Before we go someone mentioned about Santa being an audiophile, can I say I wish he was I am fed up with sox.
I am always interested in how much attention the little things attract. No disrespect meant to any of the participants.
Hifi reproduction is real life. Soundwaves are analog.Alan Hope said:
I built my T-amplifier in 2005 using the EB-TA2020 evaluation board from Tripath from which many DIY T-amps were derived. T-amps have switching frequency on the output signal. One doesn't hear the switching frequency but one can measure it. The distortion you mention is normal and seems not related to the switching frequency but to PSU voltage. I limit the output of my t-amp to about 8 Watts, where distortion is at its lowest.
Been there, done that. What conditions are you thinking of?
fredex said:
Hey, I've explained this "large" differences before, there is a large 😀 difference between a subjectivist and objectivist "large".
The one are talking about detail, the other talk about making a guitar sound like a piano. 😀 😀 😀
I may not be understanding you. The stimulus I was thinking of is the sound field in the ABX-DBTing room. What do we have to adapt to for the purpose of the tests ?jneutron said:
ravon said:
Ravon, I mostly use well recorded acoustic guitar and drumset with cymbals and bells type of stuff.
I will leave the cables for you, see what you can get. Wish I could organise you something from Ecosse in Scotland.
I suggest something like two recordings (same track) from 3 different cables and see if I can match them. Make sure your you have a CORRECT record of which cables are used for each track. 😀 😀 😀
André
Andre Visser said:
They are two separate things connected by a cable... kind of like a mic connected to a dsp unit that varies its dsp settings depending on what it picks up through a camera (it's a "subjectivist dsp unit. 😀) ... and the blind person's experience is like having the camera shut off, which frees up some of the dsp's memory to read the mic input at higher sampling speed and resolution.
tnargs said:
Yip, something like that, that's the reason why I listen with closed eyes to free up resources in my asp. 😀
nickmckinney said:
Yah, I guess the lamination NRE is a killer. I wonder if there is much savings to have standard lamination plates made and machined later to size..
nickmckinney said:
Ah.. In my apps, 2% is unacceptable. Last project, the vendor subbed a different melt halfway through, and we ended up putting 1 mil thick shims on 11 thousand pound magnets in order to equilize all the magnets.
nickmckinney said:
Well heck, ya put that copper sleeve in, whaddya expect??
Nice design.
Cheers, John
KSTR said:
Agreed. That is why it is useless to use mic's within the ears, or even a pair of mikes at ear spacing. Considering only first arrivals (all four*), it takes a heck of a lot of processing to try to duplicate what the brain does naturally..
First arrivals* for a mono signal..
1. left reaches left, right reaches right..image is centered. this is headphone as well as listening with a septum to prevent crosstalk.
2. Left reaches right, right reaches left..this is enough information for each ear to image the speakers directly (two), as well as a direction inverted imaging. Luckily, we tend to ignore the last three images, but I certainly agree they are confounders.
My test uses the human "processer" to really duplicate as best as possible, the placement of an image. The goal in stereo is to pretend the speakers aren't there at all, so one should try to achieve that, no?
KSTR said:
I didn't mean it in a bad way. When I said "fail", I mean that the recording engineer has not been given the tools to work with that eliminate the need for special talents. So I consider it a failure of engineering...it's just my way of thinking...
AJinFLA said:
If you consider that humans adjust to stimulus changes in a user transparent fashion, then how would one compare subtle changes in the system...if the measurement tool adjusts without telling us?
If one really is interested in comparing cables, human adjustment capability must certainly be considered, not ignored.
Tinkering is entirely acceptable. But in itself, does not advance SOTA. When tinkering produces good results, the reasons MUST be investigated, analyzed, and reported. This was in essence, the "beef" I had with John Curl. However, I can certainly understand his attitude, as he makes a living in audio, and the better his edge, the better off he is..I cannot blame him..
fredex said:
The purpose of the abx/dbt testing is to find small changes using humans as the test instrument. Since we adapt to alterations in stimulus (this case being localization parameters), we must consider our adaptation..
Using a synthetic, not found in nature stimulus for audibility is not a good thing..
Cheers, John
@ jneutron,
One could argue, that our brain is detecting time differences between the arivals at both ears, and so, it there is no time difference it draws the conclusion that the sound source must be in front of us.
Not so much processing power needed so far it seems at a first glance.
But our brain is able to do the same analyzing routine for numerous sound sources at the same time and for threedimensional spaces.
So if we consider several instruments located in front of us but next to each other spanning for example 120 degress we are able to pinpoint to each source and if we add an instrument distance right in front of us our brain is still able to seperate this from the others and to tell us it´s right in front of us, but at a greater distance.
And still listening in a reverberand space does enhance our ability to do so.
Regarding additional test setups, suggested by you and KSTR, i´d have some doubts at least in the beginning.
As we are holistic systems it is sometimes difficult to predict what our reactions to alterations will be.
And the transferability of results to considerations about `normal listening`would become problematic.
(The old rule of testing- the more controlled the less the practicability of the results)
As the claim in our cable discussion is " it´s possible to hear differences after exchanging (for example) the interconnects" (all safeguarding included i.e. measurement, differences, hearing thresholds and so on) in a given setup, the first step has to be succesfull listening tests in _this_ setup.
Everything else would be helpful in better analyzing and understanding of the underlying working principles later on, if the audibility is proved.
Just for the discussion in the hifidelity reproduction scheme, in general the suggested experiments would be very interesting.
One could argue, that our brain is detecting time differences between the arivals at both ears, and so, it there is no time difference it draws the conclusion that the sound source must be in front of us.
Not so much processing power needed so far it seems at a first glance.
But our brain is able to do the same analyzing routine for numerous sound sources at the same time and for threedimensional spaces.
So if we consider several instruments located in front of us but next to each other spanning for example 120 degress we are able to pinpoint to each source and if we add an instrument distance right in front of us our brain is still able to seperate this from the others and to tell us it´s right in front of us, but at a greater distance.
And still listening in a reverberand space does enhance our ability to do so.
Regarding additional test setups, suggested by you and KSTR, i´d have some doubts at least in the beginning.
As we are holistic systems it is sometimes difficult to predict what our reactions to alterations will be.
And the transferability of results to considerations about `normal listening`would become problematic.
(The old rule of testing- the more controlled the less the practicability of the results)
As the claim in our cable discussion is " it´s possible to hear differences after exchanging (for example) the interconnects" (all safeguarding included i.e. measurement, differences, hearing thresholds and so on) in a given setup, the first step has to be succesfull listening tests in _this_ setup.
Everything else would be helpful in better analyzing and understanding of the underlying working principles later on, if the audibility is proved.
Just for the discussion in the hifidelity reproduction scheme, in general the suggested experiments would be very interesting.
Jakob2 said:
Yes. For the first arrival, and the direction inverted, the virtual is center. The alternates due to crosstalk, seem to be ignored.
Jakob2 said:
No prob..
Jakob2 said:
Difficult to predict? You are a master of understatement.
Jakob2 said:
Agreed and agreed. You sound like my old statistics prof..
My points within the thread regarding audibility testing..
1. How do we account for human ability to adapt transparently to a subtle change in localization parameters.
2. Why do we use synthetic artificial, not seen in nature signals to test what humans can detect? No ITD..
3. Given purely synthetic stimulus to create an artificial source image, what level of changes are humans sensitive to specifically with respect to a real source. With the third speaker generating the real source, and two generating the artificial image, what artifacts within the artificial image system will alter the image w/r to the real one.
4. How tight do we have to control the artifacts of #3?
Jakob2 said:
I still point out the confounder......human adaptation.
Assume one has created a test signal which can fool the subject into believing the center speaker (offset some arbitrary angle) is creating the sound. If we then rapidly switch between the derived image to the real speaker, will we notice an adaptation time constant?
If we use simple panning to derive the virtual image, is adaptation the same time constant?
If we drop in some ITD, then pan that to get the virtual, is the adaptation the same, different, no change.
If we find the correct mix of ITD and IID to eliminate any adaptation, THEN we use that mix and the CUT (cables under test) to determine if the cables cause the systems to diverge.
Cheers, John
Andy Graddon said:
Yeah, I agree.. there are too many people making snide remarks.
I always assume when people do that, or start yelling, they are not thinking rationally, or objectively 😉
and..........
it reminds me too much of my Ex
Lock 'er up Cal !! I mean the thread .......
Ok Andy, we'll put you down as "Did No Measurements Whatsoever" (in your own system, in your own home) while listening and "Hearing differences with cables that measure the same". So now it's up to Andre, Rabbitz, AnalogSa, etc.
Gentlemen?? Your cable measurements please???
Lock the thread? Uncomfortable with the discussion Andy? How about not reading it and allowing those who wish to discuss to continue? Is anyone forcing you to listen to the blasphemy of the heretics?
cheers,
AJ
AJinFLA said:
AJ, I have not measured for LCR differences between cables, what's the use, I can't change it anyhow without making it shorter.
What are the chances of finding two different cables with the same LCR measurements anyway? Then wire diameter is also important, in the end every cable is a compromise.
André
jneutron said:
Thanks. My understanding of your position may have improved. I thought the purpose of abx-dbt testing is to establish the 'audibility' of small changes not to establish whether they exist or not. As you say we adapt transparently, so even if there is a change we might not hear it. So it would seem obvious you can't use human ears as a test instrument to establish whether a small change exists.
cheers
DOUBLE BLIND TEST RESULTS
**************************
DOUBLE BLIND TEST RESULTS
**************************
Ok guys, I've just done the DBT.
2 ICs used, 1 copper (VdH D102III), 1 silver (Glasshouse kit 1)
Randomisation by coin toss, test run by my 13yr old daughter across 3 rooms (one listening room, one waiting room for me, another for her) - ie we did not see each other at any point during the test.
Amp volume same throughout.
System as sig.
Music: studio recorded pop - Lenny Kravitz, listening for:
Overall clarity, freq balance, guitar harmonic structure, sibilance, percussion dynamics, mid-bass tone and evenness.
20 tests, same part of same two tracks each time.
I attempted to identify each cable.
RESULTS
Null hypothesis = 10/20
Test cables correctly identified = 5/20.
(On tests: 5,7,17,18,19)
So rather muddy water - I expected either a null or a positive result.
If analysed as identifying audible DIFFERENCES between cables then it was statistically significant (two tailed p=0.0414) ie this was not a chance result.
But if looked at as an audiophile identifying silver/copper differences with confidence then it was a failure. I thought I could hear the generally described differences attributed to silver (lean but extended bass, slightly raised treble, smooth but detailed highs).
I can't.
I consistently mis-identified the cables. Not randomly but wrongly.
Interesting - and irritating!
**************************
DOUBLE BLIND TEST RESULTS
**************************
Ok guys, I've just done the DBT.
2 ICs used, 1 copper (VdH D102III), 1 silver (Glasshouse kit 1)
Randomisation by coin toss, test run by my 13yr old daughter across 3 rooms (one listening room, one waiting room for me, another for her) - ie we did not see each other at any point during the test.
Amp volume same throughout.
System as sig.
Music: studio recorded pop - Lenny Kravitz, listening for:
Overall clarity, freq balance, guitar harmonic structure, sibilance, percussion dynamics, mid-bass tone and evenness.
20 tests, same part of same two tracks each time.
I attempted to identify each cable.
RESULTS
Null hypothesis = 10/20
Test cables correctly identified = 5/20.
(On tests: 5,7,17,18,19)
So rather muddy water - I expected either a null or a positive result.
If analysed as identifying audible DIFFERENCES between cables then it was statistically significant (two tailed p=0.0414) ie this was not a chance result.
But if looked at as an audiophile identifying silver/copper differences with confidence then it was a failure. I thought I could hear the generally described differences attributed to silver (lean but extended bass, slightly raised treble, smooth but detailed highs).
I can't.
I consistently mis-identified the cables. Not randomly but wrongly.
Interesting - and irritating!
- Status
- Not open for further replies.