There is nothing wrong with SINAD chasing

[babadono]

surprised that even a well measuring amplifier had noise coming out of the tweeter of his speakers.

Yep...chasing silence at no signal on very efficient horn loaded speakers is a REAL thing....

 

[RexrothPigeon]

I am also impressed with D class performance overall. But would not expect it to last 30 years.

Why’s that?

 

[Oddball]

Why’s that?

Just a hunch from overall consumer product industry. My best guess is that nothing is intended or specd to last that long any more. Just a sign of times.

 

[RexrothPigeon]

Just a hunch from overall consumer product industry. My best guess is that nothing is intended or specd to last that long any more. Just a sign of times.

Audio Hunch Research is down the hall

 

[Oddball]

Oh just got lost as there are many halls. Any specific one?

 

[RexrothPigeon]

Oh just got lost as there are many halls. Any specific one?

It's right past the old man yelling at the clouds about the Good Old Days

 

[amirm]

On a slightly different, but related topic...
I find the idea of having a DSP capable AVR (for example) and then disabling all features and running it in pure direct mode (just because this maximises SINAD) is pretty stupid.

Well, I could say that about your comment. If we want to see how good an AVR's DAC implementation is, we have to compare it to stand-alone DACs. In that sense, all processing must be turned off to make it a proper comparison. It is through this testing that we have discovered how deficient AVR DACs are. Prior to this, people spent extra money for "high end" AVRs and processors thinking that got them same level of performance as stereo product. Now they know different.

Second, no one knows what manipulation you are going to do so no way to create a defensible scenario to test.

Third, common manipulations such as high pass filtering have very small impact on measurements.

Finally, pure mode has not made a difference in any of my AVR DAC testing that I recall. I turn it on anyway as to avoid people wondering if performance would improve. But it just doesn't as the limitation is in the DAC design, not the rest of the AVR. The main impact that mode has is in analog input where Pure direct disables digitization and feeds the audio directly to the amplifier. Otherwise, it is a marketing feature.

 

[Oddball]

It's right past the old man yelling at the clouds about the Good Old Days

I guess you might have gotten lost in the great halls as they are truly endless. It might be that you are just stuck in the limbo until you actually get to advance to the next level. Much like you might associate with a game, in your young level "speak". Level 1, is not bad, ways above level zero.

 

[RexrothPigeon]

in your young level "speak".

Ser, I too am old

 

[Oddball]

Ser, I too am old

Older implies wiser - so not sure what to think of it. Any particular point you want to make or can we make this uncomfortable and odd encounter pass?

 

[Chrispy]

Audio Hunch Research is down the hall

Then again much of the class D stuff being used hasn't gotten 30 years old yet....

 

[Chagall]

Well, I could say that about your comment. If we want to see how good an AVR's DAC implementation is, we have to compare it to stand-alone DACs. In that sense, all processing must be turned off to make it a proper comparison. It is through this testing that we have discovered how deficient AVR DACs are. Prior to this, people spent extra money for "high end" AVRs and processors thinking that got them same level of performance as stereo product. Now they know different.

Second, no one knows what manipulation you are going to do so no way to create a defensible scenario to test.

Third, common manipulations such as high pass filtering have very small impact on measurements.

Finally, pure mode has not made a difference in any of my AVR DAC testing that I recall. I turn it on anyway as to avoid people wondering if performance would improve. But it just doesn't as the limitation is in the DAC design, not the rest of the AVR. The main impact that mode has is in analog input where Pure direct disables digitization and feeds the audio directly to the amplifier. Otherwise, it is a marketing feature.

I think there’s a bit of talking past each other here.

The post you were replying to wasn’t really about measurement methodology. It was about real-world usage: people running AVRs in Pure/Direct mode specifically to chase higher SINAD numbers, while giving up the very DSP features that make an AVR valuable in the first place. In that context, the criticism isn’t about how ASR measures gear, but about how some consumers use it.

More broadly, this thread feels a bit messy because there are at least three different kinds of SINAD chasing being mixed together:

Measurement perspective – necessary and useful for fair, repeatable comparisons.

Manufacturer SINAD chasing – which has clearly pushed the industry forward.

Consumer SINAD chasing – where higher numbers are pursued even when there’s no noise or audible distortion with the actual speakers in use.

As many have said, modern electronics are already transparent, in no small part due to your work Amir. That’s a genuinely positive outcome and something to be proud of.

But from a consumer perspective, once there’s no audible noise or distortion in a given system, further SINAD chasing can easily turn into buying for the sake of a bigger number, especially if it comes at the cost of room correction, bass management, or other DSP features that have a clear, audible benefit.

That’s the distinction I think some posts were trying to make.

 

[Everett T]

Ser, I too am old

Let's not remind us of our age, lol.

 

[TurtlePaul]

Well understood. But your amp could go for another 10 years as well?

I am also impressed with D class performance overall. But would not expect it to last 30 years. So that would kind of weight in into the price/year equation.

If you are looking at it in price/year and think that old school amps last 30 years, keep in mind that recapping has costs. When somebody paid $2,200 in 1999 dollars for a Bryston 4B-ST, you better bet that they are going to get it serviced. When a McIntosh or Accuphase starts to hum, it goes to a workbench when a black box AVR would go to the graveyard.

