Testing audiophile claims and myths

[castleofargh]

moderator rant:
guys. I know that it's rarely possible to only fight the ideas instead of the people when the arguments are about very personal opinions. but please try anyway. just replace liar/dishonest/ignorant/etc by "wrong" or something less offensive.

@gregorio IMO you're going too far. if you read the all thing, Keith is as always, just impressively optimistic about what we record and what we can, or will make use of later on. he expects some ultrasonic content to be music(not right out impossible), he expects that high res may have a small audible difference in some ways(also not impossible, although probably not a concern for adults), and no matter how subjective his experiences, he stated many times the nature of said experiences, how he's only sharing impressions, not claiming to prove anything. he even mentioned a few times how even those impressions of his are about tiny variations.
that general mindset doesn't deserve the dishonest BS liar treatment that you've been delivering(also those can very much be seen as personal attacks and you know Head-fi doesn't want to host those). just like you, I see trouble(and sometimes danger) in including unsupported possibilities into our decisions and technological choices. because in practice, many of those "we haven't disproved it so we shouldn't dismiss it", will turn out to be unfalsifiable ideas. meaning that those ideas once accepted, are never going to be disproved and are never going away. and that's bad for everybody!!!!!
so don't go thinking that I'm troubled by your position. as usual the issue is how aggressive you are. a few people have complained about it so I get out of my cave. analogsurviver defending the superiority of vinyl over CD with a calm demeanor, is better than you insulting people while trying to defend a solid fact. that's probably not what you think, but it doesn't matter because:

Our forum rules have one major purpose: they allow a respectful exchange of ideas.

if you can't do that, the rest becomes irrelevant. I enjoy learning things from you, and on several occasions I've had people PM me to basically say that they'd rather have me ban 2/3 of the forum than losing you and the knowledge you bring. so as a modo trying to satisfy most people, there is only one working outcome, and that's a Gregorio who can argue a point and stop there.



non modo whatever:
@KeithEmo if I have one advice for you, it's to lighten up on the analogy combos. you've brought up a lot of them in the last few days and IMO they're not serving your argument very well. aside from the usual deathtrap that an analogy often turns out to be, there is the consistent comparison between stuff that may or may not be, and some clear proved facts from the analogy. and while you don't try to force it like a guy claiming that we need high res audio because we see a difference on a 4K TV screen, the analogy does associate things on very different levels when it comes to having evidence. for guys like Greg and myself, who are now wired to be suspicious, you can be sure that we're reading those posts and thinking that you're trying to mislead people into associating facts with maybes as if they're all facts. and you being an industry insider probably doesn't help our bias, whether we know it or not. sorry ^_^.

 

[Davesrose]

1. Yes, provided ALL the following are true:
A. Consumers actually have some old scratched slides to start with.
B. The slides contain some infrared information to start with.
C. The infrared information (when there is any) on those slides actually contains the details necessary for it to be of any use in the removal of scratches.
D. Science has figured out a theory for how to extract, analyse and use those details in the infrared information to remove scratches.
E. Technology exists which actually implements that theory.
F. Consumers actually own that technology (an infrared slide scanner).

That's not how ICE works. The premise of it is that the slide doesn't have any relevant information in the infrared band. A scanner that is ICE compatible will have an infrared band, which then records dust and fingerprints that show up in the infrared band and will be subtracted and filled in by the software. It's clearly a different technology, so not really an analogy in recording ultra-high sound frequencies (which that argument seems to be if there's more precision/ more realistic model of environment).

 

[KeithEmo]

I think you're looking at it from a very odd point of view.

In general, when we look at a slide, we ASSUME that everything we're seeing is "relevant image information".
However, we also accept that this is an assumption, rather than a certainty.
(We may not be able to visually tell the difference between a scratch, a power line, and a far-away contrail.)

But your assumption that the ORIGINAL SLIDE didn't have any relevant IMAGE information on it in the IR band is incorrect.
If I'd done an ICE scan on it, it would have shown WITH CERTAINTY that there were no scratches, which is better than as assumption.
That useful information is knowing that: "There are no scratches, so everything you see is actually intended image content".
(Maybe that vertical line is a power line, or a contrail from a UFO, but we would KNOW it wasn't just a scratch.)

