Who's afraid of DBTs

[spdif-usb]

I wouldn't call picking correctly 60% of the time "completely reliable".

There is no mention of "picking correctly 60% of the time" anywhere in the paper. The confidence level is above 95% for the results of all tests except one test, IIRC.

 

This study used 459Hz transition band. Sox would use a default value of 1.1kHz. What transition band were you thinking of that was needed for 44.1?

For 44.1 kHz sampling frequency, it simply doesn't matter what the width of the transition band is because then the filter artifacts can, and often will become audible under normal listening conditions in a decent Hi-Fi setup playing high quality recordings. Participants of these tests were not all golden ear type listeners BTW, so... That said, I am able to reliably and clearly hear differences even when using the far superior filtering of iZotope RX Advanced (latest version) instead of SoX. This is regardless of which filter settings are used. On a playback system that some would describe as Mid-Fi instead of truly High End. Despite the fact I do not have better than average hearing.

 

To you, these observations might be completely worthless, and that's perfectly fine. To me, personally, however, the fact I was able to pass the tests with very little trouble using foobar2000 and ABX Comparator component for foobar2000 is telling me everything I need to know plus more. I actually even have removed the ABX Comparator component from my foobar2000 a long time ago because I am not the type of person who keeps looking for answers to questions that have already been answered correctly. It's because, unlike some other people, I am not afraid to accept the final truth. The final truth being that the audible difference between 44.1 kHz sampling and higher than 44.1 kHz sampling is NOT a product of people's imagination or expectation bias, and that all opposite claims are, typically ironically in fact, subjective accusations made by those who would rather die painfully than admit they were wrong, as was greatly to be expected.

 

Uhhh, I don't see any part in the discussion section where they call their main testing method worthless.

The fact the Meyer & Moran testing method was simply wrong is very extremely easy to demonstrate, so I am not going to dig into that on here, but re ABX testing [AES Convention Paper 9174] states the following:

ABX tests have a high sensitivity, that is, the proportion of true-positive results out of total positive results is high. However, ABX tests also have low specificity, meaning that the proportion of true-negative results out of total negative results can be spuriously low. Translating this into outcomes in psychophysical tests, the proportion of the time that a listener scores well on an ABX test by chance is low, but the proportion of the time that a listener can score poorly on a test in spite of being able to discriminate the sounds is high. An ABX test requires that a listener retains all three sounds in working memory, and that they perform a minimum of two pair-wise comparisons (A with X and B with X), after which the correct response must be given; this results in the cognitive load for an ABX test being high.

That's super.

Isn't it?

[PeterSt]

Here is another pose of which I would say that DBT can not work. The "blind" part is crucial now, because in practice it will be A-A and you won't know it.

 

I for myself (meaning my own judgement) will be fairly confident that I dig out differences. At least when there's one round only. This is because, and as I told earlier on in the thread, your focus won't stay on the same things.

 

I can't guarantee it, but I guess I would come up with some difference, while both are exactly the same.

So although I am capable of a few things in this realm, I certainly wouldn't trust myself.

[trithio]

OK, but this is a risk reduction methodology based on two flawed tests.

Others will have different approaches & work on the basis that if they make a mistake they can probably sell on the 20K device with not too much hit to their pockets (if they are paying 20K they can probably take some hit anyway) & use it as a learning experience.

I just don't let risk reduction methodologies affect my decisions.

 

After the whole 'discussion' with sbgk about that post, I am afraid of it . But thanks for confirming that it was at least very clear.

And yes that is basically a risk reduction methodology. And you are subestimating the risk a bit because the pocket-hit on reselling a 20K device may be quite severe. There is a thread here something like 'your most crazy audio story'. User Daphne has a story there about buying a ~40K amp for ~16K. 4 months old.

And look. I'm not saying that mine is the only way. Or the best. You can also rent/try devices at home and you'll also have almost no risk or loss. Or many other ways to get the same results. My post was just a sample logic-anecdote to ilustrate that even imperfect DBTs may be better than no test and/or simple sighted tests.

 

With all that aside. DBTs are a tool and same as any other tool are neither inherently bad, nor inherently good, nor perfect, nor good for everything. How you use it is always the key factor. And to me the whole DBT bashing campaign sounds like a dubious form of "knives are evil cause joe cut his fingers. lets ban knives everyone'

[spdif-usb]

You are kidding, right? Certainly you don't mean to accredit the paper based on AES self serving response. In what marketing universe would AES possibly support ABX or DBT?

