Audiofools turned objectivists or vice-versa?

[gsp1971]

Anyway, to get back to the topic of the thread, there are a few examples of subjectivists who turned objectivists, but I wonder are there people who went the opposite way?

 

[SIY]

I think I want to test whether short or long listening times give better identification performance...

By whom, and using which test format?

 

[ahofer]

Anyway, to get back to the topic of the thread, there are a few examples of subjectivists who turned objectivists, but I wonder are there people who went the opposite way?

One of the Stereophile guys said he did for a while, but he was miserable so he shifted back. I don’t remember if it was Atkinson, who is here on ASR from time to time. Or maybe Holt.

 

[DimitryZ]

By whom, and using which test format?

I think I will be the subject and AB format. 10 trials?

I think on IOS, nonplaying DAC has to be physically off. I will need to fashion an opaque shield that will hide the DAC from view.

I will also need to show my wife how to turn the DACs on and off and how to use the switch box.

 

[SIY]

I think I will be the subject and AB format. 10 trials?

I think on IOS, nonplaying DAC has to be physically off. I will need to fashion an opaque shield that will hide the DAC from view.

I will also need to show my wife how to turn the DACs on and off and how to use the switch box.

A lot of potential non-auditory cueing. I’d look at a sorting format. Sufficient trials to get you to better than 5%.

Before getting fancy, do a level matched comparison. You may be surprised.

 

[DimitryZ]

A lot of potential non-auditory cueing. I’d look at a sorting format. Sufficient trials to get you to better than 5%.

Before getting fancy, do a level matched comparison. You may be surprised.

What is a sorting format?

And it turns out that my Liberty and M2TECH output 5v from their balanced analog outs.

 

[SIY]

What is a sorting format?

And it turns out that my Liberty and M2TECH output 5v from their balanced analog outs.

That is not sufficient to determine level matching. You have to measure the outputs with identical signals and match to 0.1dB.

 

[DimitryZ]

That is not sufficient to determine level matching. You have to measure the outputs with identical signals and match to 0.1dB.

I only have a digital voltmeter. That's probably not enough - I will need to clearly distinguish 5 vs. 5.06 volts.

I do have one of those calibrated microphones that plug into your phone and work with audio apps. But I don't think it's dependable to 0.1 dB.

It sounds like I need to get a measuring interface for PC and some software?

 

[SIY]

I only have a digital voltmeter. That's probably not enough.

I do have one of those calibrated microphones that plug into your phone and work with audio apps. But I don't think it's dependable to 0.1 dB.

Voltmeter will likely work fine.

 

[DimitryZ]

Voltmeter will likely work fine.

Will give it a try! 1 KHz tone is the common test point?

I do have a test CD with that stuff.

 

[Wes]

pink noise may be better

 

[DimitryZ]

pink noise may be better

Centered on 1KHz?

 

[rdenney]

Pink noise isn't centered--it covers the entire audio spectrum. But it is random noise, so the reading may wiggle a bit.

But 1KHz will also work fine, and it will not likely explore the bandwidth of the AC-reading part of your DVM.

It's not important that the reading is really accurate, but rather that it's repeatable so that matching the devices under test will really match them.

If you use a DVM make sure it is a true RMS-reading meter. Most better ones are if they claim to be. Most can read power properly, but the meter needs the bandwidth to read 1000 Hz, not 60 Hz.

Rick "who uses an old Trio SSVM for this task" Denney

 

[DimitryZ]

Pink noise isn't centered--it covers the entire audio spectrum. But it is random noise, so the reading may wiggle a bit.

But 1KHz will also work fine, and it will not likely explore the bandwidth of the AC-reading part of your DVM.

It's not important that the reading is really accurate, but rather that it's repeatable so that matching the devices under test will really match them.

If you use a DVM make sure it is a true RMS-reading meter. Most better ones are if they claim to be. Most can read power properly, but the meter needs the bandwidth to read 1000 Hz, not 60 Hz.

Rick "who uses an old Trio SSVM for this task" Denney

I need to locate my test CD, I think it has pink noise and 1/3 octave white noise, if memory serves.

In the old days I used it to manually measure my Magnepans with RS meter, which was really bad and one had to use corrections table...

 

[audio2design]

I don't think this is true at all. Since at least the 1970s, any mass-market receiver, run below clipping, has been sonically indistinguishable from the most expensive amp/pre-amp combinations that were properly designed for flat frequency response and low distortion (which includes almost all solid-state products and many tube products). And DACs in CD players have been sonically perfect since the very first Philips player (which I owned).

That is not totally true. Many of those older receiver / amps had rising often significant distortion at high frequencies. You will note it is rare to find specs from that age with distortion measured anywhere but at low frequencies.

 

[Blaspheme]

One of the Stereophile guys said he did for a while, but he was miserable so he shifted back. I don’t remember if it was Atkinson, who is here on ASR from time to time. Or maybe Holt.

It was JA, he includes that story in a column Who Watches the Watchers? The overall subject matter is relevant here.

 

[SIY]

A sine wave will be fine given that DACs are all pretty flat except in the top octave. 200-1000 Hz will work great.

 

[MarkS]

Many of those older receiver / amps had rising often significant distortion at high frequencies. You will note it is rare to find specs from that age with distortion measured anywhere but at low frequencies.

How old is "older"?

From the February 1975 issue of Stereo Review, page 32:

“Sherwood’s next-to-lowest-price stereo receiver, the Model S-7110 … can deliver 17 watts per channel into 8-ohm loads at less than 0.9 per cent harmonic distortion between 40 and 20,000 Hz.”

https://worldradiohistory.com/Archi...iFI-Stereo/70s/HiFi-Stereo-Review-1975-02.pdf

 

[Mart68]

'Into 8 ohms load' - now give it a speaker with impedance down to 2 ohms and see if it sounds identical to a Krell. It won't.

 

[gsp1971]

It was JA, he includes that story in a column Who Watches the Watchers? The overall subject matter is relevant here.

I find it quite strange that a technical editor in charge of measurements does not advocate a pro-measurement approach to component evaluation.
I am sure JA knows better than that, he has been in the business for way too long and he has massive technical knowledge, but he needs to balance component reviews with advertising revenues, hence the more cautious approach.