Two problems:
1. If you believe we should try to use correct terminology, and IIUC you do, then the term 'proof' is defined in mathematics, and in law, but IIUC not so much in science. Again IIUC, in science there is only 'evidence' which can serve to reduce doubt as to the truth of a proposition or hypothesis. That type of thing.
2. Intuition that two protocols superficially seem more or less equivalent to someone doesn't mean the scientific evidence indicating otherwise can safely be ignored. Its probably that under experimental conditions humans don't turn out to respond to different protocols as common intuition may suggest.
Serious persons in the audio field well know that there are audible differences among equipment of a given type.
This has even been mentioned here in this thread. They have long and profound experience, and don't have
to "prove" this to anyone to progress in their work.
Instead of such diversionary tactics, a constructive person would contribute to progress in the audio field.
Many of us have been doing just that for quite a number of years, and we will continue, and we will not be censored.
I recall reading about Hans' work in the Zip Cord for Speaker Test thread. He took measurements, made calculations, implemented hardware changes, and reported listening results in terms of A/B comparison with other cable. I don't hold it against him that he didn't do ABX DBT.
Also regarding ABX DBT, IIRC correctly a certain forum member raised a great stink about the sound of two wave files with the same four sine waves in them, with two of them phase shifted in one case. I asked the person if he confirmed his subjective listening impression using ABX DBT. He responded that he didn't need to because the difference was obvious to him. IIRC the member nick started with the letter 'b' and ended with some numbers.
The files are at: https://purifi-audio.com/2019/12/07/amfm/
My point here is that there seems to be a double standard: If Hans hears an obvious difference then he is obligated to provide proof or he should be censored. If the other guy hears an obvious difference then he gets a free pass. Isn't there is a word for that, is it hypocrisy?
Also regarding ABX DBT, IIRC correctly a certain forum member raised a great stink about the sound of two wave files with the same four sine waves in them, with two of them phase shifted in one case. I asked the person if he confirmed his subjective listening impression using ABX DBT. He responded that he didn't need to because the difference was obvious to him. IIRC the member nick started with the letter 'b' and ended with some numbers.
The files are at: https://purifi-audio.com/2019/12/07/amfm/
My point here is that there seems to be a double standard: If Hans hears an obvious difference then he is obligated to provide proof or he should be censored. If the other guy hears an obvious difference then he gets a free pass. Isn't there is a word for that, is it hypocrisy?
But Hans,
It is all about OTHER people being unhappy with YOUR gear, didn't you get the memo? lolol Sorry I couldn't resist.
I agree about the listening test methodology issue, for better or worse I have been dragged into hundreds of A/B, ABX/ Random long selection repeats, etc. and in my experience Mark is right, they show excessive negative results, due to lack of training in hearing the difference, as well as the mental confusion of having things switch up. They do show decent correlation for a large number of people for gross differences. Also, IMHO none of these tests do much to eliminate pre-existing biases towards certain SQ contours (the "I'm used to my own speakers" thing).
The only way I have been able to get reliable results and consensus is by weeks of training with specific exaggerated SQ problems (missing bits, odd noise floor contour, mechanical noises on magnetic recordings, etc.) and then reducing them to near-inaudibility. Done this way people train their brains to identify specific sounds. In a similar way we hired and trained pre-press graphics people by placing cards in front of them with small color variations and asked them to pick out the outlier. Some show an ability immediately, but many can learn by repeated testing, resulting in a crew of artists who could pick out microscopic color variations I personally could not see.
But yeah, let's not go down the testing rabbit-hole please!
Cheers,
Howie
That's considered the correct way of doing it. IIRC the term for that process is called 'graded learning.' Don't remember the source of it though.
Its also very important that feedback as to right or wrong answers be available immediately, rather than having to wait. One problem I had with Foobar2000ABX was that I didn't find out if an answer was right or wrong until I finished a sequence of questions. For training purposes that's bad since by the time I finished the sequence I couldn't remember exactly what was my perception was when I answered the first question. Thus I learned nothing from most of my mistakes.
https://link.springer.com/content/pdf/10.3758/MC.37.8.1077.pdf
Last edited:
People buying high end gear dish out serious moolah for the looks and bragging rights, not the performance. That’s is the truth of it. Otherwise, how do you explain a 10 W 2% distortion $9k Audio Note amp?
Are you talking about some subset of people, or generalizing from a few people to all people who have more expensive systems that you do?
BTW, they took down your excellent 360-degree review graphic. Canceled it, basically.