Come on dude, you're smarter than that?

No, I am not smarter than people from the likes of Bob Stuart, nor would I dare claim to be. Then again, I feel smart enough to understand that your subjective remarks are not helping me ignore the fact you clearly aren't smart enough to trick me into changing my view... so no real reason for me to start being afraid still........

[spdif-usb]

DBTs are a tool and same as any other tool are neither inherently bad, nor inherently good, nor perfect, nor good for everything. How you use it is always the key factor.

Amen!

[PeterSt]

To you, these observations might be completely worthless, and that's perfectly fine. To me, personally, however, the fact I was able to pass the tests with very little trouble using foobar2000 and ABX Comparator component for foobar2000 is telling me everything I need to know plus more. I actually even have removed the ABX Comparator component from my foobar2000 a long time ago because I am not the type of person who keeps looking for answers to questions that have already been answered correctly. It's because, unlike some other people, I am not afraid to accept the final truth.

 

Well said and in my view, all very well said. However, there's one part I can't agree with per se, hence it requires more to be definitive :

 

The final truth being that the audible difference between 44.1 kHz sampling and higher than 44.1 kHz sampling is NOT a product of people's imagination or expectation bias, and that all opposite claims are, typically ironically in fact, subjective accusations made by those who would rather die painfully than admit they were wrong, as was greatly to be expected.

 

What's wrong with this is that you assume the perfect 44.1 filter, which clearly is not that, no matter iZotope RX etc. I don't claim I have that perfect filter (because can not exist), but this is my experience :

 

So I have my Arc Prediction. All what is out there is so much "not on par" that you give me 1 second only and I'll still have a 100 out of 100. Nothing new, because you can do it too, as you say. But now the difference :

Arc Prediction heavily relies on measurement. Alas, measurements to my own standards and ideas, but still. Now, this is from many years ago by now, and some times I try to improve on it. Not because I am in lack of anything, but because I know the perfect filter does not exist. So I can slightly improve on one aspect (say the frequency domain) but can see (measurement) how it degrades the other somewhat. I also know what to look for, and some times I can see I must have done right because what's hardly visible should not be audible.

But it always is. And a bit depending on the music, I still can hear it in one second.

 

That I can hear it in one second is nice (can be for the better I suppose), but it is about how after a while you can hear the flavor of the filter and if there's on thing I can't stand it is flavors. So neutrality is my motto. But now what I am really saying :

When something is not right then it WILL express in flavors. Not at day 1 then at day 25 if necessary. But that is what "wrong" is. Simple actually.

 

What follows from this is that I don't need Hires, nor do any of my customers that I can see. There's just no one coming up with the subject hardly (while in here (CA) it's the talk of the day). So to be clear : no Hires is going to improve anything, and sadly it can only be the other way around (because of flawed Hires).

 

So the real motto of my response : even when things seem so clear (like only Hires can solve the issues of the required filtering), this still can be wrong judgment because of some base being not right. In this case it would be SoX or iZotope or players of name and fame ... it all doesn't matter because they all hurt my ears (and not because I have sensitive ears - only because I am used to "better" if that is allowed to say). So how would I ever use those to judge that Hires is better ? It will. And if not the Hires is flawed.

 

There is so much perspective to everything !

And please do not continue on this Hires stupid stuff; other threads are for that.

[trithio]

... you clearly aren't smart enough to trick me into changing my view...

 

That is very dangerous logic. Here's a sample situation: 10 wise men cannot 'trick' a stubborn idiot into doing the right thing.

Just saying . I have no bone in your 'fight' with mayhem. Dont want any either.

[mmerrill99]

If I understand you correctly, Peter(?) - what we are talking about is an increase in perceived dynamics?

Of course this perception is as a result of a change in the audio signal stream but I'm of the opinion that it is a very subtle change that causes out auditory perception to sense a better illusion being created. I agree that it should be measurable but I don't think our measurements are sophisticated enough yet to be able to do so?

 

Although I fully agree with "perceived effects" I never dig that. I mean, from what I perceive myself. There's always a mere technical reason to be found. So on this one, "obviously not a measured increase in dB" :

 

First off I would be careful with this, because it is very easy to measure an increase in dB, where it about all places in the room where we can measure that. So would you, for example, perceive to be suddenly really in the church (but it is your listening room) - and which expresses by a very wide sound stage and being in the middle of the music all over ... this then really is so and can be measured by means of an SPL meter (I have done those experiments).

