|
|
Post by donh50 on Dec 21, 2022 17:53:19 GMT -5
Random thoughts... I would put a passive volume control after the (SS?) preamp and before the tube buffer to minimize the loading effects on the buffer. Probably not a big deal... Here's a thread discussing passive volume control in several iterations: www.audiosciencereview.com/forum/index.php?threads/attenuator-bandwidth.34268/As Keith said, soldering does not involve microwaves, but (very) hot air or IR/laser sources. I quickly grew to hate the Aphex, Linda Ronstadt notwithstanding... |
|
|
|
Post by leonski on Dec 21, 2022 18:14:00 GMT -5
Any more, so far as I know, circuit boards are constructed via wave soldering. All of the surface mount components are placed in their proper locations on the motherboard, and then the entire board (with components) is microwaved just enough to melt the solder without cooking the components. Although this results in a low reject rate for the assembly line, the solder joints aren't always as robust as they might otherwise be. And traditionally, any component that must withstand physical connection/disconnection (such as USB ports) is mounted not on the motherboard itself but rather on a metal reinforcing plate with a plug/sucket connection to the motherboard. On the NUC, this practice was abandoned in lieu of economy of construction. Oh well... Last I saw 'wave soldering'.....It was a shallow TANK of liquid solder. Spelled HOT...... It is excited in such a way as to produce vibrations in the solder. Standing waves........You can demonstate this in your Kitchen Sink.... A stuffed and ready to be soldered board is ran.....maybe a conveyor? just above the molten solder. The little WAVES reach UP and wick into where the solder goes...... DONE! I don't know that I'd expose some components TO microwaves. Any induced currents in integrated circuts may cause failure or reliability issues....down the road. Resistor and capacitors should be OK.......But I'd NEVER put a circuit board IN my Microwave Oven (The Science Oven) for ANY reason whatsoever..... Just looked and found a WIKI...... en.wikipedia.org/wiki/Wave_soldering |
|
|
|
Post by marcl on Dec 21, 2022 18:54:52 GMT -5
Any more, so far as I know, circuit boards are constructed via wave soldering. All of the surface mount components are placed in their proper locations on the motherboard, and then the entire board (with components) is microwaved just enough to melt the solder without cooking the components. Although this results in a low reject rate for the assembly line, the solder joints aren't always as robust as they might otherwise be. And traditionally, any component that must withstand physical connection/disconnection (such as USB ports) is mounted not on the motherboard itself but rather on a metal reinforcing plate with a plug/sucket connection to the motherboard. On the NUC, this practice was abandoned in lieu of economy of construction. Oh well... Last I saw 'wave soldering'.....It was a shallow TANK of liquid solder. Spelled HOT...... It is excited in such a way as to produce vibrations in the solder. Standing waves........You can demonstate this in your Kitchen Sink.... A stuffed and ready to be soldered board is ran.....maybe a conveyor? just above the molten solder. The little WAVES reach UP and wick into where the solder goes...... DONE! I don't know that I'd expose some components TO microwaves. Any induced currents in integrated circuts may cause failure or reliability issues....down the road. Resistor and capacitors should be OK.......But I'd NEVER put a circuit board IN my Microwave Oven (The Science Oven) for ANY reason whatsoever..... Just looked and found a WIKI...... en.wikipedia.org/wiki/Wave_solderingThe wave solder machine always scared me. I'd walk by and think "oh no! What if I forgot and stuck my hand in there!"  |
|
|
|
Post by leonski on Dec 21, 2022 22:53:15 GMT -5
Marci,
You never worked in a dangerous place. Wave solder machine will doubtless have guards and such preventing direct contact WITH the molten solder.
I don't blink about a wave solder machine. But SOME of the stuff I've worked with over the years? DEMAND caution and awareness.
How about 160,000volts? Or various Arsenic, Phosphorus, or Boron compounds? Or Hydroflouric Acid, and half a dozen other acids and proprietary mixtures.......
And SOME used around fairly high voltages or with RF excitation....used in creating a PLASMA discharge to drive chemical reactions?
We had a furnace malfunction many years ago. Ever see SILICON run like water? Maybe 1200c or hotter.
