Hey fellas,
Having recently been made redundant it may be time to downgrade some of my Hifi gear
First to go will be my DAC. It's a Trichord Pulsar DAC that I bought used a few years ago, although it cost £1400 back in 1994.
My thinking is to replace with a modern budget equivalent (CA DacMagic), which sells for around half the price of the Trichord on the used market (£150 vs £300) - I'll pocket the difference into savings.
Reading up the on the DacMagic's specs it looks as if DACs have come a long way the 15 or so year between them so I was hoping the audible difference will not be that great. Before selling it, curiosity has got the better of me and I've decided to see what is in the Trichord component wise to compare:
To my surprise there are two toroidal transformers housed inside the main DAC, as well as an external PSU!
Compared to the DacMagic, there are obviously far more quality component with regards to capacitors and power supplies, although the circuitry seems far simpler.To be expected, given their respective age gap of 16 years:
ImageShack® - Online Photo and Video Hosting
What do you guys reckon? The DacMagic has had plenty of praise since it's release I am hoping the far more advanced chipset/DAC chips will compensate for it's relatively cheaper output stage and wall-wart power supply...
Since there is a 15 year difference between the designs, this seems feasible to me, or am I dreaming?! Since the £150 saving I will get by "downgrading" is fairly modest, I don't want to do something I regret.
Thanks for any advice
Having recently been made redundant it may be time to downgrade some of my Hifi gear
First to go will be my DAC. It's a Trichord Pulsar DAC that I bought used a few years ago, although it cost £1400 back in 1994.My thinking is to replace with a modern budget equivalent (CA DacMagic), which sells for around half the price of the Trichord on the used market (£150 vs £300) - I'll pocket the difference into savings.
Reading up the on the DacMagic's specs it looks as if DACs have come a long way the 15 or so year between them so I was hoping the audible difference will not be that great. Before selling it, curiosity has got the better of me and I've decided to see what is in the Trichord component wise to compare:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
To my surprise there are two toroidal transformers housed inside the main DAC, as well as an external PSU!
An externally hosted image should be here but it was not working when we last tested it.
Compared to the DacMagic, there are obviously far more quality component with regards to capacitors and power supplies, although the circuitry seems far simpler.To be expected, given their respective age gap of 16 years:
ImageShack® - Online Photo and Video Hosting
What do you guys reckon? The DacMagic has had plenty of praise since it's release I am hoping the far more advanced chipset/DAC chips will compensate for it's relatively cheaper output stage and wall-wart power supply...
Since there is a 15 year difference between the designs, this seems feasible to me, or am I dreaming?! Since the £150 saving I will get by "downgrading" is fairly modest, I don't want to do something I regret.
Thanks for any advice
I've not seen the Trichord outside of your photos, but the "dac list" identifies the dac chip used as the PCM69AU. If this is true, I find it rather shocking that such an elaborate unit would be built around such a mediocre chip. The PCM67/69 dacs, which I'd say are about the worst dacs Burr-Brown ever made, are a hybrid design, allegedly using multi-bit ladder dac stage for the high level data and bitstream for the low level. I've personally upgraded the heck out of a Theta dac using this chip, but even after giving it what in my view was far better supply & peripheral support than shown in the Trichord, it was not even as good sounding as a not nearly as upgraded AA Dac in the Box. Just didn't have any liveliness or realistic soundstaging. Boring.
I also find it quite strange in the Trichord that they used such large & elaborate main power supply components, but then, where it counts most, near each chip, they used small & cruddy blob tantalum caps, making the elaborate main supply stuff just about futile. If it were my unit, the FIRST thing I would do is replace every single one of those tantalum caps with much larger value Nichicon KZ, KS or FG lytics, although I have to say I'd be a bit reluctant to put any work into it if the dac chip is accurately i.d.'d as a PCM69 or 67.
When I first read your subject line, I said to myself, "hell yeah, and old dac can beat a new one", but this may be one of the rare cases where the fundmentals of a design make that situation reversed.