 

More down to earth, think like this when music is perceived louder, when less noise would be a subject :

 

All what is more "square" but was flattened (smoothed) by the noise at first, is now more square as such (thus physically so) and this more high transient is what we perceive as more "hurting"; Easily to translate to "more loud".

That meanwhile the bass gets more solid (in the same situation) is just following the same logic. But for the longer wave the work out is different; now the noise which rides on one cycle of it gets less and thus the wave becomes literally more solid. Exactly the opposite of the smoothing happening in the higher frequencies which acts upon more cycles at once (but investigate the noise pattern).

 

PS: When I perceive changes as such, they never ever go without the reasoning of why. To that regard do notice that changes are always intended, often down to very technical elements. Some times the change works counter productive and then too I want to know why. Example of that : when I know noise is lower but sound gets nasty because of that, I'm fairly confident I hit another noise source, that now being more profound (read : less random). So on to the source of that.

In the end it gets more and more difficult to improve because chances get higher and higher that things inherently (always in there) get more profound, only because you took out other misery.

Nothing goes by accident and nothing is voodoo either.

[mmerrill99]

With all that aside. DBTs are a tool and same as any other tool are neither inherently bad, nor inherently good, nor perfect, nor good for everything. How you use it is always the key factor. And to me the whole DBT bashing campaign sounds like a dubious form of "knives are evil cause joe cut his fingers. lets ban knives everyone'

Yes, that's exactly correct & one of the points I wish to make. DBTs should not be used by anybody except those who are well trained in the area of perceptual testing - they are not for hobbyist use.

 

Unfortunately, this is not normally the case - we have all sorts of uninformed blind tests trying to "prove" that "night & day" differences do not exist. The reality is that the "normal usage" of DBTs results almost always in null results, as expected. This would not be a problem if people realised that this is mainly due to bad test design but they don't. So the prevailing impression in the mind of a large group of people in this hobby is that DBTs prove the delusion of those who claim they hear differences.

 

Agreed that the usage of the tool is the problem & I would suggest that the tool itself can be so difficult to use correctly that it is used incorrectly 99% of the time.

[PeterSt]

If I understand you correctly, Peter(?) - what we are talking about is an increase in perceived dynamics?

Of course this perception is as a result of a change in the audio signal stream but I'm of the opinion that it is a very subtle change that causes out auditory perception to sense a better illusion being created. I agree that it should be measurable but I don't think our measurements are sophisticated enough yet to be able to do so?

 

If you receive it as more loud (as you said) ? No, I think what I said applies in that case. But there is more;

 

I don't know who told about it (I think it was Paul), but while listening to music (sound regarding that), our visual system is highly active. Why ? because we always try to associate with familiar things from memory. All starts with being alert to danger, which btw is also why certain frequencies are more sensitive to us (think 1200 - 5KHz or so). The snap of a branch, human voice. So visuals sure are recorded in our memory. Now when things match, the foot tapping etc. occurs. We can associate.

 

I have told about it before - this "system" is very complex; when you listen to (and through) your gear, at first you will be able to be satisfied when hearing a cymbal (happy happy) without attack and thus actually without the real source of the sound. This, and this is the interesting part, while you surely know the cymbal from live. You can even see it happening (Cobham hitting the cymbal). Still how your system shows it is also fine. At first !

But wait for the day that the attacks on the cymbal come through as well, through your speakers. From that day on you can't be without. So if now that's taken out again, your brain will hurt and it will be very tiring to longer listen to that. It just ain't right. Not any more.

This is where the filters are crucial. So let them ring too long (compare to none previously) and you will hear it. That the absolute frequency/harmonics (which are 100s for a cymbal) isn't right ... so what. You wouldn't know better anyway. But if something which is to be sharply boundaried while it is not, there's a mismatch with ... well, some kind of combination of the visual memory and what we can synthesize from the elements of sound. At least that is how I see it.

 

Your next problem when you have this all quite right, is that the live concerts start to suck, because no way good sound comes from there (usually). Still it is as live as can be, including the visuals. But then of course someone thought to lead all through microphones, or otherwise the hall has too poor acoustics etc.

