production) allows us in effect 27 Bit of analogue levels. Sadly this chip is hampered by a SNR/Dynamic range of much less than 120dB, so much of that possible resolution resides below the noisefloor and is of no use. We woudl have to parallel humungous numbers of PCM1704 DAC's to push the noisefloor low enough to make use of the extra bits.
Of course, non of what is written above has any direct relation or mapping onto perceived sound quality. However, a personal observation is that I seem to like DAC's sonics in about the inverse of the amount of noise shaping used. Equally I also know (of) experienced listeners who have a reasonable track record judjing sound quality whose reaction is the opposite (the more noiseshaped the system the better they like it - SACD/DSD being one extreme example).
So the bottom line is - listen for yourself and select what sounds best to you.
Ciao T"
Some may disagree with this guy (like whether noise shaping is a bad or good thing), but he makes a lot of sense to my feeble mind. I think the only DAC I know that can really approach 24 bit analogue resolution is the MSB Diamond DAC (they claim 27 bit resolution), but this is a true DISCRETE MULTIBIT DAC using resistors used in the aerospace industry instead of silicon. Problem is the Diamond DAC start at $25k.
-
It seems like a lot of audiophiles get hung up on the "32 bit" sigma delta DAC chips (ESS, AKM, Wolfson, etc). Problem is these chips including the highly touted ESS chips do not provide more analogue bit resolution than the TDA1541 multibit chip from the 80s. At least this is my understanding. Here is a link to a very well written response from a different forum:
http://www.audioasylum.com/audio/digital/messages/15/153466.html (http://www.audioasylum.com/audio/digital/messages/15/153466.html)
The issue of the exact resolution of these DAC's is quite interesting at a technical level. It is true they strictly only have 6 bit resolution but at a very high frequency. Various tricks are used by preprocessing to give it an effective 32 bit resolution in the audio passband - called various things such as noise shaping or some such. In effect the truncation of the lower bits adds noise to the signal - the noise shaper shifts that noise outside the audio spectrum. This is similar to 1 bit dacs that actually have a horrid one bit resolution but it is done at such a high frequency the resolution in the audio passband is very good. They also use noise shaping. I don't know if this is a good or bad thing - I presume its good since these guys sound like competent engineers to me. The true proof of the pudding is in the eating and I can assure you even compared to Mike Lenehan's quite expensive and heavily tweaked pcm1704 DAC the WFS with the ESS Saber 32 bit DAC has the edge in detail. For example I remember one recording we listened to. It had this tapping noise. On the Havana it was quite indistinct but it was heard very clearly on the WFS and nearly as clearly on the 1704. However the WFS did have a slight sibilance control issue - the 1704 was just about perfect that way. Headphone guys describe it as removing a slight layer of grit from recordings and I believe it. It is in other areas such as musicality, sibilance control etc the 1704 is better. Not a lot better - it is close - but it is better. Speaking to Eric Hider he told me in the comparison they did with the ESS and the Tranquility even though it was close the ESS had the better detail. However in other areas it was ahead and they stuck with the DAC in the Tranquility. I think there is little doubt the ESS deserves the claim of 32 bit.
Thanks
Bill
-
Bill, how are you defining 32 bit? Did you read my entire post? I hope not in terms of equivalent analogue bit resolution which no one in the industry will claim. There is no chip that will claim 180+ dB dynamic range, not even the ESS chip. It may have more detail than the traditional multibit chip but it is not 32 bit resolution, unless you mean the chip "accepts" 32 bit words which seems me just marketing nonsense (There are no 32 bit recordings as far as I am aware). My understanding is even 24 bit DACs aren't really 24 bits because the noise floor caused by electrical and magnetic noise will drown the last several bits of info. To get it to 180+ dB dyanmic range, thermal noise will need to be taken into account (like freezing down to liquid nitrogen if not liquid helium temperatures). I used to work with highly sensitive CCD camera used in astronomy that required cooling to get that last bit of dynamic range. The only DAC that has some credibility in terms of bit resolution seems to be the proprietary MSB Platinum series DACs.
Anyone willing to fork over $25k and include the MSB Diamond DAC in the shootout? :green:
-
Anyone willing to fork over $25k and include the MSB Diamond DAC in the shootout? :green:
The rest will be shot out! :green:
-
Bill, how are you defining 32 bit? Did you read my entire post?