All sorts of nasty stuff....and NO screwing around.....
|
|
|
|
Post by Boomzilla on Dec 21, 2022 23:30:51 GMT -5
An article in a computer magazine described the contemporary methods of soldering surface-mount components to motherboards. Microwaving WAS one of them. Whether this is still done, I don’t know. It’s been a while back…
|
|
|
|
Post by leonski on Dec 22, 2022 1:34:19 GMT -5
Its OWN set of hazards. You don't want to be anywhere NEAR an open case 'transmitter' when in operation.... Probably a good way to concentrate energy at ONE tiny point to accomplish the solder joint. Is it OK to expose some fine geometry integrated circuits TO such radiation? I found a LINK to something that looks so hazardous I'd NEVER try this at home...... www.homemade-circuits.com/using-microwave-oven-parts-build-soldering-iron-heat-controller/you might ask what WOULD replace microwave? Older techniques might not be 'fine' enough while a person with a soldering iron could take a LONG time to solder even a pre-stuffed board.....Not cost effective.....? |
|
|
|
Post by Boomzilla on Dec 22, 2022 6:27:58 GMT -5
Contemplating DACs...
Topping offers their D90se MQA for about $900 with remote volume and XLR outputs.
Schiit offers their multi-bit ladder DAC, the Bifrost for about $800 with remote volume and XLR outputs but no MQA.
So do I want MQA or not? I'm willing to pay extra for XLR outputs and remote volume. If I remember correctly, the Topping PCM DACs generally measure better than the Schiit ladder ones, but that doesn't really indicate how they sound relative to one another.
For now, I'm using my Emotiva Stealth DC1 and it sounds pretty spiffy (although it doesn't do high bit-depth or high sample rate files and it doesn't do MQA). I know that Emotiva is expecting to introduce a new "premium" DAC soon, and I'll probably wait to see what features and price range it has before making a decision.
So backing up to basics and ignoring features (like MQA, DSD, and XLR outputs) for the moment, where does the current state of the art in PCM DACs vs. the state of the art in ladder-DACs stand? Does either technology have any valid claims of superiority over the other? Have high frequency filters improved over time, and what filter options (if any) should I be seeking?
Thanks - Glenn
|
|
KeithL
Administrator 
Posts: 10,273
|
Post by KeithL on Dec 22, 2022 11:06:44 GMT -5
First... Emotiva WILL finally have a new high-end DAC very soon now. You're going to have to wait another week or two for more information on that... But we WILL start releasing some very interesting details right after the new year. I don't have an actual product release date but it will be soon. And, no, we're not big fans of MQA... Now... as for "the state of the art". R2R DACs simply are NOT current technology for audio applications. The state of the art in R2R DACs for audio applications is really: "If you try really hard... how closely can you get the performance of an expensive R2R DAC to come to the performance of a low cost D-S DAC". "You can get an R2R DAC that sounds reasonably good for a reasonable price if you're not concerned about its technical performance." R2R DACs are still current technology in video applications, and in some other applications where very high speed is important, and linearity past 12 or 16 bits isn't so important... But, in audio applications, D-S DACs consistently offer far better performance at far lower cost. Now I need to address another bit of folklore... If you were even thinking of purchasing an R2R DAC "because you want to avoid things like ringing and 'time-domain inaccuracies' caused by oversampling filters".... Then you need to think again. Virtually all modern R2R DACs that offer even decent audio performance use digital oversampling filters just like D-S DACs do. You may be thinking of "non-oversampling DACs"... Those don't use oversampling filters (which is they're called non-oversampling DACs)... And almost all NOS DACs are also R2R... But the performance of NOS DACs is generally so awful that they shouldn't even be mentioned in the same conversation as "state of the art"... NOS DACs can actually be designed to perform well with high-sample-rate content but not with content at the sample rate CDs use... 44k. This is basically the whole reason why oversampling was invented in the first place. (So, if you want to play a CD on a non-oversampling DAC, and have it come out sounding good, you'll have to oversample or upsample it first anyway...   ) Ignoring the fact that D-S DACs generally offer better technical specs and cost less... There is no "inherent way either type of DAC sounds"... (However a manufacturer of an R2R DAC may choose to create a certain sonic signature "to meet their customers' expectations for that sort of DAC".) It's more a matter of the circuitry outside of the DAC chip itself and of the oversampling filter the manufacturer chooses to use. (For example, if you want a "smooth and mellow sound with a slightly rolled off high end" then you'll want a "slow rolloff" filter...) Your question about "high frequency filters" is sort of interesting... There are a wide variety of digital oversampling filter designs and characteristics... And some manufacturers have developed their own special ones instead of using the ones that are included in virtually all modern DAC chips. In fact, most DAC chips offer a choice of several different filters, and the designer can either choose the one they prefer, or allow the user to choose between several different ones. To be quite honest, most of them do an excellent job of doing what they were designed to do, so it's more a question of which one you choose to use. And, in that regard, of choosing the filter option you prefer if you have the option, or of whether you like the one the manufacturer has picked if they choose for you. A "sharp cutoff filter with little ringing" is the most accurate... but some people insist others "sound more natural" with certain types of music. (In fact the differences between filters are pretty subtle... and far less audible than things like how what you're listening to was mastered in the first place.) There is also an "analog reconstruction filter" at the output of a DAC chip. However, to be honest, with an oversampling DAC, these are pretty simple to design and implement, and usually work just fine. (They can be extremely problematic with NOS DACs... which is the main problem with NOS DACs... and the main reason why NOS DACs are not a good idea.) Contemplating DACs... Topping offers their D90se MQA for about $900 with remote volume and XLR outputs. Schiit offers their multi-bit ladder DAC, the Bifrost for about $800 with remote volume and XLR outputs but no MQA. So do I want MQA or not? I'm willing to pay extra for XLR outputs and remote volume. If I remember correctly, the Topping PCM DACs generally measure better than the Schiit ladder ones, but that doesn't really indicate how they sound relative to one another. For now, I'm using my Emotiva Stealth DC1 and it sounds pretty spiffy (although it doesn't do high bit-depth or high sample rate files and it doesn't do MQA). I know that Emotiva is expecting to introduce a new "premium" DAC soon, and I'll probably wait to see what features and price range it has before making a decision. So backing up to basics and ignoring features (like MQA, DSD, and XLR outputs) for the moment, where does the current state of the art in PCM DACs vs. the state of the art in ladder-DACs stand? Does either technology have any valid claims of superiority over the other? Have high frequency filters improved over time, and what filter options (if any) should I be seeking? Thanks - Glenn |
|
KeithL
Administrator
Posts: 10,273
|
Post by KeithL on Dec 22, 2022 11:36:29 GMT -5
Yes... pretty shiny liquid... at 700 degrees... it almost does make you want to touch it. (But, in person, the waves of heat it throws off do sort of push you back a bit.  ) And, yes, nowadays some wave soldering machines use interesting technology... Including some that use a row of small nozzles to only apply the solder to specific spots... However to correct Leonski... The original machines did not use a shallow tank and actual waves... That "wave" you see is liquid solder flowing up and over the edge of a vertical pipe... There's a pump forcing the liquid solder up the pipe from underneath... And the icky brown goo you see floating on the solder around the edges away from the wave itself is peanut oil... The oil is there to limit how much the surface of the solder contacts the air (to slow down how quickly it oxidizes). It's actually pretty ingenious... Since the wave is always flowing it's always touching the board with nice shiny clean solder... And the muck and oxide float off the sides and collect down around the edges. (And, yes, even peanut oil scorches a bit at 700 degrees... but less than most of the alternatives available in the early days of wave soldering.) The entire process for MIL-SPEC boards was even more interesting. In old-style MIL-SPEC electronics you