 

So in the end I agree with you. But it depends on what exactly is the change and how it is perceived. Increase in dynamics is usually allowed, *if* a better fundament goes along with it; without that sound gets lean and sharp and will show over-detail, sounding nasty (harsh) meanwhile (not always, the nasty, but too lean yes).

[Jay-dub]

For 44.1 kHz sampling frequency, it simply doesn't matter what the width of the transition band is because then the filter artifacts can, and often will become audible under normal listening conditions in a decent Hi-Fi setup playing high quality recordings. Participants of these tests were not all golden ear type listeners BTW, so... That said, I am able to reliably and clearly hear differences even when using the far superior filtering of iZotope RX Advanced (latest version) instead of SoX. This is regardless of which filter settings are used.

 

I presume you're saying that all lowpass filters with a stopband of 22kHz or less are potentially audible, not that the width of the transition band is irrelevant to the resulting sound quality. It still isn't a reason not to say what the transition bandwidth is for the iZotope RX Advanced filter that you like. I doubt that I'll try iZotope RX Advanced, though -- I stick with SoX because I'm cheap and lazy and like the sound I'm used to. And I'm happy to say what transition bandwidth is needed for the filter to sound practically perfect to me: I used SoX sinc with transition bandwidth of 2400-3000 Hz, stopband attenuation 60dB. Anything less than 1500 Hz definitely sounds bad to me. This is for frequency cutoffs anywhere from 18-21 kHz. With these settings I'm pretty sure that any degradation I hear is due purely to the lost high frequencies and therefore cannot be improved upon by other filters with the same frequency cutoff, and does not interfere with my musical enjoyment.

[esldude]

Well that's the second major flaw of quick A/B testing (sighted or blind), IMO - it is premised on the fact that the difference is actually a distortion that is amenable to quick A/B differences & doesn't allow for psychoacoustics.

 

Sorry, stop. Don't know where you come up with stuff. DBT is premised on whether you hear a difference. The difference could be anything. Nowhere is it premised upon distortion.

[PeterSt]

Nowhere is it premised upon distortion.

 

Maybe not. But still it always *is* about that. Or do you think sound is "created" as such when something is new for the better ?

But maybe it is harder to understand the combination with Merill's psychoacoustics. So it (or he) is all about that. I have no problem thinking along the same lines.

[mmerrill99]

Sorry, stop. Don't know where you come up with stuff. DBT is premised on whether you hear a difference. The difference could be anything. Nowhere is it premised upon distortion.

A distortion meaning a change from the original signal.

[esldude]

A distortion meaning a change from the original signal.

 

That would be one use, but that is not the premise upon which DBT is based.

 

For instance, it was used long ago to determine how much louder one signal is before it is noticed by a human listener. There was no distortion in that case. Simply investigation into the fineness of level perception in humans.

 

It is premised on finding differences. Finding if one signal is heard as different or the same by listeners. If one drug treats a disease better than another drug. While it can be used on distortion that is not the premise or only use of it. Nor does the fact it can be used to probe distortion audibility mean it is a broken tool in the sense you implied.

[mmerrill99]

If you receive it as more loud (as you said) ? No, I think what I said applies in that case. But there is more;

 

I don't know who told about it (I think it was Paul), but while listening to music (sound regarding that), our visual system is highly active. Why ? because we always try to associate with familiar things from memory. All starts with being alert to danger, which btw is also why certain frequencies are more sensitive to us (think 1200 - 5KHz or so). The snap of a branch, human voice. So visuals sure are recorded in our memory. Now when things match, the foot tapping etc. occurs. We can associate.

 

I have told about it before - this "system" is very complex; when you listen to (and through) your gear, at first you will be able to be satisfied when hearing a cymbal (happy happy) without attack and thus actually without the real source of the sound. This, and this is the interesting part, while you surely know the cymbal from live. You can even see it happening (Cobham hitting the cymbal). Still how your system shows it is also fine. At first !

But wait for the day that the attacks on the cymbal come through as well, through your speakers. From that day on you can't be without. So if now that's taken out again, your brain will hurt and it will be very tiring to longer listen to that. It just ain't right. Not any more.

This is where the filters are crucial. So let them ring too long (compare to none previously) and you will hear it. That the absolute frequency/harmonics (which are 100s for a cymbal) isn't right ... so what. You wouldn't know better anyway. But if something which is to be sharply boundaried while it is not, there's a mismatch with ... well, some kind of combination of the visual memory and what we can synthesize from the elements of sound. At least that is how I see it.