Servicing these modern class D amps would make no sense beyond replacing a bad power supply. The whole amp costs less than the recap cost of one of the old monsters. These modern class D amps are not serviceable, but $100-300 devices were never going to be economic to do so.

I drive a Mercedes and was told the secret of Mercedes reliability. They have a good reputation for reliability but they truthfully aren't reliable as a Toyota. The reason for the reputation is i) it is more reliable than BMW and ii) Mercedes owners get they timely services and do repairs. When you have a 10-15 year old luxury car you still abide by the service interval and a $5,000 repair doesn't total the car. High end amps are the Mercedes of the audio world - you drive them to the dealership instead of driving them into the ground.

Also, a lot of these cheap Chinese amps these days are using better capacitors and better cooling that Hypex and Topping, respectively.

 

[TurtlePaul]

But from a consumer perspective, once there’s no audible noise or distortion in a given system, further SINAD chasing can easily turn into buying for the sake of a bigger number, especially if it comes at the cost of room correction, bass management, or other DSP features that have a clear, audible benefit.

That’s the distinction I think some posts were trying to make.

I feel like that can be a little bit of a straw man, because in my experience the best DSP devices also have very high SINAD (MiniDSP, Topping JDS, WiiM).

I also wouldn't call 'pure direct' modes 'SINAD chasing' because it existing long before SINADs were measured and was completely driven by a competing set of audiophile superstitions.

 

[-Matt-]

If we want to see how good an AVR's DAC implementation is, we have to compare it to stand-alone DACs. In that sense, all processing must be turned off to make it a proper comparison. It is through this testing that we have discovered how deficient AVR DACs are.

Of course. I completely agee that this is how your testing should be done. And I'm grateful for all the effort that you put into it, which is why I said...

SINAD is an incredibly useful metric for us to determine which components have been well designed. Publishing this information is ASR's greatest contribution.

I'm fully in favour of the SINAD measurement that you do. For me that is not even in question. It provides excellent information for us consumers to make informed purchases.


I thought the topic here was "SINAD chasing", which (to me) is more about what some people may choose to do with that information. (I tried to explain this)...

I would define "SINAD chasing" as continually upgrading (e.g. your DAC) because the latest reviewed model achieved 1dB higher score. If people want to do that and feel good about having the latest and greatest, that is up to them.

My comment about measuring whole system SINAD wasn't intended to be a suggestion of a new testing protocol for you @amirm.

Second, no one knows what manipulation you are going to do so no way to create a defensible scenario to test.

Of course, it would be a terrible way of measuring the performance of components. Everyone's rooms and systems are different, there would be far too many uncontrolled variables.


And perhaps this was going a bit too far...

At the end of the day, it is the fidelity of the sound wave at the listening position that matters, not the fidelity of the electrical signal at some earlier stage in the chain.

Both matter. When I'm purchasing a component, I absolutely do want to know how good or bad its electrical performance is.


And yet (separate from Amir's testing), I still think if there was a relatively straightforward way for us individually to measure our system's overall SINAD (including the effect of the room) that could be quite interesting. There are some details that would need to be defined. I.e. If the target curve is anything but flat, this would presumably need to be taken into account in the SINAD calculation? What level or levels should measurements be run at? etc, etc. Anyway, that is a discussion for another thread.

I also wouldn't call 'pure direct' modes 'SINAD chasing' because it existing long before SINADs were measured and was completely driven by a competing set of audiophile superstitions.

I'm not against pure direct being there as an option (and it is correct that Amir should use it for his measurements). However, whenever I hear its use being recommended for normal listening, it does set off my "audiophile superstition" alarm bells!

SINAD chasers would presumably not allow themselves to make use of DSP room correction and use only pure direct mode. In doing this they may achieve a tiny benefit in SINAD, but by forgoing room correction a massive penalty in the actual sound at the listening position.

 

[Holmz]

Both matter. When I'm purchasing a component, I absolutely do want to know how good or bad its electrical performance is.

SINAD is one number, so it is a bit like comparing the grey scale of.a Van Gough with the grey scale of a cartoon.
The distortion profile and the noise are bundled together.

The main saving grace is that as the number go to zero the noise be one of the things that end up at zero.

Somewhere in the <100 dB are some great amps and some amps that hiss like vipers.

 

[Tell]

If it's nothing wrong with chasing SiNAD, then I take it there's no limit for when you upgrade? A 90dB SINAD amp is not enough when there's a 100dB one? 110dB is even better? 120dB still not enough? 130dB is too low? 140? 150? How long are you going to chase that dragon?

Tbh getting new stuff based purely on better SINAD than the "old" thing you have that's already audible transparent is just as bad as those pure subjectivist buying new stuff because Paul McGowan said so in a video.

 

[Mort]

Tbh getting new stuff based purely on better SINAD than the "old" thing you have that's already audible transparent is just as bad as those pure subjectivist buying new stuff because Paul McGowan said so in a video.

Hard to take such exaggeration seriously.

 

[Tell]

Hard to take such exaggeration seriously.

So why should you chase more SINAD when there's no audible different?