Hi jstang,
I could do a destructive HV test using a HV unit from a multi wave oscillator (up to 1 Mega volt)
It will probably survive a tube output, but make sure to sand paper the PCB edges carefully, in order to make sure there is no short or DC leakage current.
If you need more space in your chassis, simply remove the tube output stage. I tested various tube output stages with the very transparent sounding ISD player, and the tube sound coloration is very audible now. I tested single tube stage a la Lampizator, balanced input stage with modified Broskie buffer and an Aikido output stage.
Tubes are fine for covering up things that should have been tackled at the source.
TDA1541A output however is problematic as output compliance can't be exceeded very much without increasing THD. With the TDA1543 output compliance is much more relaxed, due to a different on-chip output stage. This chip however requires clean reference voltages for both passive I/V resistors, and a clean power supply (poor PSRR).
I am still thinking about best possible TDA1541A output stage,
My current favorite is a 4 JFET balanced, semi complementary active stage (gain approx. 50x). All source resistors need to be non-inductive wire wound types. No current sources allowed, using hybrid chokes in series with resistor instead. The DAC ac output signal is tapped as close as possible to the I/V resistor (approx. 15 Ohms) terminals using a Kelvin coupling. never underestimate the sound quality degradation that can be induced by incorrect grounding, especially in digital playback equipment.
For best clarity / refinement, only a single TDA1541A can be used. Multiple paralleled DAC chips never change output samples at exactly the same moment (tolerances), this introduces extra sample timing jitter, something we just want to avoid at all cost. Multiple DAC chips also means multiple clock loads, so jitter performance is likely worse compared to a single clock load (every attached clock load, plus required wiring, usually leads to increased jitter levels). Multiple DAC chips also increase interference levels (summed noise from on-chip logic circuits).
That's why I use single DAC chip designs now.
I could do a destructive HV test using a HV unit from a multi wave oscillator (up to 1 Mega volt)
It will probably survive a tube output, but make sure to sand paper the PCB edges carefully, in order to make sure there is no short or DC leakage current.
If you need more space in your chassis, simply remove the tube output stage. I tested various tube output stages with the very transparent sounding ISD player, and the tube sound coloration is very audible now. I tested single tube stage a la Lampizator, balanced input stage with modified Broskie buffer and an Aikido output stage.
Tubes are fine for covering up things that should have been tackled at the source.
TDA1541A output however is problematic as output compliance can't be exceeded very much without increasing THD. With the TDA1543 output compliance is much more relaxed, due to a different on-chip output stage. This chip however requires clean reference voltages for both passive I/V resistors, and a clean power supply (poor PSRR).
I am still thinking about best possible TDA1541A output stage,
My current favorite is a 4 JFET balanced, semi complementary active stage (gain approx. 50x). All source resistors need to be non-inductive wire wound types. No current sources allowed, using hybrid chokes in series with resistor instead. The DAC ac output signal is tapped as close as possible to the I/V resistor (approx. 15 Ohms) terminals using a Kelvin coupling. never underestimate the sound quality degradation that can be induced by incorrect grounding, especially in digital playback equipment.
For best clarity / refinement, only a single TDA1541A can be used. Multiple paralleled DAC chips never change output samples at exactly the same moment (tolerances), this introduces extra sample timing jitter, something we just want to avoid at all cost. Multiple DAC chips also means multiple clock loads, so jitter performance is likely worse compared to a single clock load (every attached clock load, plus required wiring, usually leads to increased jitter levels). Multiple DAC chips also increase interference levels (summed noise from on-chip logic circuits).
That's why I use single DAC chip designs now.
I live in the same town as Tesla did in NJ... I could try to check his basement to see if he left any mega Volt units behind....
I agree that less parts are almost always better sounding....
On the multi DAC error, moving to S1s provided more improvement than I had expected. Guess it helped remove the added errors.
I was doing some scope testing the other night of a single dac and my 4 DACs looking at a 18K hz test signal on a CD. That really sold me on the 4 DAC linear interpolation benefits.
I am sold on that trade off... the up side of more parts is worth more than the down sides of more parts to my ears. AT LEAST UNTIL YOU WRITE A FEW MORE POSTS AND CHANGE MY MIND....
johnk
I agree that less parts are almost always better sounding....
On the multi DAC error, moving to S1s provided more improvement than I had expected. Guess it helped remove the added errors.