I am defining it as the ultimate resolution that can be achieved in reproducing a very simple signal such as say a 20 hz tone where the various tricks such as noise shaping will be very effective in getting rid of the noise introduced by the 6 bit truncation. Even if it was only one bit resolution at 40mhz such a signal will be reproduced with very high resolution. But this is only tech talk. I was really keen on this sort of stuff once and studied stuff like Dirac Delta functions and Generalized Fourier Transforms. But these days can't really be bothered. Listening to Nat King Cole while I am typing this is more my speed now. Sounds entrancing through the ESS DAC and a Rose Voix Tripath amp on my ML1's. Man that combo is fast, really fast especially up-sampled to 192/24 by J river. Interesting to see if Hugh's new Naksa can match the speed of that Tripath job. It will undoubtedly clobber it in tons of other areas but for speed and acceleration these are killer and really reveal the detail of that ESS chip. It's addictive.
Oh yea - did read it - very nice - thanks for posting it.
PS - Just switched to his daughter Natalee - equally as entrancing and the detail I hear is unforgeable - same as the title of the CD - and her fabulous voice. Yes the musicality of other DAC's may be a bit better but this speed and detail has its merits.
Thanks
Bill
-
Listening to Nat King Cole while I am typing this is more my speed now. Sounds entrancing through the ESS DAC and a Rose Voix Tripath amp on my ML1's. Man that combo is fast, really fast especially up-sampled to 192/24 by J river. Interesting to see if Hugh's new Naksa can match the speed of that Tripath job. It will undoubtedly clobber it in tons of other areas but for speed and acceleration these are killer and really reveal the detail of that ESS chip. It's addictive.
Thanks
Bill
Bill: can the little Tripath amp drive the 86 dB ML1's to reasonable SPLs? I guess the Rose Voix Tripath is very similar to the Kingrex T20, rated at about 10 watts into 8 ohms.
-
Bill: can the little Tripath amp drive the 86 dB ML1's to reasonable SPLs? I guess the Rose Voix Tripath is very similar to the Kingrex T20, rated at about 10 watts into 8 ohms.
Depends on what you mean by reasonable. It can drive them to moderate levels where you would swear the singer was in the room with you no problem - even to the point where you have to turn it down a bit on some recordings. Beyond that it has problems. But since that is the type of music I listen to that is not much of a problem. I am replacing it with another amp while I am waiting for Hugh to finish my Naksa not due to power problems but purely to get a remote volume control to make my testing of the Tranquility easier.
Get one and try it out - they are dirt cheap.
One thing I will also mention is I find there is a tendency to always wind up the volume on highly neutral speakers like ML1's. Resisting that can be a bit of an effort.
Thanks
Bill
-
Yea Bill I get ya. :wink: Tech talk is fun at times, but ultimately it's what you like that matters. Anyways, thanks for having the DAC shootout. I'll go sell my soul and get the Diamond DAC for the shootout. :icon_twisted:
-
Get one and try it out - they are dirt cheap.
This one i used. One of the cheapest and best i know.
http://cgi.ebay.nl/DIY-2x25W-Class-T-Amp-TRIPATHS-TA2021B-KIT-BAUSATZ-/230509579789?cmd=ViewItem&pt=LH_DefaultDomain_77&hash=item35ab710e0d#ht_3560wt_1139
http://www.art0.de/
-
BTW, the poster of the link I provided is none other than Thorsten of AMR:
http://www.amr-audio.co.uk/html/cd_individual.html
I like his CDP using the TDA1451.
-
This one i used. One of the cheapest and best i know.
http://cgi.ebay.nl/DIY-2x25W-Class-T-Amp-TRIPATHS-TA2021B-KIT-BAUSATZ-/230509579789?cmd=ViewItem&pt=LH_DefaultDomain_77&hash=item35ab710e0d#ht_3560wt_1139
http://www.art0.de/
Thats actually a bit more powerful than the one I used. Yea they really are cheap and it is a real hoot to see them compete against much more expensive amps. Those amps are better but the Tripath is not shamed and in the area of detail retrieval and acceleration they are up their with the best at any price.
