Objectivist or Subjectivist? Give Me a Break

[DonH50]

If my reading is correct, using FFTs, we cannot be accurate in both frequency & phase - from wiki "When applied to filters, the result is that one cannot achieve high temporal resolution and frequency resolution at the same time; a concrete example are the resolution issues of the short-time Fourier transform – if one uses a wide window, one achieves good frequency resolution at the cost of temporal resolution, while a narrow window has the opposite trade-off." http://en.wikipedia.org/wiki/Uncertainty_principle

So again, the point is that the most used tool in signal processing, FFTs are less resolving than our hearing in regards to determining the full audio envelope. Decomposing this into an individual elements, such as frequency, and then showing that indeed measurements can be more accurate than the ear, in frequency determination, is missing the point - in the case of hearing it would appear that the whole is very much more than the sum of the parts!

The problem is you cannot get infinite resolution in zero time, and it is difficult to perform an FFT over infinite time... That is perhaps the fundamentla issue with FFTs as an analysis tool, you only get a window of time, and it acts on the whole window. They are great for steady-state, less so for moving (time-varying) waveforms.

 

[DonH50]

Don, I believe this paper bears more attention than a quick skim... <elided>

Agreed, that's all the time I had over lunch'ish, and I have rehearsal tonight so will have to revisit it later. I have been called to task a couple of times recently because I have been swamped at work and home, thus am a "fly-by" WBF'er lately, sorry.

 

[DonH50]

@Groucho -- On making the null: There are a variety of ways to implement distortion generators using analog circuits, from limiters and clippers to multipliers. Abd of course you'd also need filter circuits, phse adjust circuits, gain control, etc. Most of the circuits, or at least similar, can probably be found in an old National Semiconductor Op-amp App Guide. Of course I do not know exactly what Bob implemented.

 

[opus111]

Sure, there have been lots of times I was proven wrong and immediately changed my opinion.

Examples from WBF - ever recanted here? Also that was only half of your claim - how about evidence for the other half - the subjectivists that don't admit they are wrong even when faced with compelling evidence?

 

[DonH50]

...tough crowd!

 

[opus111]

Another thing I noticed is that even though -70 dB = inaudible has now been settled with absolutely no room for disagreement*, nobody here has acknowledged changing their opinion.

Remind me as I'm sometimes a bit slow - how was this settled so conclusively? Was it settled according to the opinion of Ethan Winer only?

I already listed two times I changed my opinion after learning something, and nobody acknowledged that either.

When you're blowing your own trumpet why do you need someone else to blow it too?

Nor has one person emailed me their choices about the various Wave file comparisons on my web site to show they can hear stuff at a lower threshold than my claims. I'm not trying to be a jerk here, but jeez guys, can't you man up and admit you were proven wrong?

If you'd persuade us of the relevance of those files then I'd have a go - so far, from the marketing you've done for them, they look to me totally irrelevant. Hence I have better things to do. If you really want to get people to listen then complaining about how people are wimps for ignoring you you'll find counter-productive. They'll chuckle and ignore you all the more.

*Even John Atkinson at the end of his article admitted he had been wrong:

He's a subjectivist so stands as one falsification of your earlier claim about subjectivists not being willing to admit they're wrong.

 

[Groucho]

On making the null: There are a variety of ways to implement distortion generators using analog circuits, from limiters and clippers to multipliers. Abd of course you'd also need filter circuits, phse adjust circuits, gain control, etc. Most of the circuits, or at least similar, can probably be found in an old National Semiconductor Op-amp App Guide. Of course I do not know exactly what Bob implemented.

Hello Don.

Thanks, yes, it might be possible to throw something together out of a few analog chips and bits of wire, but those old op amps and multipliers etc. were quite noisy and (in my imagination at least) far from 'hi fi'. It's a strange story altogether. I'd be imagining that the reference amp would be almost perfect anyway, and he might only need to passively mix in a fraction of a 'bent' signal into his amp to duplicate it..? - a bit far fetched though!

 

[microstrip]

Most of the circuits, or at least similar, can probably be found in an old National Semiconductor Op-amp App Guide. Of course I do not know exactly what Bob implemented.

Don,

Are you referring to the thick NI blue book on linear circuits? I still have it in my office!