I also find it quite strange in the Trichord that they used such large & elaborate main power supply components, but then, where it counts most, near each chip, they used small & cruddy blob tantalum caps, making the elaborate main supply stuff just about futile. If it were my unit, the FIRST thing I would do is replace every single one of those tantalum caps with much larger value Nichicon KZ, KS or FG lytics, although I have to say I'd be a bit reluctant to put any work into it if the dac chip is accurately i.d.'d as a PCM69 or 67.
When I first read your subject line, I said to myself, "hell yeah, and old dac can beat a new one", but this may be one of the rare cases where the fundmentals of a design make that situation reversed.
Regarding your critique of it's liveliness, I can see where you are coming from... It certainly has an unharsh/analogue signature which IME is ideal for taming harsh/bright systems. It's certainly not unpleasant, but it is less forward than other DACs I have owned. In it's favour, the generous power supply does give it impressive bass control.
I recently had a listen to a AVI S2000MD DAC (earlier still, 1991) which utilises the famed Philips TDA1541A S1 DAC with a discrete output. Needless to say it is a firm favourite of mine and I can see why it still has such a reputation! Too bad it is 16 bit / 44.1kHZ only, otherwise I would of chosen that for my system.
I recently had a listen to a AVI S2000MD DAC (earlier still, 1991) which utilises the famed Philips TDA1541A S1 DAC with a discrete output. Needless to say it is a firm favourite of mine and I can see why it still has such a reputation! Too bad it is 16 bit / 44.1kHZ only, otherwise I would of chosen that for my system.
hi mate the trichord research Dac is in a total different leauge to the theta. itoo have modified the theta dac
Graham fowler builds hi quality phono stages and guess what the analogue outputstage on the trichord is I excellent 👌
I have spent a week fighting a DAC upgrade.
My source is a dedicated Windows 10 PC. Mostly CDs on a SSD, but want to investigate HD streaming.
My DAC is a very old MUSE.
So I bought a SMSL Scanskrit 10 based on objective testing reviews. NOWHERE did the instructions say you need to download drivers ( you must if using over 24 bit/96K) As I have my laptop setup with my Focusrite 2i2 for verifying re-caping my amplifier was valid, I decided ot see if I could measure the THD of the new DAC, my old, Focusrite loopback baseline, PC DAC and my Woofer tester.
All measures were 24 bit/96K as that is all Windows supports natively.
Using the FOCUSRITE drivers, the loopback measured .001% with a noise floor around -120 dB. USB powered, but clean for power harmonics.
My old Muse measured around .025% and .008% across the bandwidth with considerable power artifacts. Improving the external supply by adding a LT1083 regulator eliminated any measurable power artifacts. No change to THD.
So I bought a SMSL and tested it. It had two terrible noise spikes suggesting it was defective. Distortion measured roughly .03% THD+ noise. As this is more than 30 times my test loopback, I suggest it is valid. I saw no AC artifacts even being USB powered. Native Windows drivers. Sent it back.
Bought a Topping e-30. No spikes. Power artifacts were horrible USB powered so whipped up a clean regulator and again, the dropped below the noise floor. Excellent.
THD was still way higher than spec and 10 times my baseline. As it refused to play anything other than 44.1, I dug in and found Windows settings as well as their driver. Installed, and now it switched rate and depth, but no change in distortion.
What I learned is SMSL and Topping may make good hardware, but their instructions and business acumen are terrible. Over on the test sites, the denigrate me for not being a super experienced hack the internals of Windows expert when what I bought is sold as a consumer plug and play product. Lost total respect for them.
Searched the WEB and found many suggestions on changes to Windows settings. Many address the likelihood of glitches and pops, but none had any effect on distortion.
The E-30 dashboard does have settings for filtering, but no mention of why to choose one setting over another. I suspect they may.
No mention if it does force USB profile 2 async over native profile 1. No discussion at all on WASAPI. ( seems only some players support it, JRiver, Foobar etc) Why a consumer product would support 32 bit/768 Khz I do not know unless it is just how they get advertising specs as I find no streaming sources deeper than 24 bits, 192K. ( which Windows supports natively, just not with async protocol over USB.
My suspicion is the vanishing low .0003 or less specs are only for 32 bit, 768 Khz sampling. For one thing. 16 bits wil by definition all distortion woudl be below the noise floor. Where I am disappointed is why they can't do .005 to .001 or so. Is it due to the filtering that supports the deep/fast rates so compromises mere CD quality? If so, then simply it is not the right DAC for me.
