Master Thread: Are measurements Everything or Nothing?

[Robbo99999]

It's a mysterious wall of words, full of sound and fury, with no supporting facts. So, looks like we're on the internet, all right

I've just got a bit of a problem with people saying that DACS sound different when they measure fine here on ASR. Filters could make a difference if you're choosing one that rolls off way too early, but beyond that I think it's all imagined. This thread is the Wild West though, and is intended to be as such, so I can happily leave some people to waste their time discussing the nuances of different DACS.....I'll just give a brief interjection here & there!

 

[Aleksandar RS]

Does the 321 have various filters to choose from? It is possible that filters that droop a bit in treble instead sound like the bass has changed a little or vice versa.

Do you happen to know what Wadia has inside them in modern products. I had a couple of their Dacs way back when, the last being a Wadia 25. I wonder if they are using off the shelf DAC chips in them now. Back in the old days they were close to 3 db down at 20 khz because of the Spline filtering they used.

EDIT to add: Okay they are using the older ESS 8 channel Sabre Dac chip 9016S combining channels for lower noise. I think the SMSL uses the newer flagship version of the ESS9039 Dac chip.

No, the wadia 321 does not have selectable filters, they have their own sound, and that's it. Of course, the 9039PRO is newer and has better specifications. In principle, this shows the most important thing, which is that the chip is important, but the most important thing is the support for it and the implementation itself. You can't put a Ferrari engine in a Chery (Chinese car brand) and say I sell fantastic cars, here are the specs.

 

[Blumlein 88]

No, the wadia 321 does not have selectable filters, they have their own sound, and that's it. Of course, the 9039PRO is newer and has better specifications. In principle, this shows the most important thing, which is that the chip is important, but the most important thing is the support for it and the implementation itself. You can't put a Ferrari engine in a Chery (Chinese car brand) and say I sell fantastic cars, here are the specs.

The most important thing is the fidelity of the analog output signal. The Chery example is a bit of false equivalence.

 

[Blumlein 88]

@bok What even is phase noise, would i see it in the distortion measurement or frequency response, if it where really, really high?

Edit: I'm not a pro, so be gentle mate.

Example of high phase noise. This is a 11,025 hz tone at 44.1 khz sample rates for a cheap AV converter. Ideally you would have a low fairly even noise floor. Instead you have a rather high noise floor spreading around the central tone. Some jitter tones are right at 70 dbFS. The whole noise floor is elevated. This was with a 128K FFT.

 

[Aleksandar RS]

The most important thing is the fidelity of the analog output signal. The Chery example is a bit of false equivalence.

Chery was the first that came to my mind, it's a bit of an exaggerated comparison.

 

[bok]

@bok What even is phase noise, would i see it in the distortion measurement or frequency response, if it where really, really high?

The introduction in this paper covers it quite well:

https://tf.nist.gov/general/tn1337/Tn190.pdf

Normally the term phase noise includes fluctuations caused by both phase modulation and amplitude modulation. Jitter is sometimes also called phase noise but that excludes AM phase noise which is caused by e.g. noise in DAC Vref supplies. Close-in phase noise refers to the phase noise close to the carrier or fundamental frequency (e.g. seen as noise skirts in FFT). High close-in phase noise is often claimed to cause temporal smearing. I'm not aware of any studies that have defined an audibility threshold to this. Anyhow many high-end DACs claim to have oscillators with low close-in phase noise.

Though not strictly related to phase noise the book "An Introduction to the Psychology of Hearing" by Brian Moore refers to a study that showed close-in AM to have lower threshold of audibility (i.e more audible) than FM.

 

[TheBatsEar]

This thread is the Wild West though

This thread is a containment thread. They are needed and do a vital function for the community, soaking up a certain kind of emotionally invested participants.

It's supposed to be the Wild West to be effective.

 

[HarmonicTHD]

Why do you keep putting words in my mouth? I have not said it is audible. Though that does not mean it is inaudible.

Are we putting words in your mouth?

Master Thread: Are measurements Everything or Nothing?

My only opposition about the ones I listen and know well are the proportions of voices and instruments,they sound enormous if you get closer than 4 meters. But his space is big and that's not a problem,they are an absolute fun and the peaks can through you from your chair :facepalm:. Like any...

www.audiosciencereview.com

 

[bok]

Are we putting words in your mouth?

Master Thread: Are measurements Everything or Nothing?