Thanks
Bill
-
BTW, the poster of the link I provided is none other than Thorsten of AMR:
http://www.amr-audio.co.uk/html/cd_individual.html
I like his CDP using the TDA1451.
Thanks for that. Just one technical thing I will mention. It is well known noise shaping easily gives 16 bit resolution to one bit DAC's eg:
http://web.mit.edu/jonas/www/6.341/paper.pdf
Now the exact detail of these latest 32 bit DAC's is not generally known but that is understandable because its propriety. I do not believe it is unreasonable to take the engineers word and while it is only 6 bit resolution believe it has an effective resolution of 32 bit using more modern techniques than what was done with the older one bit DAC's.
Thanks
Bill
-
Not to promote ECD, but what i appreciate is that he discusses all his knowledge on the forums. He even published the full schematic off his latest 1541DAC. A part of the joy testing DAC's is imho also to learn why it is like it is.
-
Hi Guys
Just returned form Mikes and the DAC shootout is still all go. The only thing, due to the extra work required if a lot turn up, is it may not be blind.
Thanks
Bill
-
Hi
You should really try to maintain the blind testing as it looks like the designers of 2 of the DAC's will be present.
Brian
-
Hi
You should really try to maintain the blind testing as it looks like the designers of 2 of the DAC's will be present.
Brian
Strongly agreed. It usually works out better if one person is designated to do all the switching and kept out of the official listening panel.
One thing often forgotten is how the "unused" DAC's need to be powered up all the time while one DAC is being played. They cannot be unplugged while waiting. However, having DAC's plugged to the wall will contaminate the power line, so the best thing to do is get a power conditioner that isolates the components from wall. A simple isolation transformer will do, and I strongly recommend all the unused DAC's be plugged into the isolator in another room while in wait..
-
A fair comparison is not as easy as it seems. The power issue is something I hadn,t thought of.
There are a lot of pre conceived notions on what certain chips and implementations sound like. NOS does this well and OS dose that well etc. It will be interesting to see if these notions hold up in the face of these very serious implementations both on the NOS and OS sides.
My instincts tell me that at this level these DAC's will sound more alike than different and that as a group they will be much better than lesser efforts of either type. It will be interesting to see the outcome. Hopefully you can keep the blind testing in the mix.
Brian
-
Yesterday spend all day in John Brows lab to update my 1541DAC with his the latest tweaks. We talked about dac technics and there main differences in sound perception. He just posted a short explanation in another forum. But it might be interesting for readers here:
"NOS vs OS
Unfiltered NOS offers highest accuracy in the time domain and poorest in the frequency domain.
Filtered NOS offers reasonable accuracy in both time and frequency domain.
OS offers highest accuracy in the frequency domain and poorest in the time domain.
Linear interpolation (multiple DAC chips in parallel fed by delayed I2S signals) works excellent at lower frequencies, but distorts higher frequencies causing early trebles roll-off among other things. It can work very well in combination with a digital brickwall filter like used by Cambridge Audio in their 4 x TDA1541A CD player.
The question would be what's most important for realistic sound reproduction, highest precision in the time domain or highest precision in the frequency domain. Based on many years of experimenting and listening to equipment designed by other developers it seems that accuracy in the time domain is much more important for achieving most realistic sound reproduction than we realize.
This is the reason why I choose unfiltered NOS. Compromise I have to make is reduced accuracy in the frequency domain.
The problem is, we can't have both, using 44.1 / 16 format, it's either one, the other or a combination of both. It looks like we have to make our own choices based on our personal preferences as there is no perfect way to go.
We also have to realize that sound quality is determined by audio component matching. Connecting some random equipment together is very likely to cause mismatches and resulting sound quality degradation. It is also very important to understand specific audio component properties in order to achieve optimal matching.
Example, unfiltered NOS won't work optimally with class-D or comparable switch-mode power amps as frequency spectra of both DAC and amplifier will inter-modulate."
-
The problem is, we can't have both, using 44.1 / 16 format, it's either one, the other or a combination of both. It looks like we have to make our own choices based on our personal preferences as there is no perfect way to go.
Interesting. I wonder how up-sampling to 192/24 using programs like J River would change things - would you get the best of both worlds with a NOS then?