I was doing some scope testing the other night of a single dac and my 4 DACs looking at a 18K hz test signal on a CD. That really sold me on the 4 DAC linear interpolation benefits.
I am sold on that trade off... the up side of more parts is worth more than the down sides of more parts to my ears. AT LEAST UNTIL YOU WRITE A FEW MORE POSTS AND CHANGE MY MIND....
johnk
Hi jstang,
I could do a destructive HV test using a HV unit from a multi wave oscillator (up to 1 Mega volt)
It will probably survive a tube output, but make sure to sand paper the PCB edges carefully, in order to make sure there is no short or DC leakage current.
If you need more space in your chassis, simply remove the tube output stage. .............
..................
For best clarity / refinement, only a single TDA1541A can be used. Multiple paralleled DAC chips never change output samples at exactly the same moment (tolerances), this introduces extra sample timing jitter, something we just want to avoid at all cost. Multiple DAC chips also means multiple clock loads, so jitter performance is likely worse compared to a single clock load (every attached clock load, plus required wiring, usually leads to increased jitter levels). Multiple DAC chips also increase interference levels (summed noise from on-chip logic circuits).
That's why I use single DAC chip designs now.
What?!
Hi all
It's not easy for a DIY man to take actions an try to do something for yourself if you follow Mr. -ecdesign- I was, at first, very confused when I saw these last posts.
All that effort on the multi DAC reference and the "perfect sound" in this ultimate thread, is not any more? Amasing to say the least. All the interpolation talk and stuff? gone? It did sound good or didn't it ? 
I am very amased and confused how the earth you get the energy to do all this stuff -ecdesign- without going to the end, the nearvous breakdown
Please take all this as a complimant -ecdesign-. Please keep it up. You are one of a kind that are an extinct specie
I am myself a very dedicated person, sometimes, with all my interrests but you are absolutely one of a kind.
Hats off.
Hi all
It's not easy for a DIY man to take actions an try to do something for yourself if you follow Mr. -ecdesign-
I was, at first, very confused when I saw these last posts.All that effort on the multi DAC reference and the "perfect sound" in this ultimate thread, is not any more? Amasing to say the least. All the interpolation talk and stuff? gone?
It did sound good or didn't it ? I am very amased and confused how the earth you get the energy to do all this stuff -ecdesign- without going to the end, the nearvous breakdown
Please take all this as a complimant -ecdesign-. Please keep it up. You are one of a kind that are an extinct specie
I am myself a very dedicated person, sometimes, with all my interrests but you are absolutely one of a kind.
Hats off.
I think we need to keep the 4x upsampling and keep doing a opamp i/v. ecdesign says the truth and especially since he did so many research. Still a twin dac could be interesting for real balanced output ... I also wonder if it is worth keeping the tda1541 because of the new high sampling 96khz music , it might be worth to shut the door and concentrate efforts on the AKM for example. It can sound good too... I don<t know any tda1541 lovers who prefered the new dacs???? The cirus ones are very good, better bass than the tda1541, but tda1541 wins in different textures and separation, GOSH its hard...
Hi Gabdx,
I have a pair of AKM4396 DAC chips (originally bought to replace my M-Audio superDAC's (the reason to join DIYaudio) AKM4393) conveniently stored on its original bags...as John has explained, it is so difficult to implement a jitter free-noise free environment for the humble TDA1543, imagine how difficult it will be for the "HiResolution" DACs...they always sound sterile in comparison.
One of the problems with digital formats lies in the mastering step: once it is badly executed even the best of the gears won't do miracles.
Cheers,
M
I have a pair of AKM4396 DAC chips (originally bought to replace my M-Audio superDAC's (the reason to join DIYaudio) AKM4393) conveniently stored on its original bags...as John has explained, it is so difficult to implement a jitter free-noise free environment for the humble TDA1543, imagine how difficult it will be for the "HiResolution" DACs...they always sound sterile in comparison.
One of the problems with digital formats lies in the mastering step: once it is badly executed even the best of the gears won't do miracles.
Cheers,
M
I like the ending on this song, well rendered !, (Love those old recordings were the voice is 10e10 louder than everything ) it seems to me that its not like until the 1990 that recordings started to be less unbalanced, still many exceptions.