My only opposition about the ones I listen and know well are the proportions of voices and instruments,they sound enormous if you get closer than 4 meters. But his space is big and that's not a problem,they are an absolute fun and the peaks can through you from your chair :facepalm:. Like any...

www.audiosciencereview.com

You seem to have reading comprehension disorder. Try reading again what I said.

 

[HarmonicTHD]

You seem to have reading comprehension disorder. Try reading again what I said.

You are running out of arguments now getting personal. Nice move.

 

[bok]

Vice versa.

 

[Aleksandar RS]

I've just got a bit of a problem with people saying that DACS sound different when they measure fine here on ASR. Filters could make a difference if you're choosing one that rolls off way too early, but beyond that I think it's all imagined. This thread is the Wild West though, and is intended to be as such, so I can happily leave some people to waste their time discussing the nuances of different DACS.....I'll just give a brief interjection here & there!

If you have a little problem with people who say that DACS sound different when they measure fine here on ASR, I don't see why you initially replicated my post (I noticed it just now). I can't solve those problems. I criticized the sound of the D400ES and wrote about it long before the measurements came out. Primarily about the filter OFF option (which is practically unusable, as it turned out later in the measurements). In addition to cutting a large part of the audible spectrum, it often forces the amplifier into clipping. I wrote about it both here on ASR and SMSL directly. I received a firmware update from them, which, as expected, did not solve the problem).

This was all before the publication of the measurement results, which only confirmed that the option was not usable. Even Amir wrote that way (for the filter OFF option), I quote: "Not that you would want to use it anyway as it cuts way into the audible frequency response". Also, I understand you mean that all DACs sound the same. If that's the case, then there's no point in discussing it because I don't think so and it's not worth discussing, so I can't help there either.

 

[eyekaiser]

Not all DAC's are made equal. Between two pieces of shit, sure, maybe you can tell. Going back to my point, you work with what you have. Going back to my earlier point, what is it all for? These DAC's are okay for the price and task. Absolutely the same? Probably not. But you are talking about the difference of two things that do not really matter. One piece of shit probably smells a little different from the other. But both are shit. Enjoy.

 

[dshreter]

Not all DAC's are made equal. Between two pieces of shit, sure, maybe you can tell. Going back to my point, you work with what you have. Going back to my earlier point, what is it all for? These DAC's are okay for the price and task. Absolutely the same? Probably not. But you are talking about the difference of two things that do not really matter. One piece of shit probably smells a little different from the other. But both are shit. Enjoy.

If a piece of shit has a 110 SINAD, it will still sound exactly the same as every other high SINAD DAC even if it has flies landing on it.

No, it’s not intuitive that a $200 interface would perform so admirably, but that’s just how it is.

 

[HarmonicTHD]

Not all DAC's are made equal. Between two pieces of shit, sure, maybe you can tell. Going back to my point, you work with what you have. Going back to my earlier point, what is it all for? These DAC's are okay for the price and task. Absolutely the same? Probably not. But you are talking about the difference of two things that do not really matter. One piece of shit probably smells a little different from the other. But both are shit. Enjoy.

Again just anecdotes and opinions. No evidence.

There are plenty of other forums where people like to tell each other stories. Why did you come here? What did you associate the word science in the forum title?

Believe what fairytale you want. I am wasting my time as you guys chose to ignore the facts.

 

[Blumlein 88]

Not all DAC's are made equal. Between two pieces of shit, sure, maybe you can tell. Going back to my point, you work with what you have. Going back to my earlier point, what is it all for? These DAC's are okay for the price and task. Absolutely the same? Probably not. But you are talking about the difference of two things that do not really matter. One piece of shit probably smells a little different from the other. But both are shit. Enjoy.

So what would you give as an example of a not POS DAC (from studio gear or wherever) which is good or is a reference that gives superior results to good measuring low priced POS devices?

 

[bok]

On a lighter note let's see what ChatGPT has to say about these issues.

Q: How does close-in phase noise impact audio dac sound?

ChatGPT:
Close-in phase noise can impact the sound of an audio DAC (digital-to-analog converter) by introducing distortion or noise in the audio signal. This noise can manifest as jitter in the signal, which can cause audible artifacts such as harshness, smearing, or a lack of detail in the audio. The closer the phase noise is to the signal frequency, the more significant the impact on the sound quality will be. To minimize the impact of close-in phase noise on the audio quality, it is important to use high-quality clock sources and carefully design the digital signal processing stages in the DAC.