And, by doing that infrared scan on my damaged copy, I now KNOW where all the scratches are.
This is useful information because it prevents me from confusing those scratches with lines that belong on the picture.
Even with no further action this information may prove quite useful.
However, even beyond that, this information lets me make higher quality corrections of those scratches.
The result is an image that is both visually more attractive and more accurate.

Here's the analogy:
The original recording venue is "the original perfect slide".
The final mixed down copy is "the potentially scratched slide".
Assuming that I can extract information from whatever exists in the ultrasonic band.....
1) it is simply information that I might simply find useful
2) specifically it might help me understand and/or correct flaws in the original recording

For example, by analyzing the ultrasonic reflections in aggragate, I may determine that the main venue was fifty feet square.
Furthermore, I may be able to conclude that the drums were located somewhere near the center.
And, if, on one particular track, there is a return from a drum hit at 12 milliseconds, that would be an anomaly.
(It would correspond to an echo from something twelve feet from the drums.)
Perhaps this would indicate that some extra drums were "layed in" that were recorded elsewhere.
Perhaps someone added some studio drums, and added some reverb to attempt to match them, but was a little sloppy.
And, perhaps, my new super-duper decoder can remove the anomalous reverb, and so make the recording sound "more natural".

The analogy is simply to "using out-of-band information to make better corrections of in-band information".
And I think "invisible infrared light" is quite analogous to "inaudible ultrasonic information".

That's not how ICE works. The premise of it is that the slide doesn't have any relevant information in the infrared band. A scanner that is ICE compatible will have an infrared band, which then records dust and fingerprints that show up in the infrared band and will be subtracted and filled in by the software. It's clearly a different technology, so not really an analogy in recording ultra-high sound frequencies (which that argument seems to be if there's more precision/ more realistic model of environment).

 

[Davesrose]

I think you're looking at it from a very odd point of view.

In general, when we look at a slide, we ASSUME that everything we're seeing is "relevant image information".
However, we also accept that this is an assumption, rather than a certainty.
(We may not be able to visually tell the difference between a scratch, a power line, and a far-away contrail.)

But your assumption that the ORIGINAL SLIDE didn't have any relevant IMAGE information on it in the IR band is incorrect.
If I'd done an ICE scan on it, it would have shown WITH CERTAINTY that there were no scratches, which is better than as assumption.
That useful information is knowing that: "There are no scratches, so everything you see is actually intended image content".
(Maybe that vertical line is a power line, or a contrail from a UFO, but we would KNOW it wasn't just a scratch.)

It's not a POV: it's clarification on the technology. ICE uses IR to detect dust and oil on a color slide or negative (so it's limited to only certain photographic media). It's based on the assumption that there is no relevant visual information that was recorded in the IR band, and only physical defects on the surface of the "recording" will be isolated in an IR scan. No IR information was recorded in the slide...and your original arguments have been asking if one should record/obtain higher frequencies in the recording itself.

 

[KeithEmo]

I like your comparison to the GPS system.

However, I will note that, in many situations where we don't happen to have beacons already in place, we EXTRACT information from existing sources.
For example, we calculate the locations of cosmic events, based on information extracted from background radiation levels, pulsar emissions, and other cosmic events.
And, long before we had GPS satellite beacons, people were navigating based on the position of naturally occurring "beacons".
(For example, by pointing a sextant at the sun and the moon, and calculating our position based on information we extracted about their positions.)

This is analogous.

Yes, it's EASIER, and often more accurate, when you have the opportunity to use accurate beacon signals, specially designed for the purpose.
So, yes, a GPS reciever is much better at the job than a sextant.
However, people have been achieving "adequate" and "useful" results with sextants and similar instruments for a long time.
Likewise, having a transmitter, or a full array of ultrasonic beacon signals, would be the BEST way to determine this sort of information.
But that in no way suggests that it's the only way, or that other ways cannot yield useful results.

Yes, if you want to measure a large room, the best method is to use an ultrasonic rangefinder, or a LASER measure.
However, you can still obtain a useful approximation by clapping your hands and timing the echoes.
(Or timing the echoes on a recording that was made in that cave.)

