Focus on SINAD Gives Vendors Hiding Space

[3125b]

@JohnYang1997
the critique is not directed at you or your excellent work I‘m sure.

 

[JohnYang1997]

@JohnYang1997
the critique is not directed at you or your excellent work I‘m sure.

I understand that this specific issue of the measurement of L30 triggered these threads. As, it's being mentioned. I think I would response. Also as the issue showed I investigated this issue with more units. So this is not a bad thing overall.

 

[boXem]

I understand that this specific issue of the measurement of L30 triggered these threads. As, it's being mentioned. I think I would response. Also as the issue showed I investigated this issue with more units. So this is not a bad thing overall.

Indeed I have the same questions in mind since the A90 and M400 measurements.
You clearly showed your skills. I believe nobody will argue against this.
But I am wondering if you and some of your colleagues are not going a too far away for the sake of SINAD.
In example, both A90 and L30 have an very low input impedance (2k and 2.5k if I remember well). The one and only advantage I can see is gaining a few nV in noise. For the user it's just potential issues.
But you are right, people buy SINAD. They seem crazy about your products (A90, L30), they consider the ABH2 as largely superior to any other amplifier because it's SINAD is unequaled (only due to noise) while it's equaled or beaten on all other metrics. That the way it is, I am not the one who will blame you for designing products that sell.

 

[maverickronin]

You asked for examples, and I like some aspects of my Nuprime Dac-10H, but its USB i/f performance is bad: ground loops, adverse reaction to drop outs (whereas for example my Arcam dac behaves much better), and a driver that has more drop outs than just using the vanilla driver from Windows. I've had bad pot mismatch on a Gilmore HD amp ("highly engineered"), capacitor failures on a $1.5k dac integrated, on a sub and on a Hypex power supply due to cheapo caps; driver problems with an ESI dac as well ("pro" gear). Most people just sell their stuff and move on but there are many user reports of similar issues and usability issues have been widespread in gear I've bought.

I do think you're right about the focus on SINAD is making people temporarily forget other aspects of a product. Several DAC threads start out full of hope and promise for a new budget champ and then degenerate into wailing and gnashing of teeth as weird issues pop up.

All that information is very useful, but it's not really anything Amir can test for, at least if we want any decent rate of reviews. Amir does mention those kinds of things when they happen to him, but he's just one sample and probably only uses most products for an hour or two.

Honestly, this is where more normal "subjective" reviewers come in. Someone who thinks you need to hear a new DAC in your system for a couple weeks to review it my not end up with anything useful to say about how it sounds but they can certainly identify usability problems or flaky drivers. Pretty much anyone can do that. Not everyone has an TOTL AP unit.

 

[maverickronin]

What are the benefits of using an analog volume pot rather than using a digital one?

Generally the benefit to analog volume control is safety since it can't be affected by misbehaving software and protects your downstream gear from up upstream faults.

 

[A Surfer]

Each resistor element on each substrate layer was individually laser trimmed to match. When you pull them apart, there are tiny laser cuts on each step in the material. Must have been an early application of laser trimming?

Unfortunately, people squirt Deoxit into those wonderful pots and well, you know what happens after a bath in Deoxit...

Actually I don't know and would appreciate some information. I haven't used Deoxit on my gear, but my brother used to swear by it and seemed to use it to good efect. Thanks in advance.

 

[odyo]

Indeed I have the same questions in mind since the A90 and M400 measurements.
You clearly showed your skills. I believe nobody will argue against this.
But I am wondering if you and some of your colleagues are not going a too far away for the sake of SINAD.
In example, both A90 and L30 have an very low input impedance (2k and 2.5k if I remember well). The one and only advantage I can see is gaining a few nV in noise. For the user it's just potential issues.
But you are right, people buy SINAD. They seem crazy about your products (A90, L30), they consider the ABH2 as largely superior to any other amplifier because it's SINAD is unequaled (only due to noise) while it's equaled or beaten on all other metrics. That the way it is, I am not the one who will blame you for designing products that sell.

What low input impedance does ? My house has shtty electricity i'm worried about ground loop or something though i don't have much knowledge about it. Generally my speakers does lots of hissing noises, ocassional popping noise when powering on my pc or shutting off. My psu blow up or gone bad couple of times before. My surge protector power strip with 6 outlet making electrical noise. My monitor's stupid back ambient light also making electrical noise if i use it. My pc is also very busy with high end components and coil whines. Though my SBX AE5 soundcard is completely silent at every volume step even with super sensitive iems and headphones. No issue with it but i guess it gets it power from pcie socket via motherboard<PSU. My psu is quite good. I wonder are these external dacs and amps gonna be clean via usb and power strip ?

 

[KPH]

Generally the benefit to analog volume control is safety since it can't be affected by misbehaving software and protects your downstream gear from up upstream faults.