Q: Besides frequency response and output impedance what other factors influence audio dac sound at low frequencies?

ChatGPT:
There are several factors that can influence the sound of an audio DAC at low frequencies, in addition to frequency response and output impedance. Some of these factors include:

• Linearity: A DAC's linearity, or its ability to accurately reproduce different levels of input signals, can impact the sound of low-frequency signals. Non-linearities in the DAC can cause distortion, which can degrade the sound quality of low-frequency signals.
• Dynamic range: The dynamic range of a DAC, or the difference between the loudest and softest signals it can reproduce, can also impact the sound of low-frequency signals. A DAC with a high dynamic range will be able to reproduce low-frequency signals with more detail and clarity.
• Noise floor: The noise floor of a DAC, or the level of background noise it generates, can also impact the sound of low-frequency signals. A DAC with a low noise floor will be able to reproduce low-frequency signals with less background noise and greater clarity.
• Jitter: Jitter, or small variations in the timing of the digital signals, can also impact the sound of low-frequency signals. Jitter can cause distortion in the audio signal and can reduce the clarity of low-frequency signals.
• Power supply noise: The noise generated by the DAC's power supply can also impact the sound of low-frequency signals. A clean power supply that has been properly filtered and regulated will help to minimize this impact.

Overall, to get the best sound of low-frequency signals, a high-quality audio DAC should have a combination of good frequency response, output impedance, linearity, dynamic range, low noise floor, low jitter, and clean power supply.


And finally the million-dollar question.

Q: Can audio dac sound be characterized by measurements only?

ChatGPT:
Measurements can be used to characterize the performance of an audio DAC, but they do not provide a complete picture of the DAC's sound quality. Measurements can provide objective information about the DAC's frequency response, distortion, noise, and other parameters, but they do not necessarily correlate with subjective perceptions of sound quality.

For example, a DAC with very low distortion measurements may still sound harsh or bright to some listeners, while a DAC with higher distortion measurements may sound more pleasing to others. Similarly, a DAC with a very flat frequency response may not sound as natural or engaging as one with a slightly rolled-off response.

It's important to note that while measurements can provide important information about a DAC's performance, they do not replace the need for listening evaluations by trained listeners. A combination of measurements and subjective listening evaluations is often used to determine the overall sound quality of an audio DAC.

Additionally, it's also worth noting that different listeners, environments, speakers or headphones can also change the perception of sound, which makes it even more complex to determine the overall sound quality of a DAC based solely on measurements.

 

[solderdude]

Measurements can provide objective information about the DAC's frequency response, distortion, noise, and other parameters, but they do not necessarily correlate with subjective perceptions of sound quality.

The problem here is 'subjective perceptions'. This is where the discrepancy occurs. The human perception. Almost always sighted.
This warrants the need for blind level matched listening tests with statistical relevant number of attempts.

 

[eyekaiser]

Well, there is what's on paper and what you actually hear. Take your favorite CD for example. Now copy that CD to a CD-R. Now copy that copy. On paper, all discs sound exactly the same. Right? But if you listen, you will find the copy doesn't sound that great. But that's impossible! It's just a copy of ones and zeroes, right? Other things are at play here. You have two ****** DACs (and keep in mind I have and use one of them) which sound great for voice over or spraying your juices on chaturbate to then be compressed to your viewers and fan base (are we getting good signal-to-noise ratio in capturing those moans and when the fluids hit the lens?) but if we are capturing a symphony orchestra or an opera singer or some rare archival-level ethno recording...in short, when you are really recording, you may want to reach for something a little nicer. These are two hundred dollar components, are they shielded? Did the manufacturer cheap out anywhere? Was everything perfectly assembled? Is the firmware okay? Because if I am really recording something important, I don't want to worry about this. "Oh but the graph says..." yeah, whatever. Which captures the squelching of your juices better, the audient or motu? Does it matter? Depends whose juices I guess..Lots of things look great on paper, but in reality are not. There's 'good enough', which is fine. I don't mind a little cheap and dirty sometimes, but if I am really recording for keeps...

 

[Robbo99999]

This thread is a containment thread. They are needed and do a vital function for the community, soaking up a certain kind of emotionally invested participants.

It's supposed to be the Wild West to be effective.

Indeed!