1. Yes, provided ALL the following are true:
A. Consumers actually have some old scratched slides to start with.
B. The slides contain some infrared information to start with.
C. The infrared information (when there is any) on those slides actually contains the details necessary for it to be of any use in the removal of scratches.
D. Science has figured out a theory for how to extract, analyse and use those details in the infrared information to remove scratches.
E. Technology exists which actually implements that theory.
F. Consumers actually own that technology (an infrared slide scanner).

2. Wow, a willingness to concede that something that doesn't exist, really doesn't exist, that's novel!
2a. And here we go with the suggestions, suggestions which are not only unsupported by any evidence but that actually contradict all the evidence.
2b. Yep, you keep doing that. It all sounds reasonable and therefore, your assertion that "it seems like a pretty obvious next step" isn't too much of a logical leap and seems entirely reasonable. I can't prove it, but being a ProTools expert of many years, it seems "pretty obvious" to me that ProTools will NOT include an infrared slide scanner in the foreseeable future and actually, I find it pretty absurd to even suggest that it might, regardless of how reasonable it might sound to others! Wait a minute ... are you NOT asserting that ProTools will include an infrared slide scanner? Was the slide scanner thing just an analogy, another one of your completely non-analogous analogies?!

Let's look at YOUR analogy shall we, and apply it to commercial music/sound: Assuming "A" is a commercial music/sound recording, then it is true. "B" is only true sometimes (only sometimes is there anything other than useless information in the >20kHz band). There's no shred of evidence to suggest that "C" is ever true, in fact all the evidence indicates that it MUST be false. D, E and F are also false. So out of the 6 requirements, ALL of which have to be true, in fact only one and a half are. Clearly then it's a terrible analogy, so how come it sounds so reasonable? Simple: You gloss over the fact that "B" is only sometimes true. You omit to even mention "C", deliberately ignore every request for any evidence that "C" exists and instead misrepresent other unrelated facts as providing that evidence. You claim "D" is true and continue to do so even when the scientific evidence is presented proving that it's false. You get away with E and F by stating they may one day be true but that too is false because E and F can NEVER be true until after C and D are both true.

We're all used to the rapid advance of science and digital technology, that provides products/solutions to things that seemed impossible or even unimaginable. It's this expectation of science and technology that KeithEmo is fallaciously abusing to make his suggestions seem so plausible! In reality, science and technology rapidly advances and achieves the seemingly impossible by creating detailed information, then extracting and using that information. For example, the global positioning system (GPS) works by having satellites provide extremely detailed timing information and then us owning technology which extracts and uses that information. How would the GPS system work if there were no satellites? How would it work if there were satellites but they weren't providing any timing information? Even if the technology of extracting and using satellite timing information advanced to a level that isn't even imaginable today, that still wouldn't make the slightest bit of difference if there were no satellite timing information in the first place! To bring this analogy back to music/sound: Once we start using radar or sonar transmitters (or bats) in the studio and equipment to capture the reflections of those signals, then I'll be happy to record at a high sample rates to capture that detailed information, provided it doesn't affect what I'm able to do in the frequency band that is audible and provided there may actually be the potential for that information to be beneficial and a demand for it. So far, only 2 potential benefits have been suggested, one which wouldn't actually be a benefit and the other that's too laughably ridiculous to even repeat!

G

 

[KeithEmo]

Apparently there have been several variations on the ICE process over the years... each of which works to different degrees on different types of slides and negatives.
However, most of them DO in fact claim to detect and repair "dust and scratches".
If you check the literature, you will find descriptions about which variations, present in which scanners, do or don't work with specific film types.

- On B&W slides you can use the fact that the emulsion is opaque to IR to identify scratches, which extend through the emulsion, as being "perfectly bright in the IR spectrum".
(However, since the emulsion is opaque to IR this will only work to detect scratches that extend through the emulsion to the base layer.)
- With some color processes, the dyes and emulsions are somewhat transparent to IR light, and so both dust and scratches may show up as dark (opaque).
(In this process, the dust and scratches contrast because they are MORE opaque to IR than the emulsion.)
- And, apparently, certain clor print emulsions are more or less opaque or translucent to IR, and it doesn't work at all with them..