Thanks
Bill
-
Yesterday spend all day in John Brows lab to update my 1541DAC with his the latest tweaks. We talked about dac technics and there main differences in sound perception. He just posted a short explanation in another forum. But it might be interesting for readers here:
"NOS vs OS
Unfiltered NOS offers highest accuracy in the time domain and poorest in the frequency domain.
Filtered NOS offers reasonable accuracy in both time and frequency domain.
OS offers highest accuracy in the frequency domain and poorest in the time domain.
Linear interpolation (multiple DAC chips in parallel fed by delayed I2S signals) works excellent at lower frequencies, but distorts higher frequencies causing early trebles roll-off among other things. It can work very well in combination with a digital brickwall filter like used by Cambridge Audio in their 4 x TDA1541A CD player.
The question would be what's most important for realistic sound reproduction, highest precision in the time domain or highest precision in the frequency domain. Based on many years of experimenting and listening to equipment designed by other developers it seems that accuracy in the time domain is much more important for achieving most realistic sound reproduction than we realize.
This is the reason why I choose unfiltered NOS. Compromise I have to make is reduced accuracy in the frequency domain.
The problem is, we can't have both, using 44.1 / 16 format, it's either one, the other or a combination of both. It looks like we have to make our own choices based on our personal preferences as there is no perfect way to go.
We also have to realize that sound quality is determined by audio component matching. Connecting some random equipment together is very likely to cause mismatches and resulting sound quality degradation. It is also very important to understand specific audio component properties in order to achieve optimal matching.
Example, unfiltered NOS won't work optimally with class-D or comparable switch-mode power amps as frequency spectra of both DAC and amplifier will inter-modulate."
There is a long discussion on his topic over on HeadFI.com forum titled NOS DAC -marketing bs where Dan Lavry goes into basic sampling theory that governs DAC and ADC operation. It is worth a read for anyone who has an interest in the technical side of the NOS/OS discussion. I can also recommend National Semiconductors Application Notes 236 and 237 as a basic primer.
In the quote above there is reference to time domain and frequency domain as separate charecteristics. When viewing the same signal these 2 domains give you the same information in the end.
When viewing the output of an unfiltered NOS DAC you are not just seeing the audio signal but also the image energy which is related to the sampling frequency. This additional HF energy is what appears to give NOS a faster impulse response.
Much like in the tube/solid state amplifier debates, the end result of listening and studying these issues will result in better DAC's in the future.
Brian
-
Interesting. I wonder how up-sampling to 192/24 using programs like J River would change things - would you get the best of both worlds with a NOS then?
As far that i know there are no DAC chips who are capable to do 192/24 NOS DA converting.
-
As far that i know there are no DAC chips who are capable to do 192/24 NOS DA converting.
The PCM1704 can and is in fact done in the Overdrive DAC:
http://www.empiricalaudio.com/products/overdrive-dac
Thanks
Bill
-
There is a long discussion on his topic over on HeadFI.com forum titled NOS DAC -marketing bs where Dan Lavry goes into basic sampling theory that governs DAC and ADC operation. It is worth a read for anyone who has an interest in the technical side of the NOS/OS discussion. I can also recommend National Semiconductors Application Notes 236 and 237 as a basic primer.
Dan Lavry does have some interesting things to say. Now I will lay my technical biases on the table. I am with Chord when they say it is timing errors that are the important thing. That is why higher frequency sampling rates sound better even though mikes simply don't, nor is there any reason for them to go that high. That said the Tranquility DAC guys say there is much much more to this than the dac conversion technology used and they play as big if not bigger role. They optimized them and by doing that achieved excellent results with a simple 16 bit dac chip.
Thanks
Bill
-
The PCM1704 can and is in fact done in the Overdrive DAC:
http://www.empiricalaudio.com/products/overdrive-dac
Thanks
Bill
Look in the datasheet. Its maximum 48 khz/16 bit. It is a delta sigma DAC with oversampling digital filter. Not really a pure NOS approach i.m.h.o.
-
Look in the datasheet. Its maximum 48 khz/16 bit. It is a delta sigma DAC with oversampling digital filter. Not really a pure NOS approach i.m.h.o.
