I haven't kept up with either the latest DAC technology or DIY... I've just got a parts bin of now-ancient DACs.
Can't seem to come to any sort of "conclusion" -- if there is one -- on which classic or modern DACs are worth using (regardless of $$).
For the sake of this thread, let's keep things simple ...
...given careful/"optimum" implementation ... which would you rather have as your desert-island DAC:
ESS -- ES9018K2M (best modern DAC???)
...or ...
Philips TDA1541A (double crown)
Thx!
P.S. This thread may evolve as necessary ... e.g., you can add any comment about NOS, etc.
Can't seem to come to any sort of "conclusion" -- if there is one -- on which classic or modern DACs are worth using (regardless of $$).
For the sake of this thread, let's keep things simple ...
...given careful/"optimum" implementation ... which would you rather have as your desert-island DAC:
ESS -- ES9018K2M (best modern DAC???)
...or ...
Philips TDA1541A (double crown)
Thx!
P.S. This thread may evolve as necessary ... e.g., you can add any comment about NOS, etc.
What' a good DAC?
Anyway, the question was not about GETTING ... it was about HAVING either DAC ...
I dunno ... DIY forums are so confusing to me me now ... what with all the personal (uncontrolled) experiments and projects ... and not a lot of statistical polling.
And then there are the DIY egos ...micro-managers ...anal-nigglers ... yada ... yada ...
Don't get me wrong ... I'm not a huge fan of ABX or pure-objective testing ... but I still don't know what's a good DAC ...
...may be barking up the wrong tree ... that is, and as many suggest, that it really is the IMPLEMENTATION which is key ...i.e., you can make any DAC sound decent. E.g., a few years ago, the Zoran all-in-one chip -- video, audio, Dolby, the whole 9 yds. -- that was in $30 Toshiba SD 3990 DVD players-- remember that?!. It got a lot of modding attention on DIY forums...and that was at a time (2004-2006) when many of the current high-enders (WM8470, CS4398, PCM1792, PCM1704, AD1862 as well as all the classic Philips) were readily avail. WTF? Is DIY that trendy? Why, that trendy pathology smells a lot like the same affliction suffered by DIYer's favorite target of ridicule: audiophiles ... spending all their $$
....so ... DIY ... what have you done for me lately?
I have a few genuine dual crowns in my stash ... they were purch'd around Y2K...
Anyway, the question was not about GETTING ... it was about HAVING either DAC ...
I dunno ... DIY forums are so confusing to me me now ... what with all the personal (uncontrolled) experiments and projects ... and not a lot of statistical polling.
And then there are the DIY egos ...micro-managers ...anal-nigglers ... yada ... yada ...
Don't get me wrong ... I'm not a huge fan of ABX or pure-objective testing ... but I still don't know what's a good DAC ...
...may be barking up the wrong tree ... that is, and as many suggest, that it really is the IMPLEMENTATION which is key ...i.e., you can make any DAC sound decent. E.g., a few years ago, the Zoran all-in-one chip -- video, audio, Dolby, the whole 9 yds. -- that was in $30 Toshiba SD 3990 DVD players-- remember that?!. It got a lot of modding attention on DIY forums...and that was at a time (2004-2006) when many of the current high-enders (WM8470, CS4398, PCM1792, PCM1704, AD1862 as well as all the classic Philips) were readily avail. WTF? Is DIY that trendy? Why, that trendy pathology smells a lot like the same affliction suffered by DIYer's favorite target of ridicule: audiophiles ... spending all their $$
....so ... DIY ... what have you done for me lately?
Sorry, that was not clear to me..
I guess the 32bit ESS might have some advantage in linearity over the TDA, but I have no idea to what degree that's audible. Most influence on the SQ imho is the quality of the DAC's implementation (PSU, PCB design and components).
Encapsulated DAC -- it's the implementation, stupid!
In the non-DIY and "audiophile" crowd there seems to be no similar "consensus". E.g., the latest Astell and Kern AK240 $3k (!!) DAP uses Cirrus Logic CS4398 x2 (Dual DAC) ... my $120 Colorfly DAP uses a single CS4398.
The TDA1545A/TDA1387 are avail in 8-pin SOIC .... so if "classic" Philips segmented-current-architecture sound is so great, Apple or Sony would've put these DACs into their products ...
That's not so far-fetched ... popular Chinese manuf. Head-Direct (HiFiman) uses a classic TDA1543 (8-pin DIP) in its NOS model 601 DAP.