 

Your next problem when you have this all quite right, is that the live concerts start to suck, because no way good sound comes from there (usually). Still it is as live as can be, including the visuals. But then of course someone thought to lead all through microphones, or otherwise the hall has too poor acoustics etc.

 

So in the end I agree with you. But it depends on what exactly is the change and how it is perceived. Increase in dynamics is usually allowed, *if* a better fundament goes along with it; without that sound gets lean and sharp and will show over-detail, sounding nasty (harsh) meanwhile (not always, the nasty, but too lean yes).

Yes again, I think I understand(?) you, Peter - I'm agreeing with your other post that it would seem the reason for better perceived dynamics have to do with the time domain aspect of the signal. The better the risetime of the signal, the more accurately the start of the sound is perceived - the better we hear the layering & separation of different instruments in the sound stage. Conversely, the more time smearing in the signal, the less accurately we hear the instrument onset, the less layering, solidity in the sound stage, the less dynamic are the perceived dynamics.

 

I'm not talking about added distortions which emphasises detail.

[PeterSt]

For instance, it was used long ago to determine how much louder one signal is before it is noticed by a human listener. There was no distortion in that case. Simply investigation into the fineness of level perception in humans.

 

As long as you realize that playing back music at around realistic levels also click-in. Not exactly related to what you say here, but true it is.

 

Deviation from reality is also a distortion.

[mmerrill99]

That would be one use, but that is not the premise upon which DBT is based.

 

For instance, it was used long ago to determine how much louder one signal is before it is noticed by a human listener. There was no distortion in that case. Simply investigation into the fineness of level perception in humans.

Yes but that is a different use to what is being discussed here - that is threshold DBTs for establishing JNDs

 

It is premised on finding differences. Finding if one signal is heard as different or the same by listeners. If one drug treats a disease better than another drug. While it can be used on distortion that is not the premise or only use of it. Nor does the fact it can be used to probe distortion audibility mean it is a broken tool in the sense you implied.

So, can I ask this - do you suggest that a DBT would expose differences to anybody (of normal hearing) or do you think that knowing what to listen for, where to listen will have a significant effect on the outcome? In other words, is prior training in identifying distortions of consequence or not?

[PeterSt]

I'm not talking about added distortions which emphasises detail.

 

... which is the biggest pitfall and even hard to recognize (hello NOS users). At first. But this is with most of the distortions; they "add" like a photo may look more sharp when noise is added.

[esldude]

Yes but that is a different use to what is being discussed here - that is threshold DBTs for establishing JNDs

 

 

So, can I ask this - do you suggest that a DBT would expose differences to anybody (of normal hearing) or do you think that knowing what to listen for, where to listen will have a significant effect on the outcome? In other words, is prior training in identifying distortions of consequence or not?

 

No it has been reliably shown, that for some artifacts or differences training improves results. For others like level differences it seems to not matter. Where the training makes a difference is only at the very margins of perceptibility. One should be careful not mix up pure bulk experience with training. What I mean by this is often audiophiles or pro audio people will think of how long they have listened carefully, and how many years, and assume they are something of a trained listener. Sorry, nope, not hardly. Mainly because their perceptions were without any verifaction of whehter or not their perceptions were in fact correct.

 

You could do the Philips Golden Ear challenge training course.

 

https://www.goldenears.philips.com/en/introduction.html

 

In forums some have reported getting to the top level. Some report they don't quite get there. I don't recall reading anyone saying they just breezed right through it from all of their years of experience as an audiophile.

 

I believe Harman has something similar online.

[mmerrill99]

No it has been reliably shown, that for some artifacts or differences training improves results. For others like level differences it seems to not matter. Where the training makes a difference is only at the very margins of perceptibility. One should be careful not mix up pure bulk experience with training. What I mean by this is often audiophiles or pro audio people will think of how long they have listened carefully, and how many years, and assume they are something of a trained listener. Sorry, nope, not hardly. Mainly because their perceptions were without any verifaction of whehter or not their perceptions were in fact correct.You could do the Philips Golden Ear challenge training course.

 

https://www.goldenears.philips.com/en/introduction.html

 

In forums some have reported getting to the top level. Some report they don't quite get there. I don't recall reading anyone saying they just breezed right through it from all of their years of experience as an audiophile.

 

I believe Harman has something similar online.