Looks very ambitious to replace the dac on the M-audio !!! ,
Could we could find a way to split the 24 bit 96khz signal between two tda1541 ^.^
Looks very ambitious to replace the dac on the M-audio !!! ,
Could we could find a way to split the 24 bit 96khz signal between two tda1541 ^.^
Hi gabdx,
Upsampling "adds" nothing, it's similar to scaling a low resolution photograph in an attempt to get more detail. It's even likely that errors are added in the upsampling process. Upsampling is used for smoothing the signal (not adding any extra information) This is often required to prevent problems with flawed audio sets.
In my humble opinion even the very best (discrete) audio OP-amps available today, are not the best solution for I/V conversion, nor for any other audiophile analogue signal processing. They lack dynamic resolution and sound rather analytical.
I used both, discrete (DI4MJ) and integrated (OPA627, LM4562, THS4031 and so on) OP-amps, as documented on this thread, and now I am not using them anymore, I don't even use integrated voltage regulators for similar reasons. I/V OP-amps also require different properties (settling time, bandwidth) compared to OP-amps used for signal amplification.
OP-amps have / need global feedback, when comparing circuits with and without global feedback, the circuits without global feedback have higher THD (less correction), but offer way higher dynamic resolution and refinement. Think of well designed tube power amps that have high THD of around 1%, yet can still produce magnificent perceived sound quality (A300B SE for example). Striving for lowest possible THD may be focusing on the wrong parameter.
The fewer components in the signal path, the higher the dynamic resolution and refinement. Perceived sound quality also depends heavily on circuit dynamic resolution, this is probably much more important than vanishing low THD. Keep in mind that even the best speakers / headphones introduce way higher THD than DAC chips with worst THD specs.
Unfortunately, modern DAC chips are designed for lowest possible THD and highest possible bit resolution, they are definitely NOT designed for achieving highest possible dynamic resolution and refinement.
They use very high on-chip frequencies (required for higher bit resolution and driving the Delta Sigma modulators). This increases on-chip ground-bounce, and interference. Modern DAC chip (on-chip) analogue output circuit usually consists of switched-capacitor filters, containing (OP-amp) buffers, so more than plenty components in the analogue signal path, and greatly reduced dynamic resolution. Worst of all, this is integrated on the DAC chip and cannot be corrected externally.
About bit resolution, consider a 96/24 source, time resolution is increased by approx. factor 2 while bit resolution is increased by factor 256, isn't this a bit odd? one would expect factor 256 for both, resulting in 24596/24. It would be more logical to use 96/17, 192/18, and 384/19. I think 24 or even (theoretical) 32 bit resolution is complete overkill and serves no practical purpose whatsoever in digital audio PLAYBACK systems.
I have reason to believe that not bit resolution but time resolution is most important parameter in digital audio playback (jitter sensitivity is already a hint). So striving for ever higher bit resolution is not going to help. Besides I yet have to see the playback system capable of resolving true 24 bit resolution with 0.0001% THD at the speaker output.
TDA1541A performance, like all other DAC chips, depends heavily on application. One can only make valid comments about DAC chip performance when comparing DAC chips that are used in best possible application, this is seldom the case. In other words, you are usually listening to a fraction of DAC chip potential. Based on listening tests, I would say most applications are only able to extract a fraction of DAC chip potential.
This means that a TDA1543 could have the potential to challenge modern DAC chips (perceived sound quality!), but only when used in a super optimized circuit. I don't dare to think what's possible with a super tuned TDA1541A-based DAC. I am working on this, based on the extraordinary performance (perceived sound quality) possible with a single TDA1543.
I also have to add that I use a different approach by applying large bandwidth signal processing, all the way up to the speakers.
Then have speaker crossocer filter, speaker bandwidth, and human auditory system perform required brickwall filtering.
Although this approach also has drawbacks (non-linear response of the connected audio components). It's easier to achieve much better perceived sound quality this way, instead of going the OS / digital brickwall filter route.
Upsampling "adds" nothing, it's similar to scaling a low resolution photograph in an attempt to get more detail. It's even likely that errors are added in the upsampling process. Upsampling is used for smoothing the signal (not adding any extra information) This is often required to prevent problems with flawed audio sets.
In my humble opinion even the very best (discrete) audio OP-amps available today, are not the best solution for I/V conversion, nor for any other audiophile analogue signal processing. They lack dynamic resolution and sound rather analytical.
I used both, discrete (DI4MJ) and integrated (OPA627, LM4562, THS4031 and so on) OP-amps, as documented on this thread, and now I am not using them anymore, I don't even use integrated voltage regulators for similar reasons. I/V OP-amps also require different properties (settling time, bandwidth) compared to OP-amps used for signal amplification.
