You may feel that is the case, but if we all to start admonishing each other on the assumption that everyone else should calibrate to our own personal outlook, where does it end? As the saying attributed to the fictional Cardinal Richelieu goes, ‘you give me six lines written by the hand of the most honest of men, I will find something in them which will hang him.’
-
Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
"Things that cannot be measured"
- Thread starter ShiZo
- Start date
You may feel that is the case, but if we all to start admonishing each other on the assumption that everyone else should calibrate to our own personal outlook, where does it end? As the saying attributed to the fictional Cardinal Richelieu goes, ‘you give me six lines written by the hand of the most honest of men, I will find something in them which will hang him.’
Pithy, this has some interesting ramifications. IQ is a sort of estimate of intelligence, isn't it? It's only a measure because it assigns a number--which may or may not be accurate at times other than when the number was assigned (and in fact, may be given as range rather than as a single hard number).
Stupidity, on the other hand, cannot be quantified, but it can be directly objectified: it is anyone who disagrees with my preferred choices for the best-sounding speakers. Right?
If your talking about home audio, sure. Studio/instrument devices, not so. Compressors, distortion pedals, reverbs even mics with full measurements need to be listened to. These are however devices that dont necessarily need flat freq or low distortion, except the dynamics processor, but how do you measure the distortion of a modulated signal.
But these might also fall in the perceptual effects category.
Has anyone seen anything about measuring instruments? These days, that speaker measurement rig and a robot player should give repeatable results. Do you think a Stradivarius measures different than a $300 violin? It must, but Is there a correlation? If there was it might make some subjectivists think twice about measurements.
Well, I suppose we could count election poll results as a useful heuristic for societal stupidity, if not a precise measure or one for individuals.
KSTR
Major Contributor
The AKM <--> TEAC co-operation in developping the AK4490 DAC chip would be one example :
https://velvetsound.akm.com/us/en/voices/interviews/teac-round-table-discussion/
Of course this is AKM's marketing version of the story and has to be taken with a grain of salt but shows the mentality. I've also had personal contract to a Japanese audio engineer and he shared similar stories. Products delevelopped is primarily measurements-driven first but then later subjective audio impressions (even with the risk of them being arbitrary) are used to change things, to "fit to taste" of whatever listening panel. Even if that compromises some measurements as long as it remains in the commonly accepted "good enough" regions. And they do all of that with great care and seriousness and respect for each other. By this, neither party gets hurt.
- Joined
- Oct 1, 2018
- Messages
- 826
- Likes
- 1,228
Emotion can be heard. And I’m not talking about the emotional connection m between the listener and the program. The emotion I speak of is what separates a good piano performance to a great one.
Go to your local opera or orchestra sometime. Nothing in that venue can be measured by a machine...
Go to your local opera or orchestra sometime. Nothing in that venue can be measured by a machine...
Now that's provocative! Anyhow, it wasn't an "answer." Lighten up for a bit of jokery, why dontcha.
- Joined
- Oct 1, 2018
- Messages
- 826
- Likes
- 1,228
I fully agree with everything written in this article. I tend to ignore posts and threads about the discussing of sexism but I thought it would be nice to share with some of you. https://www.soundandvision.com/content/death-“waf”-wife-acceptance-factor
The Mathematical concept works if implemented correctly.
But i'm referring to "FFT- analysis in Audio Measurements" in the way they are "actually" executed.
like this:
https://www.audiosciencereview.com/...nderstanding-fft-in-audio-measurements.20814/
Most audio devices used for hifi are minimum phase. For some applications, the phase plot ends up being useful (or the equivalent use of Re and Im plots) and that's a mouse click away.
- Joined
- Apr 24, 2019
- Messages
- 1,501
- Likes
- 2,841
No need to add to this, you can poorly apply any tool (or simply not know all the subtleties in using a particular tool). This does not make anyone ignorant or stupid but they should be willing to step back and consider input from those that do.
Sawdust123
Member
- Joined
- May 3, 2019
- Messages
- 49
- Likes
- 168
The problem we have regarding measurements is NOT our ability to measure. The problem is knowing WHAT to measure. And there is rightly considerable debate on this matter.