I did - its a 24 bit R-2R ladder type DAC that can handle up to 768 khz - it does 192/24 NOS in a doodle:
http://focus.ti.com/lit/ds/symlink/pcm1704.pdf
Thanks
Bill
-
I did - its a 24 bit R-2R ladder type DAC that can handle up to 768 khz - it does 192/24 NOS in a doodle:
http://focus.ti.com/lit/ds/symlink/pcm1704.pdf
Thanks
Bill
The datasheet says:
FEATURES
● SAMPLING FREQUENCY (fS): 16kHz to 96kHz
● 8X OVERSAMPLING AT 96kHz
Further:
Digital data words are read into the PCM1704 at eight times the standard DVD audio sampling frequency of 96kHz (e.g., 8 x 96kHz = 768kHz) to create a sinewave output of 1100Hz.
Further:
The output of the I/V converter is then connected to a 40kHz, 3rd-order GIC low-pass filter. The filter output is then passed on to a programmable gain amplifier to provide gain for low-level test signals before being fed into an analog distortion analyzer.
Another approach then the TDA1541, which is filterless.
-
The datasheet says:
FEATURES
● SAMPLING FREQUENCY (fS): 16kHz to 96kHz
● 8X OVERSAMPLING AT 96kHz
Further:
Digital data words are read into the PCM1704 at eight times the standard DVD audio sampling frequency of 96kHz (e.g., 8 x 96kHz = 768kHz) to create a sinewave output of 1100Hz.
Further:
The output of the I/V converter is then connected to a 40kHz, 3rd-order GIC low-pass filter. The filter output is then passed on to a programmable gain amplifier to provide gain for low-level test signals before being fed into an analog distortion analyzer.
The pcm1704 is designed to to be used with a separate oversampling filter that feeds 768khz to the DAC but you don't have to use it if you don't want and guys who use it as a NOS don't. Audio GD for example use its own up-sampling rather than an over-sampling filter not to 768khz but to 192khz and if you feed 192/24 into an Audio GD dac that has the appropriate receiver then it is a nos DAC. Steve Nugent with his overdrive DAC does no up-sampling or over-sampling at all.
I have zero idea why you are arguing about this. Companies have been selling pcm1704 NOS DAC's for yonks and DIY designs aboud eg:
http://www.diyaudio.com/forums/digital-source/166290-192khz-24bit-dac-no-oversampling-no-digital-filter.html
Thanks
Bill
-
I have zero idea why you are arguing about this. Companies have been selling pcm1704 NOS DAC's for yonks and DIY designs aboud.
:? It's not my intention arguing about the PCM1704. I only have some reservation about what is claimed by commercial companies. I agree that i would be interesting to hear high sampled recordings in NOS mode. upsampling a 44.1/16bit source to 192 won't bring much profit in a (NOS) DAC, because there is still the same information.
-
upsampling a 44.1/16bit source to 192 won't bring much profit in a (NOS) DAC, because there is still the same information.
Yes it is exactly the same information with up-sampling and there is some technical controversy if up-sampling is not exactly the same as filtered oversampling. A mathematical analysis shows in the limit as sampling frequency increases they are exactly the same. But because up-sampling is more sophisticated it converges quicker which is why I prefer it - but that is dependent on the algorithm used.
The thing is however Chord believes, and I tend to agree with them, that it is not the presence of higher frequency information that makes material sampled at a higher frequency sound better, it is the extra timing information you get:
http://www.chordelectronics.co.uk/products_detail.asp?id=33
Both up-sampling and oversampling can help in recovering that information even in lower res recordings. I have verified, at least to my ears, material up-sampled to 192/24 sounds better but material recorded at 192/24 sounds better again. I believe this is the reason why.