Neither SMSL not Topping has answered questions, but I am giving them time. I have also asked Cambridge (MagicDac100 is old, but respected) and Schiit if my assumptions are correct. If so, they are both better choices for me.
FWIW, the Focusrite Scarlett 2i2 preforms way better than my Muse and better than the problems with the new ones. It also has balanced outputs, so for a balanced DAC going to 24 bit/192K, it is a bargain and why I will likely move from my test kit to my stereo! Either that or buy the Shiiit Modi 3. I was looking at e-bay and used are selling for way over half price so new it would be. ( OBE model of a $100 DAC going for $70!)
Stupidly complicated. One consistent thread is that they don't have these issues on RPi or Mac. If I had not invested in a new Windows touch screen/SSD/all in one, possibly building a Linux box on a RPi would have made a better music server. I was hoping 10 or more years woud allow me an improvement of an old BB DAC. Not yet.
My source is a dedicated Windows 10 PC. Mostly CDs on a SSD, but want to investigate HD streaming.
My DAC is a very old MUSE.
So I bought a SMSL Scanskrit 10 based on objective testing reviews. NOWHERE did the instructions say you need to download drivers ( you must if using over 24 bit/96K) As I have my laptop setup with my Focusrite 2i2 for verifying re-caping my amplifier was valid, I decided ot see if I could measure the THD of the new DAC, my old, Focusrite loopback baseline, PC DAC and my Woofer tester.
All measures were 24 bit/96K as that is all Windows supports natively.
Using the FOCUSRITE drivers, the loopback measured .001% with a noise floor around -120 dB. USB powered, but clean for power harmonics.
My old Muse measured around .025% and .008% across the bandwidth with considerable power artifacts. Improving the external supply by adding a LT1083 regulator eliminated any measurable power artifacts. No change to THD.
So I bought a SMSL and tested it. It had two terrible noise spikes suggesting it was defective. Distortion measured roughly .03% THD+ noise. As this is more than 30 times my test loopback, I suggest it is valid. I saw no AC artifacts even being USB powered. Native Windows drivers. Sent it back.
Bought a Topping e-30. No spikes. Power artifacts were horrible USB powered so whipped up a clean regulator and again, the dropped below the noise floor. Excellent.
THD was still way higher than spec and 10 times my baseline. As it refused to play anything other than 44.1, I dug in and found Windows settings as well as their driver. Installed, and now it switched rate and depth, but no change in distortion.
What I learned is SMSL and Topping may make good hardware, but their instructions and business acumen are terrible. Over on the test sites, the denigrate me for not being a super experienced hack the internals of Windows expert when what I bought is sold as a consumer plug and play product. Lost total respect for them.
Searched the WEB and found many suggestions on changes to Windows settings. Many address the likelihood of glitches and pops, but none had any effect on distortion.
The E-30 dashboard does have settings for filtering, but no mention of why to choose one setting over another. I suspect they may.
No mention if it does force USB profile 2 async over native profile 1. No discussion at all on WASAPI. ( seems only some players support it, JRiver, Foobar etc) Why a consumer product would support 32 bit/768 Khz I do not know unless it is just how they get advertising specs as I find no streaming sources deeper than 24 bits, 192K. ( which Windows supports natively, just not with async protocol over USB.
My suspicion is the vanishing low .0003 or less specs are only for 32 bit, 768 Khz sampling. For one thing. 16 bits wil by definition all distortion woudl be below the noise floor. Where I am disappointed is why they can't do .005 to .001 or so. Is it due to the filtering that supports the deep/fast rates so compromises mere CD quality? If so, then simply it is not the right DAC for me.
Neither SMSL not Topping has answered questions, but I am giving them time. I have also asked Cambridge (MagicDac100 is old, but respected) and Schiit if my assumptions are correct. If so, they are both better choices for me.