Point being the implementation in which the DAC is encapsulated is more important than the DAC ... if not, prove that remark wrong ... you've go the rest of the thread
Not sure about the ultimate reason why DIYers select one DAC over another. E.g., is it form factor: e.g., TDA1545 and TDA1543 are DIP and only 8 pins ... so easy to breadboard, and veroboard ... etc.
Or are they simply bored and need to fill time with yet another "DAC" project?
Me, I gave up on DIY and picked up toilet-plunging and a crack-cocaine habit. Almost as much fun
Well, on diyaudio.com, ESS 32bit and TDA1541A (select grade) seem to be popular.Sorry, that was not clear to me..
I guess the 32bit ESS might have some advantage in linearity over the TDA, but I have no idea to what degree that's audible. Most influence on the SQ imho is the quality of the DAC's implementation (PSU, PCB design and components).
In the non-DIY and "audiophile" crowd there seems to be no similar "consensus". E.g., the latest Astell and Kern AK240 $3k (!!) DAP uses Cirrus Logic CS4398 x2 (Dual DAC) ... my $120 Colorfly DAP uses a single CS4398.
The TDA1545A/TDA1387 are avail in 8-pin SOIC .... so if "classic" Philips segmented-current-architecture sound is so great, Apple or Sony would've put these DACs into their products ...
That's not so far-fetched ... popular Chinese manuf. Head-Direct (HiFiman) uses a classic TDA1543 (8-pin DIP) in its NOS model 601 DAP.
Point being the implementation in which the DAC is encapsulated is more important than the DAC ... if not, prove that remark wrong ... you've go the rest of the thread
Not sure about the ultimate reason why DIYers select one DAC over another. E.g., is it form factor: e.g., TDA1545 and TDA1543 are DIP and only 8 pins ... so easy to breadboard, and veroboard ... etc.
Or are they simply bored and need to fill time with yet another "DAC" project?
Me, I gave up on DIY and picked up toilet-plunging and a crack-cocaine habit. Almost as much fun
As if manufacturers don't look at the costs....A smart manufacturer looks what can play back current formats (like 24/96 i.e. what the market "demands"), what chips are available and generally they choose a type that stays in production for a while. Using chips that are out of production is reserved for the "cottage industry" and that is not meant negatively !
A one chip solution and low part count combined with good specifications is considered a winner combination in commercial terms.
Last edited:
Human science ca. 2014
MOS-FETs sound somewhat similar to tubes, but MOS-FET distortion doesn't even come close to tube ... IOW, other stuff is going on ... a lot of is probably unknown to human science ca. 2014.
Do you really think (scientific/measurable) "distortion", or lack of, is an important enough reason why careful listeners choose (or simply discern) one DAC over another?
MOS-FETs sound somewhat similar to tubes, but MOS-FET distortion doesn't even come close to tube ... IOW, other stuff is going on ... a lot of is probably unknown to human science ca. 2014.
The ESS K2M parts are more about power efficiency for mobile markets than absolute maximum performance ES9012 and ES9018 are still their headline parts though their specs are pretty similar and the K2M parts may allow different PCB layout improvement optimisations vs the additional available pins on the older 9018 and 9012 components with more pins and access to additional separated supply voltages etc.
the sunk-cost fallacy
But a lot cheap (but generally high-performance Chinese DAPs use decent topology and multiple chips): e..g, Teclast T51 (Mp4Nation) $130 DAP uses dual WM8470s and the equally cheap Colorfly C4 uses a dedicated SRC for upsampling, followed by a CS4398 (AFAIK, the best DAC by Cirrus).
So, I don't think using more than "one chip", as you note, is such a big deal anymore ... not with such low per-unit cost, and the quick/cheap pick-and-place PCB machines. I.e, Chinese economies-of-scale.
Hell, even the hard part (topological design and IDEAS) are easy/cheap ... all one has to do is to hang out in DIY forums. We do all the thinkin' ... the prototyping ... the breadboarding ... the hard/brainy stuff.
Then, the skilled marketer Eagles up some designs ... hooks up with a Chinese manuf./marketing firm ... and you've got ... HiFiMan, QLS, Colorfly, etc.
The AK 240 -- that $3K Korean DAP made for rich folks -- uses many separate chips (decoders, etc.) ... i.e, task-specific (such as a dedicated DSD decoder that sends "pure" DSD to the dual CS4398s). If they can do this much, how much more is it asking to go one step further and add a TDA1387 -- with that "great classic" sound even a lot of high-rez can't compete with? (It would only take some simple switching logic, and maybe a dedicated "classic Philips" out jack?