OP-amps have / need global feedback, when comparing circuits with and without global feedback, the circuits without global feedback have higher THD (less correction), but offer way higher dynamic resolution and refinement. Think of well designed tube power amps that have high THD of around 1%, yet can still produce magnificent perceived sound quality (A300B SE for example). Striving for lowest possible THD may be focusing on the wrong parameter.
The fewer components in the signal path, the higher the dynamic resolution and refinement. Perceived sound quality also depends heavily on circuit dynamic resolution, this is probably much more important than vanishing low THD. Keep in mind that even the best speakers / headphones introduce way higher THD than DAC chips with worst THD specs.
Unfortunately, modern DAC chips are designed for lowest possible THD and highest possible bit resolution, they are definitely NOT designed for achieving highest possible dynamic resolution and refinement.
They use very high on-chip frequencies (required for higher bit resolution and driving the Delta Sigma modulators). This increases on-chip ground-bounce, and interference. Modern DAC chip (on-chip) analogue output circuit usually consists of switched-capacitor filters, containing (OP-amp) buffers, so more than plenty components in the analogue signal path, and greatly reduced dynamic resolution. Worst of all, this is integrated on the DAC chip and cannot be corrected externally.
About bit resolution, consider a 96/24 source, time resolution is increased by approx. factor 2 while bit resolution is increased by factor 256, isn't this a bit odd? one would expect factor 256 for both, resulting in 24596/24. It would be more logical to use 96/17, 192/18, and 384/19. I think 24 or even (theoretical) 32 bit resolution is complete overkill and serves no practical purpose whatsoever in digital audio PLAYBACK systems.
I have reason to believe that not bit resolution but time resolution is most important parameter in digital audio playback (jitter sensitivity is already a hint). So striving for ever higher bit resolution is not going to help. Besides I yet have to see the playback system capable of resolving true 24 bit resolution with 0.0001% THD at the speaker output.
TDA1541A performance, like all other DAC chips, depends heavily on application. One can only make valid comments about DAC chip performance when comparing DAC chips that are used in best possible application, this is seldom the case. In other words, you are usually listening to a fraction of DAC chip potential. Based on listening tests, I would say most applications are only able to extract a fraction of DAC chip potential.
This means that a TDA1543 could have the potential to challenge modern DAC chips (perceived sound quality!), but only when used in a super optimized circuit. I don't dare to think what's possible with a super tuned TDA1541A-based DAC. I am working on this, based on the extraordinary performance (perceived sound quality) possible with a single TDA1543.
I also have to add that I use a different approach by applying large bandwidth signal processing, all the way up to the speakers.
Then have speaker crossocer filter, speaker bandwidth, and human auditory system perform required brickwall filtering.
Although this approach also has drawbacks (non-linear response of the connected audio components). It's easier to achieve much better perceived sound quality this way, instead of going the OS / digital brickwall filter route.
Hi gabdx,
[quoye]Could we could find a way to split the 24 bit 96khz signal between two tda1541 ^.^ [/quote]
It could make sense to use 16 bit 192 KHz in NOS mode with a single TDA1541A DAC chip. This way at least the time resolution could be increased a bit. This however requires original 24/192 or 24/96 KHz material, upsampling from 44.1 or 48 to 96 or 192 KHz offers no advantages.
It's impossible to achieve 24 bit resolution using multiple TDA1541A DAC chips (DAC chip tolerances, summed on-chip logic switching noise from multiple DAC chips). 16 bits means 2^16 or 65536 possible analogue signal levels, with TDA1541A 4mA full-scale output current this translates to 61 nA / bit.
With 24 bits we have 2^24 or 16,777,216 possible analogue signal levels, with the 4mA full scale example we would get 238 Femto-amperes / step. It's not difficult to imagine that we cannot possibly achieve this with multiple TDA1541A chips and required wiring, plus noise contribution of each DAC chip and connected I2S interface.
When being realistic, no audio set on this planet is capable of resolving 24 bit resolution up the signal path due to practical limitations. 24 bit resolution is already lost when the DAC output signal leaves the DAC PCB.