are not allowed to have "exposed copper". This means that you cannot clip off the component leads after you solder the board (because that would expose copper on the lead ends). So... when I worked at Fairchild... back in the 1970's... - the board went onto the conveyor - it went over a "wax wave" that temporarily stuck all the components to the board - then it went over a really scary dish shaped buzzsaw blade that trimmed all the leads about .040" below the board - then it went through a heater that burned the wax adhesive off again and warmed the boards - then to a flux wave - then a heater that warmed the board a bit more and dried the flux - THEN to the solder wave itself - then under some fans to cool it - then off the other end where someone manually removed it from the conveyor - then into mesh buckets ( they still cleaned batches of boards by "hand") I've seen pictures of really fancy automated systems where the board would be automatically removed from the conveyor... Automatically tested... And even automatically shrink wrapped afterwards. However, to be fair, the really scary machine was the "Phase Four Vapor Reflow Soldering Machine". It looked sort of like a freezer with a cover on the top... And, when you looked in the top, all you saw was a sort of shimmering pool of clear harmless looking vapor... (It looked a lot like a relatively harmless vapor degreaser.) You lower the piece into the pool of vapor... the vapor condenses and drips off... and transfers heat to the piece. It actually had several layers of special sorts of Freon or something similar, which had several different, but VERY high, boiling points. It could heat a six or seven pound block of copper to over 400 degrees in about five minutes. And the heat was so well contained that it wasn't even especially hot above the top with the cover open. Last I saw 'wave soldering'.....It was a shallow TANK of liquid solder. Spelled HOT...... It is excited in such a way as to produce vibrations in the solder. Standing waves........You can demonstate this in your Kitchen Sink.... A stuffed and ready to be soldered board is ran.....maybe a conveyor? just above the molten solder. The little WAVES reach UP and wick into where the solder goes...... DONE! I don't know that I'd expose some components TO microwaves. Any induced currents in integrated circuts may cause failure or reliability issues....down the road. Resistor and capacitors should be OK.......But I'd NEVER put a circuit board IN my Microwave Oven (The Science Oven) for ANY reason whatsoever..... Just looked and found a WIKI...... en.wikipedia.org/wiki/Wave_solderingThe wave solder machine always scared me. I'd walk by and think "oh no! What if I forgot and stuck my hand in there!" |
|
|
|
Post by Boomzilla on Dec 22, 2022 11:52:43 GMT -5
,,,in audio applications, D-S DACs consistently offer far better performance at far lower cost. That's pretty much what I had concluded from reading reviews, but I appreciate your confirmation. I've also experimented with various filters on DACs, and (for the most part) don't hear much or in some cases any difference. Again, measurements I've seen seem to reinforce this experience with filters - the differences are VERY high in frequency, and usually consist of how much (or little) ringing is allowed after a transient. Of course, in some few filters, there is a bit of pre-ringing before a transient, but again at such a high frequency that I probably couldn't hear it. So since D-S DAC chips are so good (and so cheap), I'm looking forward to a REAL bargain on Emotiva's new DAC pricing!  |
|
cawgijoe
Emo VIPs 
"When you come to a fork in the road, take it." - Yogi Berra
Posts: 5,035
|
Post by cawgijoe on Dec 22, 2022 12:32:35 GMT -5
One R2R dac that has received excellent reviews is the Denafrips ARES II. Check out Audio Science Review (Amir). As a measurement guy, Amir doesn't give many products a "yes". Even the manufacturer claims that R2R dacs don't measure as well as D-S Dacs, but somehow this one does measure well. There are other positive reviews out there including from Stereophile.
There is no doubt that what Keith says is correct in terms of better performance at lower cost, however there is something really cool about how well made and organized the Denafrips circuit boards are. They are mesmerizing. I bought my ARES II from a friend who moved up the line and I thought $500 for a 6 month old dac was quite reasonable. No regrets. Sounds great to me.
On that note, really looking forward to Emotiva's next dac.