Thanks
Bill
-
Both up-sampling and oversampling can help in recovering that information even in lower res recordings. I have verified, at least to my ears, material up-sampled to 192/24 sounds better but material recorded at 192/24 sounds better again. I believe this is the reason why.
hmm.. here we get into the dark area of human perception because your experience is the opposite of my personal experience. I personally prefer native unfiltert NOS without any up- or oversampling, because it sound the most like music to my ears. To get the LSB's alive in such a basic approach is damn difficult. Why do i mention this? A commercial DAC shootout is always nice to do, but what can we do and learn from the results? There are so many DAC's on the market and every year there are new kids on the block which should be better sounding. IMO it is also interesting to study and discuss the basic theory's about what influences human sound perception and what is (scientifically) needed to achieve the optimum. On the Aspen forum not a bad idea to talk about musical performance :thumb: 8)
If you wanna read i tell my personal adventure with DAC's. Until a few years ago i was fixated on technical evolution, thought that the Sabre and Wolfson where the new walhalla of sound reproduction like the companies want us to believe. I can remember a moment that i had a bottom tweaked system based on the Twisted Pair Buffalo. I went to a (poor ;-)) friend who only played with an old simple turntable, an old little tubeamp and speakers. I was kind of shocked how i got involved by this sound with all the tic's and noise. Coming home i turned on my Sabre based system with class D amp and it was like a disappointment. That was a definite turnpoint. I was sure that i went the wrong way. I decided to go for musical performance and only that; starting the music and tap your feet. (Honestly i think al audiophiles want that in the end) The science of being involved with music through a DA proces is hard to understand, alltough, it is for me. It's commonly confessed that it is impossible to retrieve all the information of the analog recording from redbook audio because the sample frequentie is to low. The best that can be done is to get the exact analog value of each sample back. Thats plain DA conversion in the opposite way as the ADC proces. It's an well known fact that such conversion comes the closest to realistic playback of the analog source recording. Thats why a lot of audiophiles prefer NOS DAC's. Problem with NOS DAC's is that they are very sensitive for errors like jitter and noise and that they don't correct errors in the frequency domain. To get the full potential out of plain DA conversion we have to choose the best capable chip and surpress and prevent all errors as much as possible. In theory, with that approach and apart from the fact if that really would sounds the most realistic and musical, we have to create the cleanest NOS DAC circuit. John Brown made it without compromises or 'secrets'. Just with using the laws of electronics and physics. The main reason why he choose the TDA1541 over all other NOS DAC chips candidates is because this is the only chip that has a very clean and stable output current source. Thats what he told me. It can also be driven by a passive I/V. Personally it's the best NOS sound i ever heard. Better then this it can't go with plain nos imo. But it is no guaranty that it will sound optimum in every audiochain. You have to put in clean I2S data (like his SD-player. I use a Squeezebox) And not every amp will match with this DAC approach, buts thats another story. Bottomline is that if this DAC doesn't pleasure the ears, NOS is not the way to go. A DAC shootout is imo not really a like DAC contest, but a commercial product evaluation.
-
Did this shootout ever occur? Is it listed under another thread? Thanks
-
I emphatically agree with hybride. I just posted my findings with my experience with the 47 Labs Flatfish transport feeding my Audio Note (With giant upgraded diodes) DAC. There is a completely black background, without a shred of self noise by any of my components.
LET THE POOR MUSIC BREATH. It will thank you for that with unimaginable detail, dynamics, and musicality.
-
Did this shootout ever occur? Is it listed under another thread? Thanks
It certainly did. Check out:
http://www.stereo.net.au/forums/showthread.php/26992-DAC-Shootout
Sorry I forgot to update this thread. Instead of going through the meandering thread above here is the key outcomes from Mike Lenehan and his brother. First from Mikes brother then Mike himself.
Hi all, my name is Russell, Mike Lenehans brother and I had the job of running the systems during the "Dac Shootout" I changed all the discs , help set up the equipment and set the volumes. I also listened continuously for around 6 hours to all the dacs. A couple of points :
1. A tremendous time was had by all. A great diversity of equipment, an even greater diversity of opinion, and a very humbling view of the characters and personalities that we grow into over time. For me , listening to people was just as important as listening to equipment. Not that I agreed with everyone, mind you ! But I took it in.
2. This was NOT a controlled comparison, A/B , blind or whatever. It was an opportunity to listen to a variety of dacs in one system to gain some listening impressions. We at least tried to control SOME variables, the room , the system, the time. At the end of the day, there were some claims of incompatibility etc. which, of course , may be all valid. I assume that we all know that some components work well in some systems and not in others. But remember that all we did here is try to provide a fixed system and environment . And give the manufacturers an opportunity to display their stuff, of course.