[quoye]Could we could find a way to split the 24 bit 96khz signal between two tda1541 ^.^ [/quote]
It could make sense to use 16 bit 192 KHz in NOS mode with a single TDA1541A DAC chip. This way at least the time resolution could be increased a bit. This however requires original 24/192 or 24/96 KHz material, upsampling from 44.1 or 48 to 96 or 192 KHz offers no advantages.
It's impossible to achieve 24 bit resolution using multiple TDA1541A DAC chips (DAC chip tolerances, summed on-chip logic switching noise from multiple DAC chips). 16 bits means 2^16 or 65536 possible analogue signal levels, with TDA1541A 4mA full-scale output current this translates to 61 nA / bit.
With 24 bits we have 2^24 or 16,777,216 possible analogue signal levels, with the 4mA full scale example we would get 238 Femto-amperes / step. It's not difficult to imagine that we cannot possibly achieve this with multiple TDA1541A chips and required wiring, plus noise contribution of each DAC chip and connected I2S interface.
When being realistic, no audio set on this planet is capable of resolving 24 bit resolution up the signal path due to practical limitations. 24 bit resolution is already lost when the DAC output signal leaves the DAC PCB.
Its funny how so many chips can have that same date code the one you show..... over and over and over ...... I will have to check my parts bin at home....
And the other funny thing... How do so many 22 year old chips look like they have never be installed.....
jk
And the other funny thing... How do so many 22 year old chips look like they have never be installed.....
jk
Hi ECdesigns,
Thanks for putting things in perspective, unfortunately I never heard a 24 bit /96 khz record played on proper equipment, I only have a less than decent system ( for the moment) .
Our ear transmits 340000 different signals to the brain through nervous fibers, that’s quite a resolution. ≈19 bits
I have something else to put in perspective too, why tda1541 is not used anymore or abandoned by the industry, is there some sound explanation real life performance inferiority? Are we crazy or what, Give me that best tda1541 dac and ill do a serious listening session with the modern dac. I mention that because for the modern recording/mastering industry and 99 % of audio companies, TDA is considered like junk, ask them. I am very suspicious of everything in general, but buying a benchmark is extremely tempting!!!
Links :
No Over-Sampling DACs and tubes
wonder if I am going to buy this one day :
http://www.curcioaudio.com/cddacsch.gif
Here is a link to some of new best dac tested, I can<t hear any difference on my computer speakers, but those with 24/196 sound cards can help us, (warning dull music).
MYTEK DIGITAL USA
Thanks for putting things in perspective, unfortunately I never heard a 24 bit /96 khz record played on proper equipment, I only have a less than decent system ( for the moment
) . Our ear transmits 340000 different signals to the brain through nervous fibers, that’s quite a resolution. ≈19 bits
I have something else to put in perspective too, why tda1541 is not used anymore or abandoned by the industry, is there some sound explanation real life performance inferiority? Are we crazy or what, Give me that best tda1541 dac and ill do a serious listening session with the modern dac. I mention that because for the modern recording/mastering industry and 99 % of audio companies, TDA is considered like junk, ask them. I am very suspicious of everything in general, but buying a benchmark is extremely tempting!!!
Links :
No Over-Sampling DACs and tubes
wonder if I am going to buy this one day :
http://www.curcioaudio.com/cddacsch.gif
Here is a link to some of new best dac tested, I can<t hear any difference on my computer speakers, but those with 24/196 sound cards can help us, (warning dull music).
MYTEK DIGITAL USA
as John has explained, it is so difficult to implement a jitter free-noise free environment for the humble TDA1543, imagine how difficult it will be for the "HiResolution" DACs...they always sound sterile in comparison.
They sound sterile because they don't have the distortion, in the first place...
The THD of a 1543 and the THD of speakers are very different in quality.
While the THD of a speaker goes up with listening level as does the THD of classA amps and both sound natural because they behave like musical instruments and like the human ear...
... the THD / level - relationship of a bad multibit DAC chip like the TDA1543 is inverted, @ 0 dB it is fairly ok but @ low level it gets very bad.
That sounds like the crossover distortion of a cheap classB IC amp: Unnatural.
As an example, if a piano tone fades out on a 1543 DAC, it will start to have increased harmonics and the more silent the tone gets, the less pure it will sound.
That is caused by the inherent poor low level linearity of the 1543 as a result of the bad matching of the MSB in chip production.
It was designed and sold as an economy DAC and it has nothing to do with jitter or power supply quality or whatever hocuspocus you do in the DAC elsewhere.