FWIW, the Focusrite Scarlett 2i2 preforms way better than my Muse and better than the problems with the new ones. It also has balanced outputs, so for a balanced DAC going to 24 bit/192K, it is a bargain and why I will likely move from my test kit to my stereo! Either that or buy the Shiiit Modi 3. I was looking at e-bay and used are selling for way over half price so new it would be. ( OBE model of a $100 DAC going for $70!)
Stupidly complicated. One consistent thread is that they don't have these issues on RPi or Mac. If I had not invested in a new Windows touch screen/SSD/all in one, possibly building a Linux box on a RPi would have made a better music server. I was hoping 10 or more years woud allow me an improvement of an old BB DAC. Not yet.
Last edited:
tvrgeek,
Which version of the SMSL Sanskrit 10th did you get? According to ASR, the first rev of the DAC has an issue of being underpowered with the USB port for data+power, but runs much better with external power connected.
Review and Measurements of SMSL Sanskrit 10th DAC | Audio Science Review (ASR) Forum
Which version of the SMSL Sanskrit 10th did you get? According to ASR, the first rev of the DAC has an issue of being underpowered with the USB port for data+power, but runs much better with external power connected.
Review and Measurements of SMSL Sanskrit 10th DAC | Audio Science Review (ASR) Forum
Verbose, but I am really looking for help!
I saw that. It was whatever was shipped this month via Amazon. However, it is advertised and sold to run from a single cable.
I accept I am no deep internal expert on the Windows audio path, but none of the "experts" so far have given me any help other than to say I am an idiot. That may be true, but it does not mean I can't plug in a box. ( Former engineering lab technician, retired computer scientist so not totally clueless) I have searched the internet for audio tweaks and most involve preventing drops in power or higher priority interrupts. That would not be the cause of my issues.
Simple. Hook up my Scarlett 2i2 in loopback. Run RightMark. Verify levels with a Fluke.
Notice noise floor. Measure 1KHz ( seems to be a pseudo standard for specs) and record. Measured @ 24/96. If supplied USB cable, I use the supplied one. All analog DAC to Focusrite using the same 6 inch RCA to PL-55. Levels into the I/O box remained the same. About -4dB as measured by RightMark. Full volume if the DAC was adjustable, Muse is not, Topping was in DAC mode.
Focusrite Loopback: .003% THD + noise. No power artifacts seen. Focusrite drivers. Reading about noise over USB, I got a galvanic isolator, put it in front of a USB 3 hub so I could provide my clean 5V supply. ( 2A linear, LT1085 based) . I caw no difference in the loopback or the MUSE test. Spec is .002, so real world measurement is quite acceptable considering that is through both the A2D and D2A. That tells me either my ports are clean (doubtful) or the Muse and Focusrite are well designed to reject any such issues. It confirms the D2A is pretty good.
Muse (old) Native driver. .034% and had a -80dB 60 Hz spike with harmonics. Built a pre-regulator and now -110 dB. Virtually clean past second and third. No high order at all. Original spec was .02% but it is 10 years old.
SMSL Native driver. ( Did not know they existed yet, nothing in the instructions) roughly .02 THD + noise. ( I did not save the graph) , but had these two big fixed spikes. So two problems. Higher order harmonics blended with noise floor pretty quickly and ugliness around fundamental well below audible though higher than the Muse.
E-30. Topping driver. Spikes were gone, but horrible crap from 120 to 1K. External power supply cleaned that up completely. But again, THD + Noise at .029%. Many artifacts exceeding -120 dB all across the spectrum. More spikes between fundamental and second harmonic, but @ -115, so worse than the others, but not likely audible. ,
There is clearly a reason why these two products read so much higher than the DAC side of my Focusrite 2i2. I know I am not going to get .0003 running 16 bit, 44.1Khz as you flat run out of bits so the printed specifications are misleading being only valid at likely 32 bits and some rate that is irrelevant for playing music. I know distortion due to the filtering artifacts will give a little noise just around the window. If there is some magic setting, magic power, magic cable, magic software, Please speak up! I have not returned the Topping yet. The E-30 driver UI has settings for filtering, but says nothing about them. I do not know what the SMSL driver UI has for settings. If there is an optimum for low rate, shallow bits that solves the distortion, I would love to hear. SMSL is already returned.
