"I can hear them church bells pre-ringing!" - A (mostly) time domain investigation into DAC-like reconstruction filters

[Mort]

But the typical opinion here is to reject this.

We don't reject it. It's just inaudible. There's a difference.

 

[Csaba Kelemen]

We don't reject it. It's just inaudible. There's a difference.

Revelation is not a scientific argument.

 

[Mort]

Revelation is not a scientific argument.

This isn't science. It's applied science. Engineering. Have you familiarized yourself with the difference?

 

[Csaba Kelemen]

We don't reject it. It's just inaudible. There's a difference.

What you call inaudible is probably sub-audible.

 

[RandomEar]

The mega tap gurus (Rob Watt -Chord, dCS, HQPlayer, Mike Moffat) argue that supercomputing, or rather huge convolutions, are required for reasonably accurate signal reconstruction. But the typical opinion here is to reject this. This forum also claims scientific foundations in its name. In my opinion, your study did not clarify this issue, it only maintains beliefs. With an argument based on maximum deviation, one could reasonably formulate a valuable criticism. That is why I brought it up.

If they claim it, they're free to prove it - the burden is on them. Also, again: This thread/study is not about the topic you are discussing (which seems to originate in claims by Rob Watt and others). My argument in the OP mainly revolves around the frequency limits of human hearing, which is well established science. So are distortion thresholds or barely audible changes and masking. So you are not pointing out mistakes made here, you are simply opening a new discussion about a totally different topic, which is commonly seen as derailing.

If you would like to see other measures evaluated or think that current measures are not enough, you need to

• Prove that there are deficiences in current metrics like distortion
• Prove that your proposed metrics cure those deficiencies
• Establish the audibility thresholds for your proposed metrics

 

[RandomEar]

What you call inaudible is probably sub-audible.

That's not something that exists. Things are either audible or not. Things cannot exits and not exist at the same time.

 

[Sokel]

The joys of strict science

(Dr. Ignaz Semmelweis's ghost still laughs at the asylum, and not with a crazy laugh )

 

[voodooless]

Things cannot exits and not exist at the same time.

Quantum theory: hold my beer... or not

 

[RexrothPigeon]

Quantum theory: hold my beer... or not

And a thousand audio snake oil salesman wink knowingly

 

[antcollinet]

If you would like to see other measures evaluated or think that current measures are not enough, you need to

• Prove that there are deficiences in current metrics like distortion
• Prove that your proposed metrics cure those deficiencies
• Establish the audibility thresholds for your proposed metrics

So much this.

That's not something that exists. Things are either audible or not. Things cannot exits and not exist at the same time.

And this.

 

[Pausi]

Strange that we get so many filter selections in consumer grade DACs when professional studio grade like Benchmark Media Systems only use one: Linear Phase Brick Wall.
See technical explanation from John Siau here.

 

[Mnyb]

Strange that we get so many filter selections in consumer grade DACs when professional studio grade like Benchmark Media Systems only use one: Linear Phase Brick Wall.
See technical explanation from John Siau here.

That's the correct way its not an user choice it's an engineering decision IMO hats of to benchmark . I constantly forget which is the correct one so i have to google when i get a new product with this built in filter quiz

Theory specify what to do get as close as possible be done, the rest is marketing FUD

 

[KSTR]

Strange that we get so many filter selections in consumer grade DACs when professional studio grade like Benchmark Media Systems only use one: Linear Phase Brick Wall.

Some pro DACs and ADCs like RME offer all the filters that the chips support, for good reasons, especially in measurement applications. Sometimes a sharp linear-phase filter is best for the task at hand, but not always. One typical reason is latency. The min-phase slow filters have the lowest intrinsic latency.

 

[antcollinet]

Some pro DACs and ADCs like RME offer all the filters that the chips support, for good reasons, especially in measurement applications. Sometimes a sharp linear-phase filter is best for the task at hand, but not always. One typical reason is latency. The min-phase slow filters have the lowest intrinsic latency.

But that then becomes a trade off. Reconstruction accuracy (Albeit mostly inaudible) vs latency.

 

[Mnyb]

But in the pro world latency can be a problem for the work flow so i can see the need .

But not in a consumer product ?

 

[MaxwellsEq]

But that then becomes a trade off. Reconstruction accuracy (Albeit mostly inaudible) vs latency.

Yes, in some live broadcast tasks, especially where humans cope badly with delays, minimising latency is important. I assume that's why RME offer the choice.

 

[KSTR]

But that then becomes a trade off. Reconstruction accuracy (Albeit mostly inaudible) vs latency.

There are more reasons than latency.
When you use a DAC and ADC in a signal generator + oscilloscope type of usage (or many other time-domain use cases) you want a causal behavior. Linear-phase filters behave quasi-noncausal.

 

[KSTR]

Yes, in some live broadcast tasks, especially where humans cope badly with delays, minimising latency is important. I assume that's why RME offer the choice.

I'd rather think their mindset is like "we shall not make any assumption about how our customers will be using the device", and when important features are available there is no need to hide them from the user.

 

[modus2000]

NOS is not a good approximation of our ground truth signal at all.

Because your ideas restricted only by electronic measurement of DAC and you don't care of the summary result of complete music system.
When you start evaluate behaviour of the whole system (but not only the DAC), you will understand that System already have mechanical filter which do approximation - speaker, which is mechanically restrict range 20-20Khz and which approximate square wave by it mass of membrane.
The Mass of speaker do all necessary approximation for amplified signal from NOS DAC

So, NOS DAC is not needed any filter (from the listening point of view), because all square shapes will be rounded by mechanical filter of speakers or headphones.
When you apply filter to DAC you add second filter to the system

But be fair - you never hear ultrasonic pre-ringing of delta-sigma DAC, because it will be also filtered by speaker mass

 

[MaxwellsEq]

So, NOS DAC is not needed any filter (from the listening point of view), because all square shapes will be rounded by mechanical filter of speakers or headphones

Now read up about Intermodulation Distortion in the audio band caused by the cr*p let through due to failing to use a proper reconstruction filter.