|
|
KeithL
Administrator
Posts: 10,273
|
Post by KeithL on Dec 22, 2022 15:53:53 GMT -5
A bit more on filters... It's specifically worth noting that "ringing" in oversampling filters is not what you might think... Specifically the ringing itself occurs at the sampling frequency... so is itself not audible at all... The part that is at least potentially slightly audible is the "time energy smear" associated with it. (You cannot possibly hear the ringing itself... what you might hear is that the transient has a slightly different duration or profile than it should.) Another thing to understand is that the ringing is inherent in the signal once you apply the necessary band-limiting to convert the original into digital. And then, likewise, it is inherent in the process of properly converting that digital signal back into analog. The ringing is NOT "an error"; it's part of what you have to have in order for the math to work out correctly. You can create a filter with no ringing but, if you do, it can only be done at the expense of other serious and audible flaws in performance. (It's also worth noting that virtually all musical instruments and naturally occurring sounds have some ringing after them... ) Another confusing thing is that, whenever you see illustrations of transient responses and ringing, they always start with "a perfect transient pulse". The catch is that this isn't actually a VALID audio signal; by definition a perfectly square pulse cannot exist in a valid digital audio signal. (You can make one on a test set, and it's handy for visualizing what the filter is doing, but one cannot exist in an actual recording of a digital audio signal.) The simple reality is that the "ideal and most accurate" filter actually has a small amount of ringing. But what you can do is to mathematically shift the point in time where the ringing occurs. You can have a relatively small amount of ringing arranged symmetrically before and after the impulse... Or you can eliminate virtually all of the ringing before the impulse by "pushing it all after" (post-ringing)... Or you can eliminate virtually all of the ringing after the impulse by "pushing it all before" (pre-ringing)... (Of course you can also design a non-optimal filter that has more than the necessary amount of ringing... which just might sound interesting.) The most accurate choice is going to be "shortest possible ringing arranged symmetrically before and after"... But many people seem to agree that, at least with some instruments, a filter with "no pre-ringing and more post-ringing" "sounds more natural". Our Ego+ DACs offer both of these options... and the ability to switch back and forth relatively easily. And, on at least a few recordings, I can hear a slight difference between them (but only a few). Although, to be fair, I cannot claim to know which is actually better. And that's an excellent experiment that anyone with one of our Ego DACs should try... see which filter choice YOU like better... I would also be remiss if I didn't mention that all of these differences are really small. They're nothing compared to the errors introduced by microphones, and microphone preamps, and tape heads, and magnetic tape, and vinyl. And we have no idea what filter settings were used in the A/D converter when the original digital conversion was made. Or even if the various tracks in that recording were made with the same A/D converter, or in the same studio, or in the same week. You really don't want to think too hard about the jitter specs on a 25 year old "state-of-the-art" A/D converter... (or a studio tape mastering deck either). Or the number of cheap op-amps, and feet of cheap wire, that were included in the console your favorite album was mixed on... ,,,in audio applications, D-S DACs consistently offer far better performance at far lower cost. That's pretty much what I had concluded from reading reviews, but I appreciate your confirmation. I've also experimented with various filters on DACs, and (for the most part) don't hear much or in some cases any difference. Again, measurements I've seen seem to reinforce this experience with filters - the differences are VERY high in frequency, and usually consist of how much (or little) ringing is allowed after a transient. Of course, in some few filters, there is a bit of pre-ringing before a transient, but again at such a high frequency that I probably couldn't hear it. So since D-S DAC chips are so good (and so cheap), I'm looking forward to a REAL bargain on Emotiva's new DAC pricing! |
|
|
|
Post by leonski on Dec 22, 2022 15:54:13 GMT -5
Yes... pretty shiny liquid... at 700 degrees... it almost does make you want to touch it. (But, in person, the waves of heat it throws off do sort of push you back a bit. ) And, yes, nowadays some wave soldering machines use interesting technology... Including some that use a row of small nozzles to only apply the solder to specific spots... However to correct Leonski... The original machines did not use a shallow tank and actual waves... That "wave" you see is liquid solder flowing up and over the edge of a vertical pipe... There's a pump forcing the liquid solder up the pipe from underneath... And the icky brown goo you see floating on the solder around the edges away from the wave itself is peanut oil... The oil is there to limit how much the surface of the solder contacts the air (to slow down how quickly it oxidizes). It's actually pretty ingenious... Since the wave is always flowing it's always touching the board with nice shiny clean solder... And the muck and oxide float off the sides and collect down around the edges. (And, yes, even peanut oil scorches a bit at 700 degrees... but less than most of the alternatives available in the early days of wave soldering.) The entire process for MIL-SPEC boards was even more interesting. In old-style MIL-SPEC electronics you are not allowed to have "exposed copper". This means that you cannot clip off the component leads after you solder the board (because that would expose copper on the lead ends). So... when I worked at Fairchild... back in the 1970's... - the board went onto the conveyor - it went over a "wax wave" that temporarily stuck all the components to the board - then it went over a really scary dish shaped buzzsaw blade that trimmed all the leads about .040" below the board - then it went through a heater that burned the wax adhesive off again and warmed the boards - then to a flux wave - then a heater that warmed the board a bit more and dried the flux - THEN to the solder wave itself - then under some fans to cool it - then off the other end where someone manually removed it from the conveyor - then into mesh buckets ( they still cleaned batches of boards by "hand") I've seen pictures of really fancy automated systems where the board would be automatically removed from the conveyor... Automatically tested... And even automatically shrink wrapped afterwards. However, to be fair, the really scary machine was the "Phase Four Vapor Reflow Soldering Machine". It looked sort of like a freezer with a cover on the top... And, when you looked in the top, all you saw was a sort of shimmering pool of clear harmless looking vapor... (It looked a lot like a relatively harmless vapor degreaser.) You lower the piece into the pool of vapor... the vapor condenses and drips off... and transfers heat to the piece. It actually had several layers of special sorts of Freon or something similar, which had several different, but VERY high, boiling points. It could heat a six or seven pound block of copper to over 400 degrees in about five minutes. And the heat was so well contained that it wasn't even especially hot above the top with the cover open. The wave solder machine always scared me. I'd walk by and think "oh no! What if I forgot and stuck my hand in there!" Pump?? Solder? Can't EVER let it go cold or you'll probably never get it going again....... |
|
|
|
Post by Boomzilla on Dec 22, 2022 16:37:29 GMT -5
...The most accurate choice is going to be "shortest possible ringing arranged symmetrically before and after"... But many people seem to agree that, at least with some instruments, a filter with "no pre-ringing and more post-ringing" "sounds more natural". Our Ego+ DACs offer both of these options... and the ability to switch back and forth relatively easily. And, on at least a few recordings, I can hear a slight difference between them (but only a few). Although, to be fair, I cannot claim to know which is actually better. And that's an excellent experiment that anyone with one of our Ego DACs should try... see which filter choice YOU like better... ... Your ears are better than mine, KeithL - I swapped back and forth on my Big Ego+ but couldn't really hear a difference. The few times I thought I did, I did the switch again and decided that I really didn't. |
|
|
|
Post by fbczar on Dec 22, 2022 17:12:43 GMT -5
First... Emotiva WILL finally have a new high-end DAC very soon now. You're going to have to wait another week or two for more information on that... But we WILL start releasing some very interesting details right after the new year. I don't have an actual product release date but it will be soon. And, no, we're not big fans of MQA... Now... as for "the state of the art". R2R DACs simply are NOT current technology for audio applications. The state of the art in R2R DACs for audio applications is really: "If you try really hard... how closely can you get the performance of an expensive R2R DAC to come to the performance of a low cost D-S DAC". "You can get an R2R DAC that sounds reasonably good for a reasonable price if you're not concerned about its technical performance." R2R DACs are still current technology in video applications, and in some other applications where very high speed is important, and linearity past 12 or 16 bits isn't so important... But, in audio applications, D-S DACs consistently offer far better performance at far lower cost. Now I need to address another bit of folklore... If you were even thinking of purchasing an R2R DAC "because you want to avoid things like ringing and 'time-domain inaccuracies' caused by oversampling filters".... Then you need to think again. Virtually all modern R2R DACs that offer even decent audio performance use digital oversampling filters just like D-S DACs do. You may be thinking of "non-oversampling DACs"... Those don't use oversampling filters (which is they're called non-oversampling DACs)... And almost all NOS DACs are also R2R... But the performance of NOS DACs is generally so awful that they shouldn't even be mentioned in the same conversation as "state of the art"... NOS DACs can actually be designed to perform well with high-sample-rate content but not with content at the sample rate CDs use... 44k. This is basically the whole reason why oversampling was invented in the first place. (So, if you want to play a CD on a non-oversampling DAC, and have it come out sounding good, you'll have to oversample or upsample it first anyway... ) Ignoring the fact that D-S DACs generally offer better technical specs and cost less... There is no "inherent way either type of DAC sounds"... (However a manufacturer of an R2R DAC may choose to create a certain sonic signature "to meet their customers' expectations for that sort of DAC".) It's more a matter of the circuitry outside of the DAC chip itself and of the oversampling filter the manufacturer chooses to use. (For example, if you want a "smooth and mellow sound with a slightly rolled off high end" then you'll want a "slow rolloff" filter...) Your question about "high frequency filters" is sort of interesting... There are a wide variety of digital oversampling filter designs and characteristics... And some manufacturers have developed their own special ones instead of using the ones that are included in virtually all modern DAC chips. In fact, most DAC chips offer a choice of several different filters, and the designer can either choose the one they prefer, or allow the user to choose between several different ones. To be quite honest, most of them do an excellent job of doing what they were designed to do, so it's more a question of which one you choose to use. And, in that regard, of choosing the filter option you prefer if you have the option, or of whether you like the one the manufacturer has picked if they choose for you. A "sharp cutoff filter with little ringing" is the most accurate... but some people insist others "sound more natural" with certain types of music. (In fact the differences between filters are pretty subtle... and far less audible than things like how what you're listening to was mastered in the first place.) There is also an "analog reconstruction filter" at the output of a DAC chip. However, to be honest, with an oversampling DAC, these are pretty simple to design and implement, and usually work just fine. (They can be extremely problematic with NOS DACs... which is the main problem with NOS DACs... and the main reason why NOS DACs are not a good idea.) Contemplating DACs... Topping offers their D90se MQA for about $900 with remote volume and XLR outputs. Schiit offers their multi-bit ladder DAC, the Bifrost for about $800 with remote volume and XLR outputs but no MQA. So do I want MQA or not? I'm willing to pay extra for XLR outputs and remote volume. If I remember correctly, the Topping PCM DACs generally measure better than the Schiit ladder ones, but that doesn't really indicate how they sound relative to one another. For now, I'm using my Emotiva Stealth DC1 and it sounds pretty spiffy (although it doesn't do high bit-depth or high sample rate files and it doesn't do MQA). I know that Emotiva is expecting to introduce a new "premium" DAC soon, and I'll probably wait to see what features and price range it has before making a decision. So backing up to basics and ignoring features (like MQA, DSD, and XLR outputs) for the moment, where does the current state of the art in PCM DACs vs. the state of the art in ladder-DACs stand? Does either technology have any valid claims of superiority over the other? Have high frequency filters improved over time, and what filter options (if any) should I be seeking? Thanks - Glenn Given the ancient R2R technology how do you think Schiit is able to produce a DAC that performs like the Yggdrasil? I look forward to the new Emotiva DAC. Will it play DSD? |
|
|
|
Post by leonski on Dec 22, 2022 17:31:09 GMT -5
Original CD player from PHILIPS, sold in this country as a Magnevox FD-1000 featured the TDA1540 DAC or similar number.
This was a 14 BIT DAC but the player DID feature 4x oversampling. this put the turnover frequency WAY above where you'd have phase shift problems.....auduble, at least.
If I could FIND A LASER and maybe someone to fix it? I'd LOVE to compare this relic with any modern player.....
Later the TDA1541 was a full 16 bit and highly regarded until those odd 'single bit' experiments. Sony did some weird stuff, too, but that just sounded AWFUL to even regular people......
I'm not certain most of the new developers can tell the difference between 'cheaper' and 'better'.......All the while making grandiose claims. That maybe 3% of everyone can actually hear...reliably....
|
|
|
|
Post by Boomzilla on Dec 22, 2022 23:33:29 GMT -5
3% difference in DAC chips? Yes, I could think that to be accurate. But as KeithL points out, most of the "sound shaping" is not done in the DAC chip itself, but the analog output amplifiers downstream of the actual DAC. There's far greater than a 3% difference between, for example, and Audio Research output stage and a Cambridge Audio one... |
|
|
|
Post by leonski on Dec 23, 2022 1:14:15 GMT -5
Sorry, boom, what I meant....and didn't write or communicate properly, was that in the good player line? The differences are very small and only (my SHEER GUESS) is that only maybe 3% can reliably tell 'em apart.
After that? Good is good. And who knows what kind of system (and $$$) it would take to tell the difference between 'good' DACs?
My DacMagic+ has several filters. Except for a very few instances? I'd express no preference. About the same result as when I experimented with absolute phase.......
I renew my offer that IF I can get it FIXED? I'll let people listen to my FIRST GEN Magnevox / Philips 14bit player (well reviewed by Stereophile AT THAT TIME) against whatever you are currently useing.
I'm just curious as all get out.....
|
|
KeithL
Administrator
Posts: 10,273
|
Post by KeithL on Dec 23, 2022 11:17:59 GMT -5
Let me start by saying that I've heard the original Schiit Yggdrasil but not the current newer generation versions... Now, that said, the Yggdrasil is pretty much a tour-de-force for an R2R DAC... The original was beautifully made and sounded quite nice... And the newer models seem to offer several different, and interesting, options... And I have no doubt that they also sound very good... I almost bought one of the originals and, if I was looking for a modern state-of-the-art R2R DAC, I would certainly consider the new top model Yggdrasil. HOWEVER, in technical terms, the current top model STILL only claims 20 bits of actual resolution... So, while the Yggdrasil certainly offers "technically decent performance", and I'm sure it sounds great... On specs alone there are far cheaper D-S DACs that offer equal or even significantly better performance. Also, again, while I don't know the details of the new models, the original model DID have oversampling... And one of its important details was a custom-designed and somewhat unusual oversampling filter. And it also had a very good analog section... and a decent USB input section... which was upgraded several times. So there were a lot of interesting details about the Yggdrasil beyond its being an R2R DAC. In fact, I'm not so sure that it sounded good because it was an R2R DAC, rather than in spite of the fact that it was an R2R DAC. And, yes, Yggdrasil is proof that, if you put enough effort into it, you can design an R2R DAC that performs reasonably well. (But that doesn't change the fact that it's easier and cheaper to get comparable performance in a D-S DAC.) Just to be clear here... The CD recording format is limited to 16 bits of resolution... And, on virtually every CD I know of, they've improved the sound even further by sacrificing at least one more bit to dither... And a really well mastered CD can still sound just as good as even a 24/192k digital audio file... Therefore, to be quite blunt, you probably don't actually NEED more than 16 bits of resolution in a DAC... Which means that the significantly better performance numbers that D-S DACs can deliver really are overkill... That's a nice way of saying that a reasonably well designed R2R DAC can audibly be more than good enough... And so the superior performance of a D-S DAC is probably "unnecessary and pretty much inaudible"... But, if you want to actually argue performance, in an audio DAC, there's no question which none will win that comparison. HOWEVER... The other side of that argument is that R2R DACs constitute "a niche product that is marketed to a niche customer segment". In other words, most people don't buy R2R DACs because they deliver better technical performance. Just as people don't buy horses because they offer better mileage, better cargo capacity, or great air conditioning and heat. People either buy an R2R DAC because they are actually convinced that they can hear a difference... Or they buy an R2R DAC because of something they're read or heard from someone else... Unfortunately, because listeners expect an R2R DAC to sound different, this offers manufacturers incentive to deliberately design them to sound different. In other words many boutique DAC manufacturers play to the idea that: "If we make our DAC sound different at least a few customers and reviewers will equate different with better."
(Or the obverse of that: "If we make our DAC sound just like everyone else's, but it costs a lot more, it's going to be difficult to convince people to buy it.") The catch here is that the technology has gotten so good that it's difficult to make something actually sound "different" AND "better".) .................................. Given the ancient R2R technology how do you think Schiit is able to produce a DAC that performs like the Yggdrasil? I look forward to the new Emotiva DAC. Will it play DSD? |
|
KeithL
Administrator
Posts: 10,273
|
Post by KeithL on Dec 23, 2022 11:23:14 GMT -5
A short answer to your short question... Our new DAC has a few features that many folks will find extremely interesting... But DSD is NOT among them. But you'll get a lot more details in a week or two.  Given the ancient R2R technology how do you think Schiit is able to produce a DAC that performs like the Yggdrasil? I look forward to the new Emotiva DAC. Will it play DSD? |
|
)
)