Point being -- and getting back to the topic of why DIYers seem to really like classic Philips sound ... might it simply be nostalgia or a comfort-zone thing ... a lotta DIYers on this forum are the "older" crowd and have an affinity for classic DACs? Like classic cars ... polluting, poor handling, poor safety, poor fuel economy, ... etc .. but can be restored and souped up ... even greatly enhanced (even when compared to moderns) with much manual effort and after-market kits ....
... is this REALLY the reason why modern TDA1541 kits and DIY projects sound decent? **...
... did Bitstream and Sigma-Delta really take that much of a sonic step back? ...
... or is it that dirty I-word ... IMPLEMENTATION?
... or ***
Footnotes:
** I'll be the first to admit ... my TDA1541A DAC, using PMD100 filter, and Prometheus output stage -- and that overkill external PSU, sounds way better than my much-modded high-rez-capable SACD/DVD-A Pioneer Elite combi. or even Asus Xonar ST ... but the Pioneer and Asus mods are not nearly as extensive.
*** There is also the sunk-cost fallacy ... but we'll save that discussion for next time
I agree with most of this...up to a point
But a lot cheap (but generally high-performance Chinese DAPs use decent topology and multiple chips): e..g, Teclast T51 (Mp4Nation) $130 DAP uses dual WM8470s and the equally cheap Colorfly C4 uses a dedicated SRC for upsampling, followed by a CS4398 (AFAIK, the best DAC by Cirrus).
So, I don't think using more than "one chip", as you note, is such a big deal anymore ... not with such low per-unit cost, and the quick/cheap pick-and-place PCB machines. I.e, Chinese economies-of-scale.
Hell, even the hard part (topological design and IDEAS) are easy/cheap ... all one has to do is to hang out in DIY forums. We do all the thinkin' ... the prototyping ... the breadboarding ... the hard/brainy stuff.
Then, the skilled marketer Eagles up some designs ... hooks up with a Chinese manuf./marketing firm ... and you've got ... HiFiMan, QLS, Colorfly, etc.
The AK 240 -- that $3K Korean DAP made for rich folks -- uses many separate chips (decoders, etc.) ... i.e, task-specific (such as a dedicated DSD decoder that sends "pure" DSD to the dual CS4398s). If they can do this much, how much more is it asking to go one step further and add a TDA1387 -- with that "great classic" sound even a lot of high-rez can't compete with? (It would only take some simple switching logic, and maybe a dedicated "classic Philips" out jack?
Point being -- and getting back to the topic of why DIYers seem to really like classic Philips sound ... might it simply be nostalgia or a comfort-zone thing ... a lotta DIYers on this forum are the "older" crowd and have an affinity for classic DACs? Like classic cars ... polluting, poor handling, poor safety, poor fuel economy, ... etc .. but can be restored and souped up ... even greatly enhanced (even when compared to moderns) with much manual effort and after-market kits ....
... is this REALLY the reason why modern TDA1541 kits and DIY projects sound decent? **...
... did Bitstream and Sigma-Delta really take that much of a sonic step back? ...
... or is it that dirty I-word ... IMPLEMENTATION?
... or ***
Footnotes:
** I'll be the first to admit ... my TDA1541A DAC, using PMD100 filter, and Prometheus output stage -- and that overkill external PSU, sounds way better than my much-modded high-rez-capable SACD/DVD-A Pioneer Elite combi. or even Asus Xonar ST ... but the Pioneer and Asus mods are not nearly as extensive.
*** There is also the sunk-cost fallacy ... but we'll save that discussion for next time
I've been rambling ... my bottom line is:
I want classic Philips sound (with oversampling) out of a smartphone (or same-sized device)...at a reasonable $.
I haven't heard all the DACs and implementations there are -- modern or classic. E.g., TwistedPear, etc.
Some folks say the AK240 is pretty good. For $3K, of course.
Can someone (or some company) come to the rescue?
I want classic Philips sound (with oversampling) out of a smartphone (or same-sized device)...at a reasonable $.
I haven't heard all the DACs and implementations there are -- modern or classic. E.g., TwistedPear, etc.
Some folks say the AK240 is pretty good. For $3K, of course.
Can someone (or some company) come to the rescue?
CONCEPTS ... smart vs. dumb DIY ... duh!
hollowman ... WTF are you yappin' 'bout here?? ...
... and, these concepts and decisions are handed off to lower-level sections of the company/organization...
... so, in a way, the conceptual/experimental ideas introduced in "DIY" forums/blogs/sites -- or even plain ol' social media -- is the smart end of "DIY"...
