Master Thread: Are measurements Everything or Nothing?

[Robin L]

I wish I had the wherewithal to be able to answer @frullo, without picking at typoz and name calling. Too many mean ol' coots and much snobbery in the pecking order; more commonly known as "El Grosser Schwantz" syndrome. IMHO and YMMV.

"Beethoven didn't born deaf. Your argument is worthy Brandolini's law."

That's what was written by @frullo. It is incoherent on multiple levels. Can't be certain what point @frullo is attempting to make. So it's not simply being snarky, it's pointing to an inability to articulate one's thoughts. "Your argument is worthy Brandolini's law." Not "of" or " a worthy example of".

Brandolini's Law: also known as the bullshit asymmetry principle, is an internet adage that emphasizes the difficulty of debunking false, facetious, or otherwise misleading information:

Ironically or no, this internet interaction is a fine example of Brandolini's Law.

 

[audio2design]

It is a delta-sigma DAC, producing plenty of ultrasonic noise... I believe modulator running at master clock frequency, thus up to (IIRC) 100 MHz. Modulator output noise peaks at the modulator rate.

Ultrasonic noise is not RF, and that there is a clock at 100Mhz, does not mean there are those frequencies on the output, not that it would matter, it would not pass through circuitry (analog) that is going to struggle with a MHz. I can't remember how many bits the internal DAC is, but it is running at far < 100MHz.

 

[voodooless]

I give kudos to @frullo in attempts to make a few valid points

Can you elaborate as to which ones those are? Possibly you have better luck deciphering Google Translate gibberish? It's really not the right tool to use for a complex conversation.

Apologies for typographical, grammatical, contextual, and/or syntactical [?] errors!

Indeed, please don't touch the fonts, especially the colour. Black text with dark themes on the forum does not work together.

 

[Suffolkhifinut]

"Beethoven didn't born deaf. Your argument is worthy Brandolini's law."

That's what was written by @frullo. It is incoherent on multiple levels. Can't be certain what point @frullo is attempting to make. So it's not simply being snarky, it's pointing to an inability to articulate one's thoughts. "Your argument is worthy Brandolini's law." Not "of" or " a worthy example of".

Brandolini's Law: also known as the bullshit asymmetry principle, is an internet adage that emphasizes the difficulty of debunking false, facetious, or otherwise misleading information:

Ironically or no, this internet interaction is a fine example of Brandolini's Law.

What are you trying to articulate?

 

[audio2design]

Allow me to ask those question(s) differently:

• Can an APx555 show the measurement (not listening) difference(s) between two different violins, played by robotic arms which can repeatedly pluck the same strings [??] the same exact way, multiple times?
• Not different types of measurements except those that are reduced-down to the standard set that we observe here at ASR?
• The kicker>> ...as well as a trained ear (yeah, you @SIV) instantly can????

Apologies for typographical, grammatical, contextual, and/or syntactical [?] errors!

I am trying to understand how this is at all relevant .... but if you want to go there, the output of an APx555 would show subtleties between two apparently identical violins that no human would ever pick up by ear. It would need a human to interpret the results, but we are good at reading graphs ...

 

[Miska]

Send me the units and I’ll be glad to show you how to do competent and transparent measurements of them using appropriate equipment.

Why not buy those with your own money, like I do?

 

[xaviescacs]

Care to offer your measurement results instead of beating your mouth? I see lot of people making a lot of noise here, but very little measurement data of their own accompanying the noise. Those keyboard warriors...

As I see it you are wasting time answering those comments instead of explaning how we should interpret your plots. Do you claim that there are audible differences or not? To me, posting a lot of plots and not explaining what they show is making a lot of noise. This is about audible differences or any measurable difference?

 

[Miska]

Ultrasonic noise is not RF, and that there is a clock at 100Mhz, does not mean there are those frequencies on the output,

There's plenty of high frequency stuff at the DAC chip output, due to high speed switching.

not that it would matter, it would not pass through circuitry (analog) that is going to struggle with a MHz.

Here you are touching precisely the problem I'm suspecting. That the analog section is struggling with the switching noise in I/V stage vs the mixing of those channels that precisely have each different high frequency switching noise (on purpose). So the difference signal between the channels heavily increases as function of frequency.

I can't remember how many bits the internal DAC is, but it is running at far < 100MHz.

You can configure ESS DAC chips to use up to 6 bits modulator output. Using 64 unity weighted elements, thus output has maximum 65 output levels.

 

[Pdxwayne]

Everything must be questioned, no Prophets just an engineer measuring and testing. Not putting Amir’s honesty in question would trust him before most. Dissent isn’t always a threat @frullo as an example, trolling or an honest opinion?

I definitely trust Amir. Since joining ASR, I bought 5 DAC, 4 HPA, and 2 headphones mainly because of him.

 

[Miska]

To me, posting a lot of plots and not explaining what they show is making a lot of noise.

I think they are self-explanatory for this forum's audience?

I'm not going to discuss audibility things more on this forum.

 

[Blumlein 88]

I get a similar result to Miska on a D10b as I posted already. But only on 44.1 and only with Rew creating the signals. With other playback software or other sample rates it doesn't happen. Maybe Miska could use a different software to play his 44.1 kHz signals. Or try them at 48 kHz with what he is using.

