Serious Question: How can DAC's have a SOUND SIGNATURE if they measure as transparent? Are that many confused?

[Reynaldo]

This didnt measure at all well by JA at Stereophile.

I don't see this measurement issue anymore.

When Gustard X16 came out, several people I knew bought it.
I think it was the second highest rated DAC here on ASR.
There were approximately ten people. I think only one is still with him, all the others including me have sold.
After two hours no one could bear to hear.

Today I have here the IFI AUDIO ZEN ONE SIGNATURE DAC, a Musical Fidelity M1SDAC and the Oppo 205. Both the IFI AUDIO ZEN ONE SIGNATURE DAC and the Musical Fidelity I can listen to for hours and probably if they are measured they will not be well evaluated.

If you look for it, you will hardly find anyone talking bad about the IFI AUDIO ZEN ONE SIGNATURE DAC.

But this is just my opinion without any scientific evidence.

 

[Reynaldo]

I recently bought iFi Audio iPower X DC Power Supply, got better sound quality using iFi DAC.

 

[Gringoaudio1]

Hmmm. I don't really have commercial interests.

I split from iFi in '19 and have no audio related products selling, other than the work I did for iFi.

Much of my current work in electronics has shifted to industrial automation, autonomous drones etc.

I might of course have audio products again in future. But right now nothing at all.

Thor

Cool to see you here sir. You gave me some guidance on an inverted gainclone amplifier in the early 2000s on DIYAUDIO.

 

[NikJi]

That's the point of measurement.

Take any typical audiophile descriptive word and imagine the wide ranging interpretation of that word by other audiophiles, not only similar but, dissimilar, exact opposite interpretations of what you even admit to struggling to define.

One should know what they are measuring for, and then measure for it.

So do yourself and everyone a favor. Get the 2 dacs, a headphone and listen to the same music on both. If you hear a difference, then you know what you are attempting to measure and will prepare accordingly.

 

[Thorsten Loesch]

This didnt measure at all well by JA at Stereophile.

Yes, because the unit JA measured was severely faulty, due to software issues.

I pushed S&M (sales & marketing, not 50 Shades of Grey) to get SP to publish a comment to that effect and have a fixed unit submitted. But even the software fix was given low priority. AFAIK a definitive software version wit complete fixes was released late in 2021. As I have not tested it, I do not know if it REALLY fixes everything.

The iDSD Pro unit has a number of switches which I would call measure/listen switches. With JA some of the underlying circuit configuration switches inside the unit (all softwre driven) were not responding correctly and were "stuck" in "listen".

Unfortunately, after I had already signed off the "golden sample" a new programmer (who is actually a good programmer) came onto the project, after it was in effect completely out of my hands. This Programmer started from an earlier version where all the various switches worked incorrectly and not from the version I had signed off.

The "golden sample" was used in software development and the software on it altered and later the same now actually not aurum but excrementvm sample was used to set factory testing standards. Because someone in AMR/iFi had decided to play big boss capitalist and reorganised a lot of things to be more "professional' the company started to turn out more and more products with major issues at launch.

It meant my control over the product and the production process was severely eroded and placed in the hand of incompetent chinese "engineers" (I use the term engineer in the loosest of all sense, these people would not pass examinations for the lowest level of service tech in Europe, despite having chinese university EE degrees).

I actually referred to the original software the iDSD Pro shipped with as the "Kendathu" release, because of the numbers of bugs, after the famous scene from "Starship Troopers".

I even raised it at board level and gave my own version "We can ill afford another Klendathu" speech. Didn't help.

I think Zen DAC Signature V2 measures better.

Not if the software is correct.

BTW, the measurements here have the same software bug:

https://www.audiosciencereview.com/...ts-of-ifi-idsd-pro-dac-with-comparisons.6119/

There are currently some tests being done on a unit with a more recent software version:

https://www.facebook.com/groups/826563705068569/

Did you design Pro iDSD ?

Yes, among most other ifi & AMR products that needed actual design.

The WiFi/BT antenna affecting the analogue electronics? Or the DSP chip?