Do not mind about what Bob implemented - nobody knows, and even Bob probably is not interested in it any more. In his own words in TAS223

"What began about ten years ago as an interest in the simple intrinsic nature of tube amps has evolved into almost unbelievably passionate work to extract the most possible realism using vacuum-tube topology. We find that in order to evoke the absolute sound from our tube designs, we are forced to let go of long held and cherished ideas about what an amplifier should do. We find that we are no longer interested in facsimile reproduction; rather we find reproduction that sounds enchanting and real is far more
interesting and serves our passions far better. In other words, we listen for something that could have been real somewhere in time and space, even if in the moment it’s not in our living room. Great tube amps, against all odds, can deliver music with a majesty and sweetness that is truly difficult to believe."
(...)
"I design amplifiers with tubes as well as transistors, and in each case I have built my amplifiers so they present a musical performance that could have existed in another time and space. "

 

[DonH50]

@Groucho -- No, they actually weren't bad, that was about the time op-amps started to shine and be Really Useful Engines (Thomas the Tank Engine). of course he could have used mostly passive filters as well... IIRC the amp he started with was a breadboard design, not a production amp, but it was long ago and I did not re-read the whole article (work and rehearsal cutting my time short! I am eating and typing at work whilst getting ready to grab my horns and go play in the orchestra tonight.) I remember him discussing it at some conference or show but memory fades... Starting with a good SS amp and having to natch to a tube amp is about as far apart as it gets, so I imagine it was some piece of work! I also imagine he spent quite a bit of time tinkering at "home" to figure out his methodology before going to the challenge.

@ microstrip -- Yup, although mine is several volumes, and I have the orange-covered earlier versions, too!

 

[amirm]

Guys, please do not discuss each other. Focus on the technical topic.

 

[jkeny]

The problem is you cannot get infinite resolution in zero time, and it is difficult to perform an FFT over infinite time... That is perhaps the fundamentla issue with FFTs as an analysis tool, you only get a window of time, and it acts on the whole window. They are great for steady-state, less so for moving (time-varying) waveforms.

Yep & music certainly has time-varying waveforms!!

Agreed, that's all the time I had over lunch'ish, and I have rehearsal tonight so will have to revisit it later. I have been called to task a couple of times recently because I have been swamped at work and home, thus am a "fly-by" WBF'er lately, sorry.

I wasn't taking you to task, Don, just trying to ensure that you had a good read of it as your post seemed to indicate that you missed what I thought was the main message of the paper.

 

[opus111]

Guys, please do not discuss each other. Focus on the technical topic.

Ethan introduced himself as a topic by claiming things that he did, and that others did not. Isn't a topic once introduced fair game for comment?

 

[NorthStar]

---The right balance between objectivists and subjectivists is in the evidence of the listening. :b

 

[JackD201]

What do you call an objectivist with subjectivist goals/targets?

It's like being schizophrenic except your personalities get along.

 

[Phelonious Ponk]

What do you call an objectivist with subjectivist goals/targets?

It's like being schizophrenic except your personalities get along.

In amp design? Someone who smells money.

Tim

 

[Ethan Winer]

I have seen you mention this a couple of times. Would you be so kind as to provide a link?

My AES Audio Myths video plays examples of a nasty sounding noise at various levels behind two types of music. There's a link in the video's description to download the original Wave files from my web site, versus the lossy compressed audio heard in the video. That part of the video starts at 32 minutes in. This is not a challenge to identify what you can hear, but rather just a demo of how soft -40 and -60 etc really are. If you'd rather skip the video, a similar subset of files is in my Artifact Audibility Report. But the video version is better because it plays more in-between levels.

There are several articles on my web site that do challenge people to choose which files they think are more or less degraded, and email me for the correct answers:

The Converter Loop-Back Tests article has three sets of four files each, and the reader's challenge is to identify which is the original file versus copies that went through 1, 5, and 10 copy generations using a mid-level sound card.

My article The Truth About Record Levels debunks the myth that A/D converters are cleaner at lower record levels. The reader has to choose which of the two music files had all of their tracks recorded very close to 0 dBFS and which were recorded 20 dB softer.

A lot of recording engineers believe that dither is hugely important, and affects the sound beyond the known benefit of reducing miniscule amounts of distortion. So the Dither Report offers parallel sets of files that are dithered and not dithered, and the reader has to determine which files are which.

The Converter Comparison article offers three pairs of mix files whose individual tracks were recorded through different converters ranging from a $25 SoundBlaster through a very high-end model. The challenge is to identify which mix was made from which set of files.

--Ethan

 

[Ethan Winer]

Examples from WBF - ever recanted here?

I don't recall, but probably not here. I haven't been wrong about audio science basics for a very long time!

Ah, wait, here's another egregious gaff, and this did happen since WBF has been in existence, though it's recording related so it never came up here:

Ethan's Failed Reverb Live Room Project

Also that was only half of your claim - how about evidence for the other half - the subjectivists that don't admit they are wrong even when faced with compelling evidence?

That's easy, I pointed this out just yesterday:

http://www.whatsbestforum.com/showt...ive-Me-a-Break&p=139955&viewfull=1#post139955

--Ethan

 

[Ethan Winer]

If you'd persuade us of the relevance of those files then I'd have a go

My post #1523 above explains the relevance in great detail. So yes, please "have a go" and prove me wrong!

--Ethan

 

[Ethan Winer]

One's own myths are always the most difficult to acknowledge/spot

Said the guy who has zero tests / wave files to the guy who listed a bunch of them and has actually done the work.

--Ethan

 

[NorthStar]

---I like what I'm reading on this last page (this page, page 153). :b