There have even been some processes that use other types of scans - like polarized light.
However, what they all share in common is that they use the "image" obtained by scanning the negative or slide with something that is invisible to humans.
Then using the information obtained from that scan to intelligently identify scratches and dust so as to guide corrections of the visible image.

Virtually all consumer slide or negative scanners use one particular variation on this... and, from reviews, each seems to work best in only certain situations.
However, in forensic analysis and art restoration, it is not unusual to try a wide variety of visible and invisible spectra, to see which works best in a particular situation.

It's not a POV: it's clarification on the technology. For one thing, ICE doesn't detect scratches: it detects dust and oil on a slide or negative. It's based on the assumption that there is no relevant visual information that was recorded in the IR band, and only physical defects on the surface of the "recording" will be isolated in an IR scan. No IR information was recorded in the slide...and your original arguments have been asking if one should record/obtain higher frequencies in the recording itself.

 

[bigshot]

moderator rant:
guys. I know that it's rarely possible to only fight the ideas instead of the people when the arguments are about very personal opinions. but please try anyway. just replace liar/dishonest/ignorant/etc by "wrong" or something less offensive.

Your wish is my command.... Keith is wrong. So are Analoguesurvivor and our hot mastering friend (I always forget his name).

That was a lot easier than having to explain the specific reasons why their thinking is wrong. It's quicker for me too. I can just get a rubber stamp that says, "YOU'RE WRONG" and politely reply to their posts faster that way.

 

[Davesrose]

Apparently there have been several variations on the ICE process over the years... each of which works to different degrees on different types of slides and negatives.
However, most of them DO in fact claim to detect and repair "dust and scratches".
If you check the literature, you will find descriptions about which variations, present in which scanners, do or don't work with specific film types.

- On B&W slides you can use the fact that the emulsion is opaque to IR to identify scratches, which extend through the emulsion, as being "perfectly bright in the IR spectrum".
(However, since the emulsion is opaque to IR this will only work to detect scratches that extend through the emulsion to the base layer.)
- With some color processes, the dyes and emulsions are somewhat transparent to IR light, and so both dust and scratches may show up as dark (opaque).
(In this process, the dust and scratches contrast because they are MORE opaque to IR than the emulsion.)
- And, apparently, certain clor print emulsions are more or less opaque or translucent to IR, and it doesn't work at all with them..

I've revised my post to include a good video on the technology of ICE and why it doesn't work with B&W negatives (which uses silver nitrates and can block IR). The advertising isn't misleading when they say it can "detect dust and scratches" (apparently there is some success with scratches still being physical damage that can be detected). The premise is that it's detecting physical imperfections using a separate band from the actual recorded spectrum. That's a different topic than asking if >20khz frequencies already in a digital recording are relevant.

 

[bigshot]

Infra Red Herring!

 

[KeithEmo]

I don't understand the perspective from which your claim originates....

ICE intentionally uses "a separate band from the actual recorded spectrum".

If we were to agree that "ultrasonic content in a recording is totally inaudible"...
And we were to also agree that "there is still some sort of information in the ultrasonic spectrum of some recordings"...
Then that information is just "information is a separate band than the audible recorded spectrum"...
And I'm simply suggesting that it is POTENTIALLY POSSIBLE to detect and correct errors in the audible portion of the recording by using information we extract from the inaudible portion.

The ONLY difference is that the ICE process relies on taking extra steps to GENERATE the additional information rather than extracting and using "extra" information already present.

I've revised my post to include a good video on the technology of ICE and why it doesn't work with B&W negatives (which uses silver nitrates and can block IR). The advertising isn't misleading when they say it can "detect dust and scratches" (apparently there is some success with scratches still being physical damage that can be detected). The premise is that it's detecting physical imperfections using a separate band from the actual recorded spectrum. That's a different topic than asking if >20khz frequencies already in a digital recording are relevant.

 

[Davesrose]

I don't understand the perspective from which your claim originates....

ICE intentionally uses "a separate band from the actual recorded spectrum".