My mom, when she was alive, used to send money to countless charities every month. This included the barely clothed children of third world countries in those pictures they used to send around. Can you imagine one of these children being duped into buying a dCS Scarlatti in lieu of feeding his sister?
On an aside, I just heard stupendous (please excuse the superlatives) differences in my latest phono stage. I can agree that there hasn't been enough studies done on the impact of transients, those signals being defined as not DC.
By "stupendous" I mean "not remotely as bad as it used to be"
On an aside, I just heard stupendous (please excuse the superlatives) differences in my latest phono stage. I can agree that there hasn't been enough studies done on the impact of transients, those signals being defined as not DC.
By "stupendous" I mean "not remotely as bad as it used to be"
Last edited:
Serious persons in any scientific field prove that there are differences before starting to discuss what is causing the differences.
I don't know who you are referring to but I have never made such claim. You should be very careful to not misquote members unless you want to get banned.
I’m talking about people with loads of money that spend it on audio esoterica - just look at some of the stuff on Stereophile retailing at 50k+ 😊
360 degree graphic?
This isn't a scientific field of contributors, nor one restricted to those of serious persons in those fields. Unsubstantiated opinions or beliefs may lay the groundwork for research by those individuals that you are referring, to set out to prove that the earth isn't flat, or that current measurements are not necessarily adequate to describe differences that might yet be proven. This can inspire serious persons.
It seems the fundamental principle you are suggesting is that if differences cannot be proven, discussions ought not to commence. The question is why not? Each individual can weigh the value of presentations in relation to their knowledge, experiences and beliefs in order to advance those things. Such restrictions you suggest imply that you can prove that discussions can, at minimum, cause more harm than good, perhaps when a child purchases a dCS Scarlatti in lieu of spending money for food for his sister. The point being is that not all of us want to present ourselves as saviours of mankind, want to be saved, nor necessarily need saving from what could be our delusions. To what extent can harm be proven to be done?
To be clear... I never really liked my sister that much anyway....
Last edited:
I'm not suggesting that. Hans Polak suggested that we should stop talking about DBT, ABX or A/B. Why should that be censored but not the discussion about unproven audible differences? Discussions of unproven audible differences will inevitably lead to requests of proofs. I'm not sure that can be done without DBT, ABX or A/B. I suggested an easier alternative but that did not receive any response.
And may I remind you that the discussion started after reports of easy-to-hear differences (along with positive & negative comments) between very well measuring dacs (and maybe one not so well measuring). This has not been about audible differences in general. IMO it is obvious they exist (e.g. tube vs. solid-state amps).
And may I remind you that the discussion started after reports of easy-to-hear differences (along with positive & negative comments) between very well measuring dacs (and maybe one not so well measuring). This has not been about audible differences in general. IMO it is obvious they exist (e.g. tube vs. solid-state amps).
Last edited:
I have numerous DAC's with specs tested by ASR that appear beyond reproach. There seems nothing that would remotely suggest there would be any audible difference between them. If one sets an actual output level at say 90 or 100dB SPL the harmonics or intermodulation artifacts start to appear at minimum far below the threshold of human hearing, as would only be discernible at 0dB SPL if in a silent anechoic room and not the 90 or 100dB it is buried. Yet these DAC's vary in sonic signature, some more so than others. This is to state that I don't at all believe the specs are revealing such reality that I am perceiving to exist.
Ultimately it appears that modern DAC's have specs that scientifically can be considered beyond reproach, that to challenge their audibility as variant would be highly suspect within the scientific community. This is to suggest that no DBT, ABX or A/B testing isn't going to cause an uproar within the scientific community under such circumstances. Conclusions suggesting differences will be invariably condemned unless something else can be pointed to as the cause it. In agreement with Markw4 there must be more to it. My guess is time domain phenomenon not being revealed in frequency spectrums.
Ultimately it appears that modern DAC's have specs that scientifically can be considered beyond reproach, that to challenge their audibility as variant would be highly suspect within the scientific community. This is to suggest that no DBT, ABX or A/B testing isn't going to cause an uproar within the scientific community under such circumstances. Conclusions suggesting differences will be invariably condemned unless something else can be pointed to as the cause it. In agreement with Markw4 there must be more to it. My guess is time domain phenomenon not being revealed in frequency spectrums.
Last edited:
Then why not just record the DACs having different sonic signatures with an ADC as I suggested? I.e. loopback from dac to adc. Why would the sonic difference not show up in the recording when compared to the original? Of course the ADC needs to be sufficiently transparent but such are nowadays available.