** Also note that 4 of the dacs were driven by a modified Marantz CD94 mk 11 transport (Phillips CDM1 mechanism). Except Joe Rasmussens Oppo, which was used as an integrated player
3. Volume levels were different for each listening group. I set the average level and then asked " Up or down" ? In no way was this a level matched comparison. I at least tried to play rock tracks at higher level than acoustic tracks, my estimate is 90 db average for acoustic and 95 db for rock. However some listeners informed that they routinely listen a lot lower or a lot higher than on the day.
4. We did not swap equipment quickly. My estimate is an hour each dac played continously and they received thorough warmup prior ( if the manufacturer requested it)
5. Ok, my opinions. ** Note that I always sat at the back and WAY off centre. However, my opinions seemed to gel with some whose ears I trust who did have the sweet spot. Also , I am a confirmed analog man ( Garrard 401/SME 3012.Grace F9E Ruby) so you can see my priorities.
** 5 dacs , 3 with tube output stages ( Monarchy NM24 - Craig Connor mods , Lenehan PDX , Steve Garlands Killer Dac) Vs two solid state/op amp output (Oppo - Joe Rasmussen mods) and Wyred for Sound.
The 3 tube out put dacs were in a different league to the other two, they were at least all "alive" and the music flowed with a transient flexibility that the two solid state dacs could not manage. The ebbs and flows of music in real time was obvious ( at least to me).
Killer Dac - all the glories of the double crown TDA 1541A shone through. Superbly consistent top to bottom, dimensional and texturally vibrant. Not quite as gutsy and dynamic as the Monarchy and better suited to acoustic than rock. My favourite on the day.
Lenehan PDX ( Clay Geisler designer) Unbelievable flow that sounded real. Again superb texture and dimension. Not quite as driving in the bass with exceptional mid bass detailing, but soft in the lower bass. Not quite as top to bottom consistent as the Killer Dac.
Monarchy NM 24. Unmatched realism , but slightly less "delicate and refined" than the other two tube dacs. Easily the most powerful sounding and best on rock, with quite staggering bass dynamics.
Wyred for Sound. An amazing amount of detail, excellent tonal consistency and powerful bass. But it sounded like you were playing a CD and there were electronic artifacts in the form of a mid treble glare that intruded. Grey background. Microdynamically restricted. However its combination of consistent tonality and detail was impressive in a hifi sense and I can actually believe that some listeners might place this as their favourite. Not me, not by a long shot.
Oppo by Joe Rasmussen Just did not work in this system. Grey, dynamically restricted ( both macro and micro) and we kept turning it up trying to get some life happening. No detail. Joe suggested a fundamental incompatibility between the very fast and wideband output stage with the Monarchy tube preamp. Nevertheless it did not work in this system and sounded terrible. ** There will be some F/Us on this using Joes own preamp/poweramp.
Next is Mike's view
1 st . Garland KillerDAC best on the day overall
I found that when the KillerDAC was playing purely acoustic unamplified program, I mean by that naturally recorded vocals and acoustic only intruments like guitar, violin,cello,and percussion of all types the Kdac was unassailable !! When listening to the ConnorNM24 on nylon string guitar I thought to myself this is so dang good ! the instrument is there ,nothing will touch this. Ahh Haa, I was wrong the Kdac actually showed me what a nylon string guitar really sounds like.
Back to the Connor NM24 and although it was still very very good the nylon string guitar sounded just ever so slightly like someone had tightened the strings a semitone and smeared the strings with warm honey. There was a just perceptible piquedness.
Vocal and harmony presentations for the Kdac were nothing short of real ! ones mouth could sometimes become unhinged ! as Steve Garland explains it , The Beauty ! The music has heart !
Some very moderate caveat’s
1.on fast paced wideband rock material like Joe Cocker Sheffield Steel or Joan Armatrading Square the Circle the Kdac I thought trailed slightly behind the ConnorNM24 in bass control and pace resolution. It was still a virtual Katana on this material but the 24 was just a tad better.
2.Perhaps the noise floor is marginally elevated in the Kdac ! the Connor NM24 may have had a blacker deader canvas to work with.
2nd ConnorNM24 10% behind the Kdac overall
I won’t over extrapolate as it’s mainly covered above. But let me say the things a beast ! I’ve spent time with a full DCS stack and I can tell you the Connor24 smacks it down simple as that. The 24 is a DCS stack with a velvet glove.