If we were to agree that "ultrasonic content in a recording is totally inaudible"...
And we were to also agree that "there is still some sort of information in the ultrasonic spectrum of some recordings"...
Then that information is just "information is a separate band than the audible recorded spectrum"...
And I'm simply suggesting that it is POTENTIALLY POSSIBLE to detect and correct errors in the audible portion of the recording by using information we extract from the inaudible portion.

The ONLY difference is that the ICE process relies on taking extra steps to GENERATE the additional information rather than extracting and using "extra" information already present.

I'll try to re-iterate. ICE can only be used with analog color slides and C-41 based color negatives. The software is taking the first scan that's within the visible spectrum (the dust and scratches will be visible in that scan). But the ICE scanner makes a separate IR pass (and digitization) that relies on a spectrum that does not include the photograph's recorded visible color spectrum. It's physically separating the original color information from physical imperfections with certain analog sources (based on there being a limited recorded spectrum, and imperfections being detected on a different band). The premise for ICE is that there's a separate pass using a spectrum that wasn't in the original recording: and that is used to isolate. This is separate from your arguments about "ultrasonic content in a recording" (ICE is relying on physical imperfections of analog source, and is incongruous to analyzing a whole spectrum in one digital recording).

 

[Phronesis]

I think there's an issue in the discussion with how the word "audible" is being used. You need to define it precisely, otherwise there will be arguments because people aren't talking about the same thing. To me, the broadest definition of "audible" is that if a sound at the eardrum, or difference in sound at the eardrum, results in an auditory nerve signal, or a difference in that signal, it's potentially "audible" if there's also a change in the brain response. That wouldn't necessarily require a conscious perception of sound or difference in sound, though a subconscious effect would be expected.

 

[KeithEmo]

I stand corrected... it seems to be widely agreed that the commercially developed ICE process works very poorly, or not at all, with B&W emulsions.
(Although I have read early white papers suggesting that it could be used to detect and remove scratches that actually penetrate the emulsion.)

It seems that a slightly different variation, which uses polarized light, rather than IR, is recommended for detecting surface scratches in B&W film.
Here are a couple of papers on the subject (I don't know if there are any commercial products that use this.)
(Sorry about the long second Google link... but it does work.... the direct link seems to be inaccessible due to lack of permissions.)

https://www.ccaaa.org/images/tinyUpload/import/jts2010/10.Rudolf_Gschwind.pdf

https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&cad=rja&uact=8&ved=2ahUKEwjx-tvw5d7eAhWlVN8KHRUqDSUQFjACegQIBxAB&url=http://dl.acm.org/ft_gateway.cfm?id=2597894&ftid=1550494&dwn=1&CFID=808360414&CFTOKEN=19365920&usg=AOvVaw2RY3Wpb6V_HagAUCR_EYgC

https://infoscience.epfl.ch/record/188640/files/ICIP2013_defects_final.pdf

I don't quite see the significance of whether the extra information comes from an explicit "extra pass"...
Or whether it is already present, but simply assumed to be "inaudible".

I'll try to re-iterate. ICE can only be used with analog color slides and C-41 based color negatives. The software is taking the first scan that's within the visible spectrum (the dust and scratches will be visible in that scan). But the ICE scanner makes a separate IR pass (and digitization) that relies on a spectrum that does not include the photograph's recorded visible color spectrum. It's physically separating the original color information from physical imperfections with certain analog sources (based on there being a limited recorded spectrum, and imperfections being detected on a different band). The premise for ICE is that there's a separate pass using a spectrum that wasn't in the original recording: and that is used to isolate. This is separate from your arguments about "ultrasonic content in a recording" (ICE is relying on physical imperfections of analog source, and is incongruous to analyzing a whole spectrum in one digital recording).

 

[bfreedma]

When an analogy proves to be wrong, and then an attempt to make it correct requires yet another dart throw, it probably isn’t a good analogy.

But hey, another dozen posts trying to equate photo reproduction with audio reproduction will be fascinating...

 

[bigshot]

But hey, another dozen posts trying to equate photo reproduction with audio reproduction will be fascinating...

You are easily amused! I'm just waiting for when we go back to talking about things related to listening to recorded music in the home. No one ever seems interested in that. They're too busy trying to think up extreme conditions that no consumer ever has to deal with.

Ultrasonic frequencies are as useful to a home audio system as teats on a bull hog.