Thanks. But what do you mean by "misbehaving software" ?

 

[boXem]

What low input impedance does ? My house has shtty electricity i'm worried about ground loop or something though i don't have much knowledge about it. Generally my speakers does lots of hissing noises, ocassional popping noise when powering on my pc or shutting off. My psu blow up or gone bad couple of times before. My surge protector power strip with 6 outlet making electrical noise. My monitor's stupid back ambient light also making electrical noise if i use it. My pc is also very busy with high end components and coil whines. Though my SBX AE5 soundcard is completely silent at every volume step even with super sensitive iems and headphones. No issue with it but i guess it gets it power from pcie socket via motherboard<PSU. My psu is quite good. I wonder are these external dacs and amps gonna be clean via usb and power strip ?

The first things that come in mind are:
- increased sensitivity to unbalancing of the connection. In the case of XLR, both input resistors and output resistors will need to be very well matched to get proper CMMR, in the case of RCA, well, it's not balanced by definition. So ground loops can become an issue. This is worth reading since it's author explains the issue much better than I do.
- matching with other gear. When the guys designing the M400 think that high output impedance allows better SINAD and the guys designing the A90 think low input impedance allows better SINAD, you have a nice voltage divider.
- matching with other gear again. It's kind of a convention in the industry that output impedance is low, input impedance high. When designing an output stage, one can consider that the other side impedance will be quite high. Output stage performance (distortion) *could* degrade when facing a lower input impedance than expected.
- others?

 

[maverickronin]

Thanks. But what do you mean by "misbehaving software" ?

All kinds of things can go wrong, especially if you're playing music from a general purpose computer running other applications and not something dedicated to music.

Some other piece of software could raise the system volume level by itself to blast you at full volume or a program with its own volume control could crash and start outputting noise at full volume.

 

[BDWoody]

Thanks. But what do you mean by "misbehaving software" ?

Anything that puts you at risk of have a line level signal unintentionally sent to your amps with no attenuation.
This can also happen via user error with one misplaced mouse click...ask me how I know...

It's certainly not necessary...but I think of it like a loaded gun. You have to be paying attention, and with software issues, a loud pop or two at the wrong time can send the cats flying...at the least.

 

[boXem]

Anything that puts you at risk of have a line level signal unintentionally sent to your amps with no attenuation.
This can also happen via user error with one misplaced mouse click...ask me how I know...

It's certainly not necessary...but I think of it like a loaded gun. You have to be paying attention, and with software issues, a loud pop or two at the wrong time can send the cats flying...at the least.

I'm more curious about cats flying than user errors. How do you know?

 

[odyo]

Anything that puts you at risk of have a line level signal unintentionally sent to your amps with no attenuation.
This can also happen via user error with one misplaced mouse click...ask me how I know...

It's certainly not necessary...but I think of it like a loaded gun. You have to be paying attention, and with software issues, a loud pop or two at the wrong time can send the cats flying...at the least.

Yeah i remember when i first enable the exclusive mode of Tidal. My soundcard uses windows volume slider instead of volume knob. When enabling exclusive mode Tidal max volume become windows max volume.

 

[KPH]

All kinds of things can go wrong, especially if you're playing music from a general purpose computer running other applications and not something dedicated to music.

Some other piece of software could raise the system volume level by itself to blast you at full volume or a program with its own volume control could crash and start outputting noise at full volume.

Anything that puts you at risk of have a line level signal unintentionally sent to your amps with no attenuation.
This can also happen via user error with one misplaced mouse click...ask me how I know...

It's certainly not necessary...but I think of it like a loaded gun. You have to be paying attention, and with software issues, a loud pop or two at the wrong time can send the cats flying...at the least.

Ok I see. Another thing that came up to me was that, I did some searching myself and found out digital volume control has loss of bits, so thought analog control is preferred. But with many of the recent products with high SINAD measurements like say over 100 db, is it still something to worry about?

 

[BDWoody]

I'm more curious about cats flying than user errors. How do you know?

I had my 708p's on my kitchen table (maybe 3' away) playing around with the AES inputs through my motu 896 and the cuemix software...which I was not very familiar with, when I changed from some setting to another while playing music and got every bit of what those speakers had to offer (500w/speaker) from much closer than I would have liked.

Both cats were hovering nearby, as they normally do when I'm clearly messing around with stuff...and when quite suddenly Steely Dan was telling the well known story of the messed up Cousin Dupree at an SPL that left me covering both ears figuring out how to make it stop, both cats literally launched themselves and what they were perched on in opposite directions faster than I think I've ever seen anything move.
The final shotgun level blast when I turned off the motu was a special farewell reminder to be a bit more careful...