Sometimes I get the impression that the Do-It-Yourself_Audio forum is more a Let's-Talk-About-Audio forum without the important let's do it aspect.
I have coached two blind tests on DiyA,
1) To test the importance of filter length in upsampling filters, a conversion of six totally different 44.1/1 recordings to 192/24 with several filter lengths of up to 16Mio coefficients were made, without telling which filter length was used and to compare them to the original 44.1/16.
2) To test whether people can really hear the ultrasonics in 194/24 and the added bits below 16 bits, the conversion of 192/24 recordings with lots of ultrasonics were made into a 20Khz brick wall filtered and 16 bit limited 192/24 versions where above 20Khz and below 16 bits random noise was added without telling what I did.
With other postings I tried to motivate people to repeat my efforts and give their opinion.
3) In the ever ongoing discussion whether cables can make a difference in sound perception, I showed the conversion of a Zip cord into a version costing less than $25,- that was an exact impedance match of an expensive MIT LS cable, where the conversion to my subjective opinion was a clear improvement
4) To get a better insight whether Voltage or Current amplification for an MC Cart's has any advantages, I posted a Universal Head Amp that could be used as voltage amp or transconductance amp with exactly the same components, making it possible for the first time to do a real compare between the two different topologies instead of comparing a voltage preamp from make A to a completely different current amp from make B. With my gear and to my very subjective opinion, the transconductance was the winner.
Why telling all this, just because when starting these threads, lots of people participated, but when it was time to do the testing, many people sadly seemed to avoid to burn their fingers, making the outcome of these simple tests 1) and 2) far less significant.
And although 1) showed no outcome in favor of 192/24 and 2) showed a real positive effect for "the longer the better" but surprisingly much against my expectation that 44.1/16 played in NOS without output filter and with all its ultrasonics was perceived by some to be on par or even slightly better as the 192/24 with the longest filter length.
But again, both test performed by less then 10 participants, so in fact not enough to draw any conclusion.
Test 3) with the cable was only performed by one or two DiyA members, although in ongoing cable discussions tenths of people are participating, most of them telling there is no such thing as sound difference between cables, so investing in $25,- would be a waste of money. A large prejudice IMO.
Point 4) was probably to ambitious. Several people started to building the head amp, but only in its transconductance version, so it never came to comparing both topologies from others.
Hans
I have coached two blind tests on DiyA,
1) To test the importance of filter length in upsampling filters, a conversion of six totally different 44.1/1 recordings to 192/24 with several filter lengths of up to 16Mio coefficients were made, without telling which filter length was used and to compare them to the original 44.1/16.
2) To test whether people can really hear the ultrasonics in 194/24 and the added bits below 16 bits, the conversion of 192/24 recordings with lots of ultrasonics were made into a 20Khz brick wall filtered and 16 bit limited 192/24 versions where above 20Khz and below 16 bits random noise was added without telling what I did.
With other postings I tried to motivate people to repeat my efforts and give their opinion.
3) In the ever ongoing discussion whether cables can make a difference in sound perception, I showed the conversion of a Zip cord into a version costing less than $25,- that was an exact impedance match of an expensive MIT LS cable, where the conversion to my subjective opinion was a clear improvement
4) To get a better insight whether Voltage or Current amplification for an MC Cart's has any advantages, I posted a Universal Head Amp that could be used as voltage amp or transconductance amp with exactly the same components, making it possible for the first time to do a real compare between the two different topologies instead of comparing a voltage preamp from make A to a completely different current amp from make B. With my gear and to my very subjective opinion, the transconductance was the winner.
Why telling all this, just because when starting these threads, lots of people participated, but when it was time to do the testing, many people sadly seemed to avoid to burn their fingers, making the outcome of these simple tests 1) and 2) far less significant.
And although 1) showed no outcome in favor of 192/24 and 2) showed a real positive effect for "the longer the better" but surprisingly much against my expectation that 44.1/16 played in NOS without output filter and with all its ultrasonics was perceived by some to be on par or even slightly better as the 192/24 with the longest filter length.
But again, both test performed by less then 10 participants, so in fact not enough to draw any conclusion.
Test 3) with the cable was only performed by one or two DiyA members, although in ongoing cable discussions tenths of people are participating, most of them telling there is no such thing as sound difference between cables, so investing in $25,- would be a waste of money. A large prejudice IMO.
Point 4) was probably to ambitious. Several people started to building the head amp, but only in its transconductance version, so it never came to comparing both topologies from others.
Hans