You will never ever change it, although the quality aspects of your DAC may be good on jitter etc., in terms of THD quality the 1543 will be crap forever.
The 1541 is better but has always different quality & sound between left and right channel because every DAC chip and every channel has a unique distortion spectrum, the only cure are chips with perfectly low distortion but it is extremely hard to find from TDA's. Also the low level distortion rises strongly without os.
... in the first place. Not only because of the missing distortion, also because of what you wrote.
But that changes nothing for the 1543...
This is the easiest way to hear difference between a good and a bad DAC if everything else equals good in the applications environment. (f.e. good filtered power supply and decoupling.)
It's most often difficult to find out if you listen to most of todays heavily compressed CD music which doesn't at all use all the 16 bits. Producers also seems to be scared to death for silent parts in modern music today.
I have a Telarc, All Digital, recording from 1983 with Bach played by Michel Murray on an hefty pipe organ at First Congregational Church, Los Angeles.
With my more and more optimised Chinese DAC (TDA1541A) i now can hear what, before this DAC came in to my life, sounded like some very weak crackling hiss low down in the silent parts. Now I suddenly can hear what it is, dissolved in sound.
It's the airpump feeding the pipe organ!
reason for "something wrong" with the sound,
could be, from personal ex.
the fashion of usual speaker unit design...
crossover points, "indirect radiation" of low fr,
tonal balance moved to highs,
the beginning of audio event is unclear
and over emphasized, the space and micro detail is moved to the
upper while lack of informations in crucial region from 200 to 500Hz?
>
I can spot the difference only with
12-to15 in paper membrane 35-45gm of moving mass and alnico motor
central driver from roughly 100Hz to 4500-5000Hz bw.
and subs + high fr units added?
.
I can say that character of the dac chips topologies and all in all other components could be more revealing?
there is no point to comment the sound signature of dacs
with standard bas-reflex speakers, regardles of money input?
could be, from personal ex.
the fashion of usual speaker unit design...
crossover points, "indirect radiation" of low fr,
tonal balance moved to highs,
the beginning of audio event is unclear
and over emphasized, the space and micro detail is moved to the
upper while lack of informations in crucial region from 200 to 500Hz?
>
I can spot the difference only with
12-to15 in paper membrane 35-45gm of moving mass and alnico motor
central driver from roughly 100Hz to 4500-5000Hz bw.
and subs + high fr units added?
.
I can say that character of the dac chips topologies and all in all other components could be more revealing?
there is no point to comment the sound signature of dacs
with standard bas-reflex speakers, regardles of money input?
Hi Bernhard,
Link:
Welkom bij Audio Club Velsen
Translate / read "verslag van de ACV bijeenkomst Woensdag 19 november 2008"
quote from this review (translated)
"De grote klinkt wat strakker in het laag en preciezer in het algemeen, maar persoonlijk vind ik de kleine DAC wat natuurlijker klinken in deze set. Deze heeft mijn voorkeur, ondanks dat deze veel goedkoper is. Dit komt doordat de klankbalans wat rustiger wordt, en dat heeft nu eenmaal mijn voorkeur, de betere kwaliteiten van de grotere DAC ten spijt."
"The big one (DI4T) has a bit tighter bass, and more accurate in general, but personally I think the small DAC (DI4MJ) sounds more natural with this set. This one has my preference, despite the fact it's much cheaper. This is because the tonal balance is more relaxed, and that is what I prefer, despite the better qualities of the bigger DAC."
The reviewer of "Audio Club Velsen" preferred the DI4MJ (4 x TDA1543) without even knowing what DAC chips were used inside the unit.
The sound quality of the single TDA1543 in the ISD player made a quantum leap compared to the early DI4MJ and DI4T designs that were driven by an iMac (AE module with Toslink output).
And this:
http://www.diyaudio.com/forums/digital-source/140538-lossless-sd-card-player-41.html
post #406
More reviews will follow later.
Everything I do is done for good reason, double and triple checked, verified by extensive listening tests. My aim is improving performance instead of degrading it. I started this project in 2006, the ISD player was completed last month. The unit sounds like almost 4 years of extensive research was put into its design.
After the first few seconds of listening to this beauty, it becomes clear what was wrong with the theoretically perfect digital audio playback equipment, it failed to provide the impression of listening to live music.