 

[maverickronin]

Another thing that came up to me was that, I did some searching myself and found out digital volume control has loss of bits

That is true, but analog volume controls aren't perfect either. More or less anytime you run a signal though a resistive element you get extra noise added to the signal, which will probably be amplified by a subsequent stage so it's often a wash.

But with many of the recent products with high SINAD measurements like say over 100 db, is it still something to worry about?

Sending 24 bit data to a modern high-SINAD DAC (even if you're just padding 16 bit Redbook or something) makes it a non issue in most cases.

 

[DDF]

Name me a 4-gang/2-gang pot with 2k-5kohm, matched to the bottom(-60db) for every unit and yet small enough to put in the chasis. Then we will use it.
The l30 uses same pot as a90 both for wolf and amir. It's simply a occasional case.

True but sampling could throw out the lowest 5 or 10 percent clunkers. Nothing worse than listening and turning a pot and hearing sound collapse to one ear. No company is better or worse in this area than any other unless they use incoming sampling or paying extra for pots that are pre-screened.

Good idea using variable gain, I did the same in my DIY preamps. Its also possible to add a fixed resistance to one channel of the cable between pre and amp that creates minute channel imbalance at higher volumes but creates a decent compensating imbalance at low volumes. But its ugly and you have to make sure frequency response isn't impacted and that the noise of the amp isn't degraded (some amps hate high source impedance, if the input referred current noise is high). I created a model of it in excel (frequency response vs pot rotation including cable capacitance), I had one pre-amp with a pot so annoying.

If pot variation is an issue for someone, best approach is to get a DAC+Amp combo where the attenuation is digital.

Its a shame caring about maximizing SNR and then toss so much of it away because the pot has so much imbalance that it can't be used at low levels. A person would have to choose one or the other, defeating the purpose of pursuing SNR excellence.

 

[DDF]

Channel imbalance is tested, as Amir said but is only valid for THAT device.

Yup, of course. But if you get a stinker, it also means that stinkers exist and no pre-screening of pots is occurring, so that becomes a useful data point.

It is much easier to design a well performing (volpot) device when it is allowed to cost a lot more and the device can be a LOT bigger.

I'm was fascinated by the Muses device so I bought a dac+pre+HD amp that specifically used it. Perfect channel balance, no coarse settings, selectable with a remote, never gets dirty or intermittent, transparent. Its great. Given a choice, I'll never go back to a pot again. Measures like a champ:
https://www.innerfidelity.com/content/katzs-corner-episode-8-my-current-affair

Don't know why its not used more, (maybe cost?)

Trouble shooting groundloops and making cheap solutions for it generally isn't easy.
So yes, one can make a test for it but may not be conclusive because real world situations may differ a lot from home situations.

Yup, but with a proper jig, one can much better detect gear that doesn't work well with others. Alas, nothing's perfect.

 

[DDF]

I do think you're right about the focus on SINAD is making people temporarily forget other aspects of a product. Several DAC threads start out full of hope and promise for a new budget champ and then degenerate into wailing and gnashing of teeth as weird issues pop up.

All that information is very useful, but it's not really anything Amir can test for, at least if we want any decent rate of reviews. Amir does mention those kinds of things when they happen to him, but he's just one sample and probably only uses most products for an hour or two.

Honestly, this is where more normal "subjective" reviewers come in. Someone who thinks you need to hear a new DAC in your system for a couple weeks to review it my not end up with anything useful to say about how it sounds but they can certainly identify usability problems or flaky drivers. Pretty much anyone can do that. Not everyone has an TOTL AP unit.

Well, its not easy but this is the stuff that matters. Focusing on the easy stuff (120 vs 110 dB SINAD) doesn't move anything forward except screen out the bums. Lets look at these 1 by 1:

• channel balance vs volume

As mentioned, best you can do here is pop the hood and warn users (or offer praise) based on previous experience, knowing the pot

• ground isolation (risk of ground loops)

Make a jig, but no one's going to step up and I don't blame Amir for not doing it.

• for DACs, driver quality and robustness (custom driver? what are benefits? Standard xmos driver?....)

Agreed, can't test for it. Better than "published" subjective reviews (which are 1 offs), this is where the community should chime in. I'd like to see more users complain when the unit they buy doesn't just work.

• how it deals with dropped samples (overshoot/undershoot and how audible). I've had dacs sound very different in how they deal with this.

Easly testable, just need to create appropriate audio digital samples. Back when I was in Ethernet design, we'd create "killer packets" for example.

• reliability cuts corners (what brand caps and what are temp ratings?)

Pop the hood, raise concern if seeing sketchy junk.

 

[maverickronin]

for DACs, driver quality and robustness (custom driver? what are benefits? Standard xmos driver?....)

Since MS finally got on board with UAC2 do drivers even matter anymore unless you need ASIO for live recording or something?