... toss these ideas around ... veroboard and put on the 'scope ...tweak ... have nice-sounding piece of kit ... post the pics and schema and a write up here on DIYA (or any other topical forum/blog) ...
THAT WAS PART OF YOUR PLAN ALL ALONG ... RIGHT?
... i.e., you wanted it in a nice form-factor for your daily jog/subway-commute ... but didn't want to deal with messy/idiotic/"artistic" case design ... that's why you posted your IDEAS and CONCEPTS in a public forum in the first place ... right? Duh!
hollowman ... WTF are you yappin' 'bout here?? ...
Ideas and CONCEPTS ... high-level/executive stuff. The white boards and napkins and pen&ink-"CADs" found in executive meetings ...Hell, even the hard part (topological design and IDEAS) are easy/cheap ... all one has to do is to hang out in DIY forums. We do all the thinkin' ... the prototyping ... the breadboarding ... the hard/brainy stuff.
Then, the skilled marketer Eagles up some designs ... hooks up with a Chinese manuf./marketing firm ... and you've got ... HiFiMan, QLS, Colorfly, etc.
... and, these concepts and decisions are handed off to lower-level sections of the company/organization...
... so, in a way, the conceptual/experimental ideas introduced in "DIY" forums/blogs/sites -- or even plain ol' social media -- is the smart end of "DIY"...
... toss these ideas around ... veroboard and put on the 'scope ...tweak ... have nice-sounding piece of kit ... post the pics and schema and a write up here on DIYA (or any other topical forum/blog) ...
THAT WAS PART OF YOUR PLAN ALL ALONG ... RIGHT?
... i.e., you wanted it in a nice form-factor for your daily jog/subway-commute ... but didn't want to deal with messy/idiotic/"artistic" case design ... that's why you posted your IDEAS and CONCEPTS in a public forum in the first place ... right? Duh!
No DAC chip is an (desert) island
Your question is much more complex than it may have seemed. We don't experience a DAC chip's sound in isolation, instead, we experience an DAC component box's sound. We actually are hearing a complex holistic system. No DAC chip is an island, they must typically be implemented in conjunction with an digital input receiver and clock recovery circuit, an clock jitter management circuit, an digital image rejection interpolation filter circuit (if any), an A.C. power mains circuit (unless battery powered), local voltage regulators and local noise decoupling circuits, the challenging I/V circuit (if the DAC chip is current output), an analog ultrasonic output filter circuit, and an analog output buffer and signal output coupling circuit.
There is so much more to the sound of a finished DAC box than the sound of the DAC chip it contains for us to be able to identify a best sounding DAC chip, I think. Choosing a desert island built DAC box, or an DIY design is probably a more pertinent, and perhaps, interesting, question.
Your question is much more complex than it may have seemed. We don't experience a DAC chip's sound in isolation, instead, we experience an DAC component box's sound. We actually are hearing a complex holistic system. No DAC chip is an island, they must typically be implemented in conjunction with an digital input receiver and clock recovery circuit, an clock jitter management circuit, an digital image rejection interpolation filter circuit (if any), an A.C. power mains circuit (unless battery powered), local voltage regulators and local noise decoupling circuits, the challenging I/V circuit (if the DAC chip is current output), an analog ultrasonic output filter circuit, and an analog output buffer and signal output coupling circuit.
There is so much more to the sound of a finished DAC box than the sound of the DAC chip it contains for us to be able to identify a best sounding DAC chip, I think. Choosing a desert island built DAC box, or an DIY design is probably a more pertinent, and perhaps, interesting, question.
Last edited:
A "classic" Philips DAC inside a modern DAP:
An externally hosted image should be here but it was not working when we last tested it.
You ought to be banned for implying the I-word again
But it's true.. it's far more complex and trying to obtain a "classic Philips sound" out of an ESS 32b, might be a challenging if not impossible task. Especially if you also want to use a mobile device.
Whoa ...
What I want is, perhaps, mobile device that incorporates, say, an SOIC TDA1387 or TDA1545A. (Forget about TDA1541 as it's too big, complex and power hungry).
So, if we have a very power-efficient TDA1545A in an iPod, we have another problem IMO ... I don't like NOS; I like digital filtering (4x or 8x). That issue seems much easier (than selecting a mobile DAC chip) as we can put that topology into the FPGA (decoder/DSP) ... preferably using some sort of custom filter (with min. phase, apodizing, or 'Wadia', etc). My CS4398-based $130 Colorfly DAP, e.g., allows filter tweaks via menu control: sharp/slow filter, etc. I wouldn't mind being able to select 2x/4x/8x (I think 8x is the fastest the 1545A can handle).
Output stage can be very nicely done using any number of today's cheap+good OPAs and surface-mount passives. Gone are the days of overpriced opa627s and ad8620s ... the ad8397 is a great+modern+cheap device!
Whoa ... words in my mouth dude ... I never said anything remotely resembling
!!!!!!
What I want is, perhaps, mobile device that incorporates, say, an SOIC TDA1387 or TDA1545A. (Forget about TDA1541 as it's too big, complex and power hungry).
So, if we have a very power-efficient TDA1545A in an iPod, we have another problem IMO ... I don't like NOS; I like digital filtering (4x or 8x). That issue seems much easier (than selecting a mobile DAC chip) as we can put that topology into the FPGA (decoder/DSP) ... preferably using some sort of custom filter (with min. phase, apodizing, or 'Wadia', etc). My CS4398-based $130 Colorfly DAP, e.g., allows filter tweaks via menu control: sharp/slow filter, etc. I wouldn't mind being able to select 2x/4x/8x (I think 8x is the fastest the 1545A can handle).
Output stage can be very nicely done using any number of today's cheap+good OPAs and surface-mount passives. Gone are the days of overpriced opa627s and ad8620s ... the ad8397 is a great+modern+cheap device!
No dude... this is YOUR thread, but not the only contributor....
I guess your intentions were not clear to me.
I believe the answer could be simplier if we speak about a power-dac !
But here I agree with Ken. This is not the dac chip alone you listen to. A ES9023 with good layout seems able to sound better than its expensive brother in a bader layout !
Personnaly I believe first external I/V Dac chip sound better ! But am i right ? The little DAC of Subbu/JP tunned & cooked to match your own system is very musical with the good caps (and of course around : good speakers, amp, pre, good links and of course good recording).
Are you aware than a good dac can sound not good if its tonal balance doesn't match well the one of your system ! So no universal good DAC, just some better than others.
The way to answer your question is to make a bridge from your desert Island and have a very good analog source to have a reference : it means : good planar, very good pre : not simple because expensive!
But maybe i can just say than in my system : the limitation come always from the DAC !
As some Island like America or Australia are bigger than Bora Bora, you can live with two DAC, cause sound is also a cultural thing : a good colored dac like a TDA1541 with tubes "too good to be true" is not the reality. I surmise e.g. my little Subbu DAC with a Nichicon LE cap to feed the ES9023 more azcurate than a TDA1541 with classical music if I take care with tonal balance : at least of course in my system (like said no good universal result exist : you have to match the devices). But why not having a TDA1541, AD1862s, PCM56, if you want to change of sound , Do you eat your spaghetti with always the same sauce ?
So, impossible to a brand or a forum to make an universal system and best standalone devices : there is no sense like the question : no better chip but many thing to match together to try to listen a sound like a tailor could do : "sur mesure" !
My 2 cents after several years of bad hifi... still looking for but listen more and more music.... priorities changed !
But here I agree with Ken. This is not the dac chip alone you listen to. A ES9023 with good layout seems able to sound better than its expensive brother in a bader layout !
Personnaly I believe first external I/V Dac chip sound better ! But am i right ? The little DAC of Subbu/JP tunned & cooked to match your own system is very musical with the good caps (and of course around : good speakers, amp, pre, good links and of course good recording).
Are you aware than a good dac can sound not good if its tonal balance doesn't match well the one of your system ! So no universal good DAC, just some better than others.
The way to answer your question is to make a bridge from your desert Island and have a very good analog source to have a reference : it means : good planar, very good pre : not simple because expensive!
But maybe i can just say than in my system : the limitation come always from the DAC !
As some Island like America or Australia are bigger than Bora Bora, you can live with two DAC, cause sound is also a cultural thing : a good colored dac like a TDA1541 with tubes "too good to be true" is not the reality. I surmise e.g. my little Subbu DAC with a Nichicon LE cap to feed the ES9023 more azcurate than a TDA1541 with classical music if I take care with tonal balance : at least of course in my system (like said no good universal result exist : you have to match the devices). But why not having a TDA1541, AD1862s, PCM56, if you want to change of sound , Do you eat your spaghetti with always the same sauce ?
So, impossible to a brand or a forum to make an universal system and best standalone devices : there is no sense like the question : no better chip but many thing to match together to try to listen a sound like a tailor could do : "sur mesure" !
My 2 cents after several years of bad hifi... still looking for but listen more and more music.... priorities changed !
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.