 

[Pdxwayne]

I think they are self-explanatory for this forum's audience?

I'm not going to discuss audibility things more on this forum.

I am not EE engineer. Would like you to explain, please?

 

[Jimbob54]

"Dear" @frullo (troll or otherwise) has a few worthy points made in original post that is worthy of a dialogue.
But most replies (save a few) are character assassinations at a minimum.
I give kudos to @frullo in attempts to make a few valid points << regardless of the pompous replies that followed. This is an outright gang-bang by a pack of wolves.

I wish I had the wherewithal to be able to answer @frullo, without picking at typoz and name calling. Too many mean ol' coots and much snobbery in the pecking order; more commonly known as "El Grosser Schwantz" syndrome. IMHO and YMMV.

Allow me to ask those question(s) differently:

• Can an APx555 show the measurement (not listening) difference(s) between two different violins, played by robotic arms which can repeatedly pluck the same strings [??] the same exact way, multiple times?
• Not different types of measurements except those that are reduced-down to the standard set that we observe here at ASR?
• The kicker>> ...as well as a trained ear (yeah, you @SIV) instantly can????

Apologies for typographical, grammatical, contextual, and/or syntactical [?] errors!

I'm mentally scarred by the idea of being gang banged by wolves.

 

[audio2design]

I'm mentally scarred by the idea of being gang banged by wolves.

I hear it's better than crocodiles.

 

[Miska]

I get a similar result to Miska on a D10b as I posted already. But only on 44.1 and only with Rew creating the signals. With other playback software or other sample rates it doesn't happen. Maybe Miska could use a different software to play his 44.1 kHz signals. Or try them at 48 kHz with what he is using.

I'm not using REW. But I've used same way for measuring about 50 or so DACs. So I'm slightly reluctant to make changes with the testing setup as for me the important thing is to maintain kind of database of the results.

I recently measured also the SMSL M500 mk II (ES9038PRO) and SMSL M400 (AK4399) on the same setup.

And some of the Archimago's results share similarities.

 

[Jimbob54]

I hear it's better than crocodiles.

Depends if you're in water or not.

 

[voodooless]

I get a similar result to Miska on a D10b as I posted already. But only on 44.1 and only with Rew creating the signals. With other playback software or other sample rates it doesn't happen.

Then have REQ make a PCM file of the test signal, and FFT that directly. That should show if the problem is with generation.

 

[rdenney]

Me too. But, especially with the advent of DSP, I realized it was just received dogma.

It was DSP that did it for me, too. Back in the day in my home system I had a JVC SEA 10-band stereo graphic equalizer, which had a RTA display, a pink-noise generator, and the accessory "calibrated" microphone. To the extent that room corrections could be made using octave filters with a 6-dB slope, it worked well enough. But there were associated phase shifts and other outcomes resulting from the hardware filtration that it used, especially when adjacent bands were pulled in opposite directions.

I also had an SAE parametric equalizer that allowed me to adjust the center frequency and Q for one filter in the bass region. It also used hardware filters, and, of course, one band in the bass region isn't really enough.

Both of these added noise and probably phase distortions, though I never measured the latter.

In commercial systems, the guys I worked with customarily included a White Instruments 1/3rd-octave equalizer, adjusted for room response using a real-time analyzer.

When that JVC unit died, I pulled it out of the system and gave that idea up for many years, until I got the SAE. But it was also too hard to live with--every power glitch would erase the settings and it also added noise. I retired that unit, too, even though it looked way cool.

Getting the Yamaha YDP2006 PEQ changed all that. It's not computer-driven DSP like you were thinking, but it is the same thing for an analog system. The input goes through a gain stage (adjustable), a pre-emphasis filter (switchable), and ADC (based on a two cascaded 20-bit Burr-Brown PCM-1760's), to the DSP unit that works in 20 bits. Coming out of the DSP unit, it goes through a DAC based on a Burr-Brown PCM63P. Balanced ins and outs. This provides a SINAD in the mid-to-high 90's, plus digital filters that don't impose all the side effects of hardware filters. This was intended for commercial applications and it was expensive when new; they just about give them away now.

Rick "whose home stereo has been EQ'd using at least real-time analysis since the middle 80's" Denney

 

[Suffolkhifinut]

I definitely trust Amir. Since joining ASR, I bought 5 DAC, 4 HPA, and 2 headphones mainly because of him.

What are you doing with it all?

 

[rdenney]

• Can an APx555 show the measurement (not listening) difference(s) between two different violins, played by robotic arms which can repeatedly pluck the same strings [??] the same exact way, multiple times?

This is a strawman. The answer is yes (assuming truly identical setups, and assuming there is an actual difference that can be reliably heard by a microphone). But who cares?

Where a device like the AP is useful isn't in measuring the characteristics of musical instruments--whether that can be done will be more about microphones. A device like the AP compares the signal going in to the signal coming out, and tells us how (and how much) that signal was altered in the process. It's perfect for measuring the playback system, which is its role.

We are still confusing music-making with reproduction playback.

Rick "musical instrument measurement is more like what a Klippel does with speakers, not what an AP does with electronics" Denney