I don't think the testing methodology relevance problem will ever sufficiently satisfy the majority. Nevertheless, I do think there needs to be more transient test signals and more reactive load testing. In my years of selling test equipment, I regularly encountered companies that had devised their own tests to test their products. They felt their tests gave them unique insight and were a competitive advantage. They were reluctant to share their methodologies with any standards efforts.
I have sat on several audio standards working groups where new tests were being pondered. These efforts would frequently get off into pie-in-the-sky methodologies. My contribution was often a commentary of the practicality of the suggested test methodologies using the test equipment that currently existed in the marketplace. Some ideas were scuttled because there was no practical means for the average company to implement the test being considered. However, ideas that we may not have been practical 5 years ago may be practical today. The situation is never static.
- We use sine waves because they are simple to use and can be easily calibrated. Music is clearly not a sine wave.
- We test into purely resistive loads because they allow simple comparison. Speakers are not purely resistive.
- We test anechoically because room reflections are unpredictable. We never listen anechoically.
I don't think the testing methodology relevance problem will ever sufficiently satisfy the majority. Nevertheless, I do think there needs to be more transient test signals and more reactive load testing. In my years of selling test equipment, I regularly encountered companies that had devised their own tests to test their products. They felt their tests gave them unique insight and were a competitive advantage. They were reluctant to share their methodologies with any standards efforts.
I have sat on several audio standards working groups where new tests were being pondered. These efforts would frequently get off into pie-in-the-sky methodologies. My contribution was often a commentary of the practicality of the suggested test methodologies using the test equipment that currently existed in the marketplace. Some ideas were scuttled because there was no practical means for the average company to implement the test being considered. However, ideas that we may not have been practical 5 years ago may be practical today. The situation is never static.
@scott wurcer i assumed its obvious i'm not talking about FFT as the mathematical concept or algorithm in generals but "FFT in Audio Measurements" and the way they are presented here.
Obviously if one has all the data the FFT function is reversal.
But also kind of Obviously several seconds of audio capture express as one "FFT measurement" screenshot do not contain all the data.
And because of this extreme data reduction (cause for example by averaging) and because we don't see the complex output.
There is the possibility for audible flaws to "hide" in this processed data.
So wen i said there are Things that cannot be measured with "FFT"
This dose not contain the whole story and not all flaws would properly show up in this "FFT"
Obviously if one has all the data the FFT function is reversal.
But also kind of Obviously several seconds of audio capture express as one "FFT measurement" screenshot do not contain all the data.
And because of this extreme data reduction (cause for example by averaging) and because we don't see the complex output.
There is the possibility for audible flaws to "hide" in this processed data.
So wen i said there are Things that cannot be measured with "FFT"
This dose not contain the whole story and not all flaws would properly show up in this "FFT"
Last edited:
solderdude
Grand Contributor
Only if the flaws are 'random' which flaws won't be. They will repeat and stand out as 'poles' or when a bit 'random' as an elevation or skirting in the FFT.
The emotion is in the performance, not created by the equipment.
If the recording is half way to being well done any emotion in the performance will be felt in the reproduction.
In fact IME getting the emotion from a superbly emotional performance doesn't need especially high quality reproduction, since it often comes from rubato and keying/fingering subtleties which are perfectly well heard on a modest hifi system.
I would much sooner hear one of my great Artur Schnabel Schubert recordings - which were historical from a SQ pov - than some of those awfully banal high quality recordings used at shows.
The musicality is blindingly obvious despite the historic recording quality.
IME when an audiophile says "musical" or "analogue" they are usually referring to a system with rolled off treble...
Yes, that's what i said before.
Thy can be relatively random and or sparse. don't need to be perfect true random. just constantly changing.
solderdude
Grand Contributor
What would cause a constantly changing or sparse distortion ?
BinkieHuckerback
Addicted to Fun and Learning
- Joined
- Feb 16, 2021
- Messages
- 796
- Likes
- 1,283
Surely anything of import can be measured?
In general digital problems. clock drift, bit flips in some byte patterns, beat frequency of close but not locked clocks.
On the analog side maybe. DCDC converters that change the frequency randomly, maybe because of temperature or load variations.
Or because they intentionally do spread spectrum.
We could speculate a lot of reasons, whats the point?