Nope, just bad programming that turned the feedback loop switching permanently to "tube+"

The iDSD Pro literally has the same schematic for the Headphone Driver/Line Driver section as the iCAN Pro and measures essentially the same, you can see that here:

https://www.audiosciencereview.com/...fi-pro-ican-headphone-amplifier-review.12247/

The circuit is really very simple. The 4 DAC's output is "mixed" together (giving a 6dB improvement over using a single DAC Chip for HD and Noise and THD & N) followed by a passive LC filter, (optional) volume control and then the line driver.

The line driver has a circuit that allows either an overall feedback loop, or one where the input device is outside the feedback loop (that is meant to be only active in the "tube +" setting). This switching was stuck permanently on. So naturally there is a lot of HD, as you would expect from a tube or jfet stage without loop feedback.

Note I still use and love my micro iDSD Black Label, for portable use.

I designed this one as well. And honestly, IF the software configures it correctly, the iDSD Pro is a LOT better, especially subjectively with the "listen" switches set appropriately and if you "like" the kind of sound that comes from gear I design. The Pro does all that and then some more.

That said, I would not spend my own money on a "Pro" set but get the Zen's and upgrade the external power supplies to get the most from them. I still think the Pro Sounds better, but not in any proportion to the extra expense. It is an indulgence I cannot justify.

BUT, if I was already considering to spend 800 USD on an iDSD Neo, which as USB DAC is technically 100% identical to the Zen DAC (II - I if not using MQA decoder) it may be worth upping the ante and getting a second hand pro as this seems to have finally a firmware that works without major bugs.

Honestly, if the Pro was selling at 999 USD, which was the target pricing at the time of it being designed, or even 1,299 USD accounting for cost overruns I'd consider it worth spending my own money, as it has a few really unique options that produce a sound I personally prefer to the Zen DAC. But at more than double that, I do not feel the iDSD Pro is competitive as product. But get a bargain on the 2nd hand market, I'd take it at 1,200 USD over a new iDSD Neo any day.

Thor

 

[Music1969]

Yes, because the unit JA measured was severely faulty, due to software issues.

I pushed S&M (sales & marketing, not 50 Shades of Grey) to get SP to publish a comment to that effect and have a fixed unit submitted. But even the software fix was given low priority. AFAIK a definitive software version wit complete fixes was released late in 2021. As I have not tested it, I do not know if it REALLY fixes everything.

The iDSD Pro unit has a number of switches which I would call measure/listen switches. With JA some of the underlying circuit configuration switches inside the unit (all softwre driven) were not responding correctly and were "stuck" in "listen".

Unfortunately, after I had already signed off the "golden sample" a new programmer (who is actually a good programmer) came onto the project, after it was in effect completely out of my hands. This Programmer started from an earlier version where all the various switches worked incorrectly and not from the version I had signed off.

The "golden sample" was used in software development and the software on it altered and later the same now actually not aurum but excrementvm sample was used to set factory testing standards. Because someone in AMR/iFi had decided to play big boss capitalist and reorganised a lot of things to be more "professional' the company started to turn out more and more products with major issues at launch.

It meant my control over the product and the production process was severely eroded and placed in the hand of incompetent chinese "engineers" (I use the term engineer in the loosest of all sense, these people would not pass examinations for the lowest level of service tech in Europe, despite having chinese university EE degrees).

I actually referred to the original software the iDSD Pro shipped with as the "Kendathu" release, because of the numbers of bugs, after the famous scene from "Starship Troopers".

I even raised it at board level and gave my own version "We can ill afford another Klendathu" speech. Didn't help.

Not if the software is correct.

BTW, the measurements here have the same software bug:

https://www.audiosciencereview.com/...ts-of-ifi-idsd-pro-dac-with-comparisons.6119/

There are currently some tests being done on a unit with a more recent software version:

https://www.facebook.com/groups/826563705068569/



Yes, among most other ifi & AMR products that needed actual design.



Nope, just bad programming that turned the feedback loop switching permanently to "tube+"

The iDSD Pro literally has the same schematic for the Headphone Driver/Line Driver section as the iCAN Pro and measures essentially the same, you can see that here:

https://www.audiosciencereview.com/...fi-pro-ican-headphone-amplifier-review.12247/

The circuit is really very simple. The 4 DAC's output is "mixed" together (giving a 6dB improvement over using a single DAC Chip for HD and Noise and THD & N) followed by a passive LC filter, (optional) volume control and then the line driver.

The line driver has a circuit that allows either an overall feedback loop, or one where the input device is outside the feedback loop (that is meant to be only active in the "tube +" setting). This switching was stuck permanently on. So naturally there is a lot of HD, as you would expect from a tube or jfet stage without loop feedback.



I designed this one as well. And honestly, IF the software configures it correctly, the iDSD Pro is a LOT better, especially subjectively with the "listen" switches set appropriately and if you "like" the kind of sound that comes from gear I design. The Pro does all that and then some more.

That said, I would not spend my own money on a "Pro" set but get the Zen's and upgrade the external power supplies to get the most from them. I still think the Pro Sounds better, but not in any proportion to the extra expense. It is an indulgence I cannot justify.

BUT, if I was already considering to spend 800 USD on an iDSD Neo, which as USB DAC is technically 100% identical to the Zen DAC (II - I if not using MQA decoder) it may be worth upping the ante and getting a second hand pro as this seems to have finally a firmware that works without major bugs.

Honestly, if the Pro was selling at 999 USD, which was the target pricing at the time of it being designed, or even 1,299 USD accounting for cost overruns I'd consider it worth spending my own money, as it has a few really unique options that produce a sound I personally prefer to the Zen DAC. But at more than double that, I do not feel the iDSD Pro is competitive as product. But get a bargain on the 2nd hand market, I'd take it at 1,200 USD over a new iDSD Neo any day.

Thor

Thanks for all the history !

Why does the headphone output have > 4 ohms output impedance ?

I imagine Neo iDSD has a better headphone output ? Power, distortion and lowest output impedance?

 

[posvibes]

If you hear a difference,

But there can't be an "if" the differences exist according to your own experience, what those differences are is not apparent and you presume that if I had the time and frankly gave a damn those same differences would be apparent to me in the same way they are apparent to you. What if I can't hear those differences at all, or hear a differences completely different to your experience.

How can you be sure the rest of your system is just not up to snuff and creating those differences in the DAC's performance?

Nope, "if" is not an option.

 

[HarmonicTHD]

One should know what they are measuring for, and then measure for it.

So do yourself and everyone a favor. Get the 2 dacs, a headphone and listen to the same music on both. If you hear a difference, then you know what you are attempting to measure and will prepare accordingly.

Unless you do a controlled and blind test eg ABX, this exercise proofs nothing but your biases. (And BTW, but I hope that is clear anyway - proper level matching is mandatory but often not done correctly and leading to falsely perceived differences. You know how to level match?)

 

[Thorsten Loesch]

Why does the headphone output have > 4 ohms output impedance ?

It should not have. The Schematics & BOM I have should have suggest it should have ~2 Ohm balanced headphone output impedance.

It could be that the build-out resistors were changed later without my involvement.

I imagine Neo iDSD has a better headphone output ?

Of course not. The iDSD Pro was designed as 1,000USD DAC and the iDSD Neo as 300 USD DAC.

The iDSD Pro has a completely discrete headphone/line driver circuit based on classic pro-audio circuitry. With maximum gain output is 5.6V SE & 11.2V BAL. The Amplifier remains well below clipping at this level.

In principle the HP Amp could output 8V SE / 16V BAL on the +/-12V output stage power supply, but maximum gain is giemped in Software on insistence by S&M (sales & marketing) to avoid the DAC having "too good" a HP Amp. In principle, with hardware changes (resistor values for gain and output stage voltage setting)even more output is possible.

The iDSD Neo has a headphone amplifier based on the MAX97220 as power driver (but used in a multi-loop topology for lower noise and distortion). This means you get around 3.6V SE and 7.2V BAL maximum at the onset of cliping.

Noise & THD should be dominated by the DAC Chip, so the iDSD Pro should be 6dB better.

Depending on volume setting, at low volume the Headphone amplifier noise may become dominant, in which case the iDSD Pro will have around 10dB lower noise in SE and 16dB in BAL.

Overall the Zen DAC & Zen CAN are a much better solution for headphones compared to the iDSD Neo, at much lower cost. Using the Zen One loses some DAC performance, but may still be better combined with Zen CAN than the iDSD Neo, objectively and subjectively.

Thor

PS, RAA tested the iDSD Neo, look for yourself:

https://reference-audio-analyzer.pro/en/title-report.php?id=3537#gsc.tab=0

 

[solderdude]

Who designed the enclosure/looks for the Zen line ?

 

[Thorsten Loesch]

Who designed the enclosure/looks for the Zen line ?

Studio Cocktail Design Studio, of which I am one of the Principals. More specifically Julien Haziza and myself to a lesser degree. A lot of recent Products have this design team.

Thor

 

[MAB]

Unless you do a controlled and blind test eg ABX, this exercise proofs nothing but your biases.

Exactly!

One should know what they are measuring for, and then measure for it.

So do yourself and everyone a favor. Get the 2 dacs, a headphone and listen to the same music on both. If you hear a difference, then you know what you are attempting to measure and will prepare accordingly.

No!
You will think DAC A sounds one way, and DAC B sounds another way.
You'll either be unable to measure significant electrical performance differences between the DACs, and conclude that measurements don't capture the "acoustic signature" of the DACs. Or you will be able to measure subtle differences between the two DACs, and incorrectly attribute those differences you thought you heard. Or you will mess up some aspect of the experiment, starting with the level matching.

Perhaps you can try something different. Test out your perception. Calibrate what you can actually hear. One common tale is that some electronics are 'euphonic' because they have 'even order' distortion, not the nasty odd order distortion (like tubes). You can get software such as DISTORT and test out these hypothesis. Play a track and dial up the 2nd order harmonics, 3rd order, 4th, ... and any combo of odd and even order harmonics. You can check what your hearing threshold for overall distortion is, and combinations of harmonics. You can see if you can hear the difference between 2nd and 3rd harmonics, any harmonic for that matter. The output is level matched, so you aren't getting fooled there. Once you have calibrated your ear to distortion levels, you can then go and compare distortion numbers in Amir's reviews for DACs, amps, preamps, etc. My guess is you will be surprised at what level of distortion you can hear relative to the differences in electronics.

I say this because there are lots of discussions about audibility, testing with controls, ABX, etc. And those arguments are valid. But this is one way to calibrate your perception and sensitivity to distortion to some of the test measurements published here. And harmonic distortion is just one aspect of reproduction.

 

[Reynaldo]

That said, I would not spend my own money on a "Pro" set but get the Zen's and upgrade the external power supplies to get the most from them.

Reading this whole story, the best thing would be to stick with what I already have and I think it's great.
The IFI AUDIO ZEN ONE SIGNATURE DAC together with the new iFi Audio iPower X DC Power Supply.

 

[Reynaldo]

I really liked the look of the iFi Audio NEO iDSD.
I know people who have the NEO iDSD and they speak very highly of the device and could still use the iPower X DC Power Supply on it.

 

[Thorsten Loesch]

Reading this whole story, the best thing would be to stick with what I already have and I think it's great.
The IFI AUDIO ZEN ONE SIGNATURE DAC together with the new iFi Audio iPower X DC Power Supply.

Good call. Both the subjectivist and objectivist crowd are ultimately trying to get you buy more gear. Be it they claim better sound, better numbers, both or just "newer", companies, dealers, reviewers etc. exist only to keep the industry going by making you feel you need to hand over a bag of shekel and upgrade to the latest, even if what you have is adequate for what you need. Long lived equipment with a "good sound" does not help the business side, only the consumer.

A lot of people in this hobby seem stuck in this hamster wheel. Worst, they do not trust their own ears. I have seen people I know tear apart a good sounding system, sell everything at a loss and build a new system up at considerable expense, which I felt (and they eventually themselves admitted) sounded worse.

Don't fall into the trap to feel inadequate and buy what you do not need.

I really liked the look of the iFi Audio NEO iDSD.
I know people who have the NEO iDSD and they speak very highly of the device and could still use the iPower X DC Power Supply on it.

Reynaldo, it is something that took me a long time to learn, as sound engineer, DJ and Electronics designer, learning how to create "good sound" (and not just in electronics).

I build my first HiFi system in 1978, with help from an Uncle who was an EE.

It was mostly based on salvaged parts (speakers, output transformers, Tubes) from old B&W Tube TV's and Radio's for Amplifier (ECL82 SE) and Speakers (10" Wideband plus many 4" Cone Tweeters based on a Box/Crossover design of a small cinema speaker).

I had a polish ZK120 R2R deck modified for stereo, all electronics rebuild using solid state parts in stereo, record/playback head exchanged with one I had to import from west germany, a stereo turntable with an MM Pickup and an old mono tube radio ("steuergeraet" with external speaker) converted into a stereo tuner with a stereo decoder. I was able to align tape deck and radio in a private repair workshop owned by a friend of another uncle.

A year later I had progressed to building a fully solid state system, with 3-Way speakers using Hungarian "Videoton" made midrange and treble dome drivers and a pair of Czech made Tesla 20cm bass drivers, a 100W per channel Amp with 10-Band EQ and LED analyser and a "digital tuner". I could not get the parts for a full PLL, so it used a mixer to convert the frequency downwards and then used a counter with TTL IC's to have a digital readout of frequency and a simple analogue circuit to store multiple stations. I was able to get a commercially made 3-Head cassette tape deck from Akai via multiple interesting import routes.

I moved on to Pro Audio and working with bands and so on. So I have been doing this audio electronic malarke for a little while.

I hope I learned enough to make sure everything I design sounds "good" even if it is "low budget" piece.

The iDSD Neo does need a decent power supply, most products relying on external power do.

At 300....350 USD I think the Neo would be a great product, especially with the headphone specific crossfeed and bass EQ I usually include in products targeted at headphones.

Even while not really competitive with other 800 USD product, I do think the iDSD Neo happens to sound "good". Maybe to some this is worth the price, if it helps them to get off the hamster wheel and stop buying a different dac at a few 100 Bux every few month and selling them on after a few month with substantial loss..

Thor

 

[Reynaldo]

A lot of people in this hobby seem stuck in this hamster wheel. Worst, they do not trust their own ears. I have seen people I know tear apart a good sounding system, sell everything at a loss and build a new system up at considerable expense, which I felt (and they eventually themselves admitted) sounded worse.

Don't fall into the trap to feel inadequate and buy what you do not need.

You are absolutely right.
What's worse, I've made this mistake a few times.

In this hobby if you look for perfection in sound, it's just a dream.
That's why I keep the old (I bought new) Musical Fidelity M1SDAC. Even not playing DSD, only PCM 192Khz but it has an excellent quality together with my integrated Musical Fidelity M6si.

Today I'm using iFi products, but when I feel like it I take the M1SDAC out of the box and put it to work.

 

[Thorsten Loesch]

You are absolutely right.
What's worse, I've made this mistake a few times.

In this hobby if you look for perfection in sound, it's just a dream.
That's why I keep the old (I bought new) Musical Fidelity M1SDAC. Even not playing DSD, only PCM 192Khz but it has an excellent quality together with my integrated Musical Fidelity M6si.

Today I'm using iFi products, but when I feel like it I take the M1SDAC out of the box and put it to work.

If you don't use the coaxial SPDIF input on the Zen One, you can use as SPDIF out to feed the M1S input and have both available to listen to.

For DSD Files MY PERSONAL SUGGESTION (and solution) has long since been to convert to 176kHz/24 Bit, there is software to either do it permanently (which was I used) or to use on the fly conversion with a variety of software solutions and approaches. I rather like Yuki San's tools:

http://wpup.html.xdomain.jp/

Thor

 

[Reynaldo]

iFi Audio NEO iDSD

 

[Purité Audio]

From my perspective IFI have some way to go before I would consider them even remotely credible.

ifi hip dac review (DAC & Headphone Amp)

This is a review and detailed measurements of the ifi hip dac combination USB DAC and portable headphone amplifier (with battery). It is on kind loan from a member and costs US $169 from Amazon including Prime shipping. I don't know why ifi thinks this is a "hip" dac as the enclosure is a...

www.audiosciencereview.com

ifi Zen DAC and Headphone Amp Review

This is a review and detailed measurements of the ifi Zen balanced DAC and headphone amplifier. I purchased it from Amazon last week. The Zen DAC costs US $130 including free shipping. I must say, I am not a fan of the look of the Zen DAC. It hearkens back to 1960s furniture style which would...

www.audiosciencereview.com

Keith

 

[Music1969]

DSD Files MY PERSONAL SUGGESTION (and solution) has long since been to convert to 176kHz/24 Bit

So this was not your idea/push to support higher sample rate DSD and stick with DSD1793 chip ?

And even the DSD512 and DSD1024 upsampler in the Pro iDSD?