If it just a matter of configuration or instructions, then both these companies fail. Instructions are cheap. Post them on the WEB. Packaged instruction can just say " go RTFI at xxx.com". No, they don't even tell you you need the driver. That is just flat inexcusable. These are consumer products, sold through a consumer path, for use on consumer environments. If not plug and play, it is their responsibility to say so.
Next thing I learned yesterday. Schiit and JDS provided feedback on my difficulties when I inquired about their products within a day and a half. Neither SMSL nor Topping have responded at all, not even a " we got your question" automated e-mail. Cambridge at least sent the automated.
So, unless I get some help that cleans the TOPPING up to roughly the same as my IO box, I wil give either the Modi 3 or OL DAC a try. Leaning to the Schiit as it runs USB audio mode 2 with their driver. Both are appropriately targeted with 24/192 as highest rates, settings that are as high as any streaming or other consumer accusable audio. 32/768 is totally bogus in a consumer playback device so for that to be as supplied tells me something about the companies. If wondering why not the Focusrite, I use it for speaker measurements so do not want to pull the stereo apart al the time. A gen 3 in not out of the possibility though.
So, if there is something I can do that lets a product perform in the real world only 10 times worse than it's spec, and similar to my three generation IO box, please speak up. There is an explanation out there. A $3000 USB cable is not on the table, neither is building a Linux server, though I might research that. I do have a $11 Monoprice USB A to B cable on it's way, but that would not help with the Topping or SMSL as they use mini. If it is just their cable that is causing the issue, then it is still a defective product.
I notice ASR measured @ 44.1, so that leaves the question why my measurement was 40 dB worse. If their result was true, it would have been 10dB below my loopback measurement so I should not have seen much difference if at all.
I saw that. It was whatever was shipped this month via Amazon. However, it is advertised and sold to run from a single cable.
I accept I am no deep internal expert on the Windows audio path, but none of the "experts" so far have given me any help other than to say I am an idiot. That may be true, but it does not mean I can't plug in a box. ( Former engineering lab technician, retired computer scientist so not totally clueless) I have searched the internet for audio tweaks and most involve preventing drops in power or higher priority interrupts. That would not be the cause of my issues.
Simple. Hook up my Scarlett 2i2 in loopback. Run RightMark. Verify levels with a Fluke.
Notice noise floor. Measure 1KHz ( seems to be a pseudo standard for specs) and record. Measured @ 24/96. If supplied USB cable, I use the supplied one. All analog DAC to Focusrite using the same 6 inch RCA to PL-55. Levels into the I/O box remained the same. About -4dB as measured by RightMark. Full volume if the DAC was adjustable, Muse is not, Topping was in DAC mode.
Focusrite Loopback: .003% THD + noise. No power artifacts seen. Focusrite drivers. Reading about noise over USB, I got a galvanic isolator, put it in front of a USB 3 hub so I could provide my clean 5V supply. ( 2A linear, LT1085 based) . I caw no difference in the loopback or the MUSE test. Spec is .002, so real world measurement is quite acceptable considering that is through both the A2D and D2A. That tells me either my ports are clean (doubtful) or the Muse and Focusrite are well designed to reject any such issues. It confirms the D2A is pretty good.
Muse (old) Native driver. .034% and had a -80dB 60 Hz spike with harmonics. Built a pre-regulator and now -110 dB. Virtually clean past second and third. No high order at all. Original spec was .02% but it is 10 years old.
SMSL Native driver. ( Did not know they existed yet, nothing in the instructions) roughly .02 THD + noise. ( I did not save the graph) , but had these two big fixed spikes. So two problems. Higher order harmonics blended with noise floor pretty quickly and ugliness around fundamental well below audible though higher than the Muse.
E-30. Topping driver. Spikes were gone, but horrible crap from 120 to 1K. External power supply cleaned that up completely. But again, THD + Noise at .029%. Many artifacts exceeding -120 dB all across the spectrum. More spikes between fundamental and second harmonic, but @ -115, so worse than the others, but not likely audible. ,
There is clearly a reason why these two products read so much higher than the DAC side of my Focusrite 2i2. I know I am not going to get .0003 running 16 bit, 44.1Khz as you flat run out of bits so the printed specifications are misleading being only valid at likely 32 bits and some rate that is irrelevant for playing music. I know distortion due to the filtering artifacts will give a little noise just around the window. If there is some magic setting, magic power, magic cable, magic software, Please speak up! I have not returned the Topping yet. The E-30 driver UI has settings for filtering, but says nothing about them. I do not know what the SMSL driver UI has for settings. If there is an optimum for low rate, shallow bits that solves the distortion, I would love to hear. SMSL is already returned.