Indeed & Harmon show that training makes a huge difference to perception of speaker differences - hardly the margins of perceptibility.

 

What their graph shows is that even for gross differences found in speakers, training makes a huge difference. Hardly at the margins of perceptibility.

 

The conclusion being that untrained listeners miss quiet a lot of these gross differences because they don't know what to listen for & where to look.

 

Your blanket statement that DBTs are about "Finding if one signal is heard as different or the same by listeners." & your subsequent denial that training is of huge importance (& relevant "only at the very margins of perceptibility") is the very reason why DBTs are the wrong tool for the hobbyist & it's results usually very dubious.

 

It's not the tool that is dubious, it's how it's used by hobbyists & the results thereof.

[PeterSt]

Training is as fast as time permits to point out things (or differences for that matter). It is true that it is hard to meet people who can do it as easily as I myself, though they do exist. But after something is pointed out one time (only) they can do it themselves. It really is super easy. And notice : always for next tracks/albums. Never the same (attempted, which would be a daft thing).

[mmerrill99]

Training is as fast as time permits to point out things (or differences for that matter). It is true that it is hard to meet people who can do it as easily as I myself, though they do exist. But after something is pointed out one time (only) they can do it themselves. It really is super easy. And notice : always for next tracks/albums. Never the same (attempted, which would be a daft thing).

Yes, once learned (or identified by someone else) it is usually internalised & becomes a reference by which one can identify the same distortion/artefact again.

 

But (& here's my BIG point) before this learning people don't readily recognise a distortion/artefact - it remains invisible to them. Sit them down in a blind A/B test & they will almost definitely not hear a difference.

 

Let them listen over the long term to music with this artefact in it & music without this artefact in it & they have a much better chance of actually picking up on the perceptual differences between them & zoning in on the specific impairment.

[esldude]

Indeed & Harmon show that training makes a huge difference to perception of speaker differences - hardly the margins of perceptibility.

[ATTACH=CONFIG]16627[/ATTACH]

 

What their graph shows is that even for gross differences found in speakers, training makes a huge difference. Hardly at the margins of perceptibility.

 

The conclusion being that untrained listeners miss quiet a lot of these gross differences because they don't know what to listen for & where to look.

 

Your blanket statement that DBTs are about "Finding if one signal is heard as different or the same by listeners." & your subsequent denial that training is of huge importance (& relevant "only at the very margins of perceptibility") is the very reason why DBTs are the wrong tool for the hobbyist & it's results usually very dubious.

 

It's not the tool that is dubious, it's how it's used by hobbyists & the results thereof.

 

Okay, has been awhile since I looked at the Harman article. The difference was larger than I remembered. Quite large.

 

I would also note two things.

 

One is though there is more variance between trained and untrained, while the difference between good and poor performance of speakers under test was rated larger by trained listeners, in every case the untrained groups came to the same conclusions and rankings as the trained listeners.

 

The other is that when the quality of the speaker being compared is higher there is less difference between trained and untrained than when the quality is lower.

 

So now all you have to do to convince me is show me how you can know that untrained listeners using DBT to confirm differences is worse than sighted uncontrolled listening. As for hobbiests needing to do a better job when they use DBT testing I would not disagree with. The answer isn't to throw it out completely. Merely to spread the word on what good procedure is.

[mmerrill99]

Okay, has been awhile since I looked at the Harman article. The difference was larger than I remembered. Quite large.

 

I would also note two things.

 

One is though there is more variance between trained and untrained, while the difference between good and poor performance of speakers under test was rated larger by trained listeners, in every case the untrained groups came to the same conclusions and rankings as the trained listeners.

 

The other is that when the quality of the speaker being compared is higher there is less difference between trained and untrained than when the quality is lower.

 

So now all you have to do to convince me is show me how you can know that untrained listeners using DBT to confirm differences is worse than sighted uncontrolled listening.

Well I've laid out my logical arguments & that's as far as it goes - as you know, I don't believe anybody has done the research or tests that you are suggesting is needed to convince you ( a safe demand )

As for hobbiests needing to do a better job when they use DBT testing I would not disagree with. The answer isn't to throw it out completely. Merely to spread the word on what good procedure is.

No, this won't work - the tool is too complicated to use reliably for the hobbiest - it doesn't matter how often or far the word is spread - it's not meant for the usage to which it is subjected to by hobbiests - incorrect results are the outcome & arguments constructed upon this shaky ground.