If the Kdac is Frank Sinatra the NM24 is Tom Jones !! It’s bass has articulation and snap that is SOTA , pace rythym and timing were no 1 on the day. Although in the end run sounded just a poofteenth caricatured in presentation compared to the Kdac.
LenehanAudio PDX 10% behind Connor NM24
Ok I’ve gotta be brain dead to put my product 3rd ! The PDX came in at 11.30am with the solder still wet , Clay had blown up a DAC chip 3days before and we were gutted . We thought we’d have to pull it.
They’ll never believe us Clay ! everyone will say we nutted out the other DAC’s were better and went back to the drawing board .
Anyway Clay gets the thing running and walks in at 11.30am on the day (30mins before the shootout starts )and says ok it’s going ! we heard one track, which was Joe Cocker Ruby Lee ! and it sounded great , smooth punchy and communicative. So in the mix it ended up !
Basically the PDX sounds very organic detailed and fast . At the start I felt it may topple the ConnorNM24 but the PDX does’nt have the togetherness of the NM24 or the drop dead harmonic texturing of the Kdac.
We are however very happy with the outcome as the shootout PDX was built from junk parts with a pair of incorrect value Duelund VSF capacitors lobbed into it from the other side of the workshop. The IV resistors for instance were 5c metal films.
JLTI Oppo 25% behind LenehanAudio PDX
We just could’nt get this player to work at the level of the other three. It did’nt have a tube back end and I know some thought this could have something to do with it.
Joe did explain that the player does require significant warm up . Later on after the shootout when the player had been on for a few hours we listened again and it had improved significantly , producing significantly improved fluidity and pace but still to my mind retaining a mild patina of grain in the mids and lack of vocal sibilant control.
Joe was sure that this was accurate reproduction and that the player was exposing the true characters of microphone signatures.
Wyred4 Sound DAC2 50% behind JLTI Oppo
This device is a $2000 retail product and sounded it ! Sorry call me biased or whatever but it did’nt belong in this comparison.
Nowhere near the JLTI Oppo in performance , always just on the edge of the resolution verse analytical thing. Bass seemed slightly synthetic and boosted ! mids were lacking real resolution and in it’s place the DigiGods were trying to serve us up sibilance. The top end was clean ! ohh so clean, like listening to music in the fridge ! the brassy sound was absent from the high hats for instance leaving us with only an endless shimmer !! Hmmmm. We’ve heard the Sabre Dac is hard to control.
Here are the voting results for the 5 DAC’s on the day
KillerDAC 4 firsts 1 second
ConnorNM24 2seconds 3 thirds
LenehanAudioPDX 1 first 2seconds 1 third 1 fourth
JLTI Oppo 1 second 1 fourth 3 fifths
Wyred4Sound DAC2 1 first 2 fourths 2 fifths
Since then the Tranquility has arrived and not even optimized by feeding it with the recommended source of a tweaked Mac Mini it was comparable to the NM24 and about the same as the prototype PDX - with the PDX a bit better. We will be doing a much more detailed comparison between an optimized Tranquility, a PDX from the first production run, and the NM24. It is hoped both may surpass the NM24 but we will see. The Tranquility and the PDX will eventually be going on a little tour to Sydney then Perth where they will be checked out against the Killer. If Hugh is happy sending it down to him for guys in Melbourne to checkout can be arranged as well.
Thanks
Bill
-
Bill thanks for the update. Is the KDac available for sale? Website? I have a Double Crown I've been wanting to use.
-
Bill thanks for the update. Is the KDac available for sale? Website? I have a Double Crown I've been wanting to use.
It sure is available for sale. They have their own website with that information:
http://killerdac.com/forum/index.php?topic=49.0\
The guy to contact is Mario who posts as Kajak12.
http://killerdac.com/forum/index.php?action=profile;u=6
Thanks
Bill
-
Hello
Anybody did try any dac using the Burr Brown PCM1794 24bit/192kHz DAC and the SRC4192 Sample rate converter ?
I'm interest about the sound quality of the PCM1794 dac with a SRC4192 because I can have those chip and could made a dac with them.
Thank
Bye
Gaetan