If it just a matter of configuration or instructions, then both these companies fail. Instructions are cheap. Post them on the WEB. Packaged instruction can just say " go RTFI at xxx.com". No, they don't even tell you you need the driver. That is just flat inexcusable. These are consumer products, sold through a consumer path, for use on consumer environments. If not plug and play, it is their responsibility to say so.
Next thing I learned yesterday. Schiit and JDS provided feedback on my difficulties when I inquired about their products within a day and a half. Neither SMSL nor Topping have responded at all, not even a " we got your question" automated e-mail. Cambridge at least sent the automated.
So, unless I get some help that cleans the TOPPING up to roughly the same as my IO box, I wil give either the Modi 3 or OL DAC a try. Leaning to the Schiit as it runs USB audio mode 2 with their driver. Both are appropriately targeted with 24/192 as highest rates, settings that are as high as any streaming or other consumer accusable audio. 32/768 is totally bogus in a consumer playback device so for that to be as supplied tells me something about the companies. If wondering why not the Focusrite, I use it for speaker measurements so do not want to pull the stereo apart al the time. A gen 3 in not out of the possibility though.
So, if there is something I can do that lets a product perform in the real world only 10 times worse than it's spec, and similar to my three generation IO box, please speak up. There is an explanation out there. A $3000 USB cable is not on the table, neither is building a Linux server, though I might research that. I do have a $11 Monoprice USB A to B cable on it's way, but that would not help with the Topping or SMSL as they use mini. If it is just their cable that is causing the issue, then it is still a defective product.
I notice ASR measured @ 44.1, so that leaves the question why my measurement was 40 dB worse. If their result was true, it would have been 10dB below my loopback measurement so I should not have seen much difference if at all.
Some progress on the Toppin E-30. I reduced the output level by 15 dB via it's dashboard.
Distortion dropped from .028% to .0052% Only measurable was second order. Noise floor came up as expected to all of -120 dB.
Input to my Focusrite is 60K, so that should not be a loading factor.
It does not like the isolator, powered line or not.
At that level, it should SOUND just fine and is certainly a useable device, 10dB besting my old Muse, but still a far cry from the published, OEM or test site values.
So now, do I sit tight, or send it back and get the Schiit? Tough call. Going to at least instlal it in the system and listen for a bit. Amazon did give me more than a month.
Distortion dropped from .028% to .0052% Only measurable was second order. Noise floor came up as expected to all of -120 dB.
Input to my Focusrite is 60K, so that should not be a loading factor.
It does not like the isolator, powered line or not.
At that level, it should SOUND just fine and is certainly a useable device, 10dB besting my old Muse, but still a far cry from the published, OEM or test site values.
So now, do I sit tight, or send it back and get the Schiit? Tough call. Going to at least instlal it in the system and listen for a bit. Amazon did give me more than a month.
A tough call? Buy plastic random Chinese garbage a half-dozen times, or California-grown hardware from a California company once?
It also decides to not always sync with the PC on a reboot even when set as the default and when Media Server starts. Just sits t? ne, Plug and not always play.
Decided to replace everything. Schiit Vidar amp, either the Modi-3 DAC and Asgard preamp, or the DAC inside the pre. option. ( waiting to hear about the differences in performance and if I care) May add their Loki tone control box. Then my new XKitx crossovers and the only old piece wil be my O-Audio Bash. Other option is the JDS Atom and a Parasound. I wish I could hear some of this stuff, but it seems around here, they think Pyle id hi-fi but want to sell you an $800 remote and $200 to program it.
Decided to replace everything. Schiit Vidar amp, either the Modi-3 DAC and Asgard preamp, or the DAC inside the pre. option. ( waiting to hear about the differences in performance and if I care) May add their Loki tone control box. Then my new XKitx crossovers and the only old piece wil be my O-Audio Bash. Other option is the JDS Atom and a Parasound. I wish I could hear some of this stuff, but it seems around here, they think Pyle id hi-fi but want to sell you an $800 remote and $200 to program it.
With the proviso that I am a skeptic...
The question boils down to, or should, something more like:
"Can I hear a difference between budget DAC X and boutique DAC y?"
I'm one of the minority who believes that if the comparison is fair then there is no audible difference between cheap and expensive.
What does a fair comparison mean? The units should be reasonably matched in specs. Unless one unit had, for example, a markedly higher noise floor than another at a given level, one might be able to pick between the two in A/B test. The specs (frequency response, etc.) should be fairly close. Even the cheapest DACs have (claimed) specs far in excess of what's detectably by the ear, and often even by measuring gear.
The test has to be blind, ideally double blind. I think you'd find that all the perceived superiority of that $1000 DAC in the walnut case, with the gold-plated connectors, the one recommended by your audio snob buddies,vanishes for some reason when you can't tell if it or the $50 DAC is playing.
Frankly, I'd be most mpressed by a fairly conducted, double-bind listening test between two or more components. The fact that you almost never see such a thing says a lot about the lack of objectivity in this hobby.
The question boils down to, or should, something more like:
"Can I hear a difference between budget DAC X and boutique DAC y?"
I'm one of the minority who believes that if the comparison is fair then there is no audible difference between cheap and expensive.
What does a fair comparison mean? The units should be reasonably matched in specs. Unless one unit had, for example, a markedly higher noise floor than another at a given level, one might be able to pick between the two in A/B test. The specs (frequency response, etc.) should be fairly close. Even the cheapest DACs have (claimed) specs far in excess of what's detectably by the ear, and often even by measuring gear.
The test has to be blind, ideally double blind. I think you'd find that all the perceived superiority of that $1000 DAC in the walnut case, with the gold-plated connectors, the one recommended by your audio snob buddies,vanishes for some reason when you can't tell if it or the $50 DAC is playing.
Frankly, I'd be most mpressed by a fairly conducted, double-bind listening test between two or more components. The fact that you almost never see such a thing says a lot about the lack of objectivity in this hobby.
Last edited:
that is logical - - I do think differences in the analog section exist. A DAC with transformer output and no "active" device may be subjectively different if not better than DAC with tube, or opamp audio section, and even capacitor choice may influence how a single tone may be perceived. Plus whatever they feed in the audio chain contributes. Sometimes one component such as a DAC will just "sound" better with another amplifier, etc. than what one would expect.
A stepped attenuator, vs ALPs potentiometer vs active high level stage with electrolytic caps in the signal path may be closely matched but interpreted as sounding different from one another. I don't know how much of that is biased. 30 years ago with ~1dB matching, I preferred the ALPs pot to the stepped attenuator w, Corning resistors which sounded somewhat "metallic" in comparison. The line stage of a Rockford preamplifier had a bloated bass characteristic
A-B with good matching would be difficult to detect I think in a lot of cases.
With modern digital techniques and much sound being delivered I think via YouTube and the OPUS codec, there's plenty of fodder for DACs. A cheap Topping model is pretty good imo. I'm afraid of SMSL and DOUK as had products fail in a matter of hours.
Its interesting to read Audiosciencereview and Superbestaudiofriends
A stepped attenuator, vs ALPs potentiometer vs active high level stage with electrolytic caps in the signal path may be closely matched but interpreted as sounding different from one another. I don't know how much of that is biased. 30 years ago with ~1dB matching, I preferred the ALPs pot to the stepped attenuator w, Corning resistors which sounded somewhat "metallic" in comparison. The line stage of a Rockford preamplifier had a bloated bass characteristic
A-B with good matching would be difficult to detect I think in a lot of cases.
With modern digital techniques and much sound being delivered I think via YouTube and the OPUS codec, there's plenty of fodder for DACs. A cheap Topping model is pretty good imo. I'm afraid of SMSL and DOUK as had products fail in a matter of hours.
Its interesting to read Audiosciencereview and Superbestaudiofriends
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.