If someone told me that a Shaolin monk would be able to throw a single sewing needle through a glass plate I would say: impossible, this cannot be, theoretically the needle has too little mass and / or speed to penetrate the glass plate. The needle would probably bend. I could come up with all kinds of theories, tests and measurements why this cannot be done.
Then you see this:
YouTube - Shaolin Monk Throwing Sewing Needle Through a Piece of Glass 4
If someone told you a single TDA1543 can create the impression of listening to live sound quality, you would say: impossible, this cannot be, theoretically THD is too high, dynamic linearity errors are too high, it's a economy chip, not designed for these performance levels. You could come up with all kinds of theories and measurements why this cannot be done.
Then you hear this:
- attached picture -
Barriers are only in the mind, and it's all just a matter of technique.
Link:
Welkom bij Audio Club Velsen
Translate / read "verslag van de ACV bijeenkomst Woensdag 19 november 2008"
quote from this review (translated)
"De grote klinkt wat strakker in het laag en preciezer in het algemeen, maar persoonlijk vind ik de kleine DAC wat natuurlijker klinken in deze set. Deze heeft mijn voorkeur, ondanks dat deze veel goedkoper is. Dit komt doordat de klankbalans wat rustiger wordt, en dat heeft nu eenmaal mijn voorkeur, de betere kwaliteiten van de grotere DAC ten spijt."
"The big one (DI4T) has a bit tighter bass, and more accurate in general, but personally I think the small DAC (DI4MJ) sounds more natural with this set. This one has my preference, despite the fact it's much cheaper. This is because the tonal balance is more relaxed, and that is what I prefer, despite the better qualities of the bigger DAC."
The reviewer of "Audio Club Velsen" preferred the DI4MJ (4 x TDA1543) without even knowing what DAC chips were used inside the unit.
The sound quality of the single TDA1543 in the ISD player made a quantum leap compared to the early DI4MJ and DI4T designs that were driven by an iMac (AE module with Toslink output).
And this:
http://www.diyaudio.com/forums/digital-source/140538-lossless-sd-card-player-41.html
post #406
More reviews will follow later.
Everything I do is done for good reason, double and triple checked, verified by extensive listening tests. My aim is improving performance instead of degrading it. I started this project in 2006, the ISD player was completed last month. The unit sounds like almost 4 years of extensive research was put into its design.
After the first few seconds of listening to this beauty, it becomes clear what was wrong with the theoretically perfect digital audio playback equipment, it failed to provide the impression of listening to live music.
If someone told me that a Shaolin monk would be able to throw a single sewing needle through a glass plate I would say: impossible, this cannot be, theoretically the needle has too little mass and / or speed to penetrate the glass plate. The needle would probably bend. I could come up with all kinds of theories, tests and measurements why this cannot be done.
Then you see this:
YouTube - Shaolin Monk Throwing Sewing Needle Through a Piece of Glass 4
If someone told you a single TDA1543 can create the impression of listening to live sound quality, you would say: impossible, this cannot be, theoretically THD is too high, dynamic linearity errors are too high, it's a economy chip, not designed for these performance levels. You could come up with all kinds of theories and measurements why this cannot be done.
Then you hear this:
- attached picture -
Barriers are only in the mind, and it's all just a matter of technique.
Attachments
I am loving to death my Linear Interpolation DAC 4x 1541 S1.... with reclocked CD pro via I2S.....
In my opinion .... The source is finally living up to my amp ( model 500 ) and speakers ( Ohm Fs ). Fs have been reported to be able to reproduce square waves.... for the last year or so since refurbing/moding the amp and pre-amp, I felt the system was limited by the source... NOT ANY MORE.....
I guess once I break down that metal barrier an SD player will be in my future, but seems like a relocked CD-Pro is just as good from what some people have reported.
Seeing I only started building DACs about 8 months... I don't feel too bad being behind you guys by about a year's worth of technology....maybe I"ll catch at some point
johnk
In my opinion .... The source is finally living up to my amp ( model 500 ) and speakers ( Ohm Fs ). Fs have been reported to be able to reproduce square waves.... for the last year or so since refurbing/moding the amp and pre-amp, I felt the system was limited by the source... NOT ANY MORE.....
I guess once I break down that metal barrier an SD player will be in my future, but seems like a relocked CD-Pro is just as good from what some people have reported.
Seeing I only started building DACs about 8 months... I don't feel too bad being behind you guys by about a year's worth of technology....maybe I"ll catch at some point
johnk
Last edited: