Some people who have reported using their ears include JBL, ESS, Purifi, and AKM. The work at JBL was apparently done by Sean Olive, first a student and later a colleague of Floyd Toole. Doesn't necessarily look unscientific from here.
You're right, unscientific is the wrong word. I should have said "it's qualitative". If they had something quantitative, they'd definitely be using that.
Seems like one might view statistical discrimination listening tests as quantitative. That's how we end up with things like measurements of 'thresholds of audibility.' OTOH preference listening tests may be better described as qualitative.
Possibly of interest regarding measured specifications verses subjective sound quality in loudspeakers: http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html
Perhaps a parallel to the above could be drawn between ADC THD+N numbers and subjective SQ (although no analogy is perfect)?
Possibly of interest regarding measured specifications verses subjective sound quality in loudspeakers: http://seanolive.blogspot.com/2009/01/what-loudspeaker-specifications-are.html
Perhaps a parallel to the above could be drawn between ADC THD+N numbers and subjective SQ (although no analogy is perfect)?
Last edited:
Great link! One of the author's comments below the article particularly hit home:
I guess this is why some people (like our friend above) just use their ears to tweak their equipment. I don't think any of us have an anechoic chamber available to take a more quantitative approach (not to mention the relevant microphones and measurement software).
I'm quite sure that JBL, ESS, Purifi and AKM use mostly measurements for testing. THD+N is an accepted figure of merit used by the whole industry. Subjective listening evaluations can be used as additional tests but not as replacement for measurements.
There does seem to be two schools of thought here. Those who think the device should present as clean a path as possible and those who want to steer the audio in a certain direction.
I think there are two "schools" one doing things and another is talking-only 😉
bohrok2610, ES9822, as any other DAC/ADC/MCU etc is a quite fast switching silicon device that needs accurate nearly zero inductive rails decoupling. Any attempt to implement that with some pink TH Wima will add tonnes of inductance there. This is not a question of personal taste but the tech requirement.
I got it like a friendly serve to strike up the ball of truth to the dark side 😉
Audio gear (for playing/recording music) is supposed to be used as an entertainment device.
Namely, to provide pleasure by playing pre-recorded music.
It's a very simple idea.
I know there are other types of use (i.e. for communication) but in this forum i think we aren't referring to those.
-----------
Also, listening to music played through a device, is an emotional experience.
I think this is also a very simple idea, we can agree upon.
-----------
So, with the above tenets in mind:
Can you quantitatively measure the effect of such a device on an emotional experience?
Namely, to provide pleasure by playing pre-recorded music.
It's a very simple idea.
I know there are other types of use (i.e. for communication) but in this forum i think we aren't referring to those.
-----------
Also, listening to music played through a device, is an emotional experience.
I think this is also a very simple idea, we can agree upon.
-----------
So, with the above tenets in mind:
Can you quantitatively measure the effect of such a device on an emotional experience?
Last edited:
This thread is about ADCs.
Some get emotional experience by listening to the recording as such. Some need additives.
AKM does it on their evaluation boards. Same for Rohm dac evaluation board. Depends on where the components are used in the circuit.
I would not dream of using evaluation boards and claiming they are SOTA before or after hacking.
I don't think the evaluation boards I was talking about were SOA, certainly not SOA in terms of SQ. The AKM eval board may get their published measurements though.
But not what datasheet promises. Just take a look at AK4499EX evaluation board measurements.
Kind of looks like they tested the eval board with a SPDIF source, maybe with AK4191 in asynchronous mode. Not clear if the dac chip datasheet numbers were taken in the same operating mode. That is to say, its not completely clear to me if the difference in numbers has anything to do with TH parts or maybe something else possibly having more to do with configuration settings by whoever was running tests.
But its only your listening and you "internal" measurement with no reference whatsoever. You use studio productions and electronic instruments tracks a reference, sounds which no human has ever heard in real life.
//
Nonsense. The recent visitors from Italy are regular classical concertgoers. One of them teaches piano. Another amateur performer has a Steinway baby grand at home. They agreed they heard everything essentially the same as I did. We have been through all this stuff before, as you already well know.