OK I found an ABX tool for linux and have just done a test. Not a perfect score but 7/10
The difference I am picking up on is what I will call richness or fullness in the sound. I first listened to A and B in their entireity and for the ABX'ing I only listened to roughly the first 10 seconds.
One to me sounds like it has more depth, and perhaps slightly more reverb/decay, the other sounds slightly flatter. That's the best I can do with a subjective description. I think 7 out of 10 is probably good enough to try the sorting part of the test. I could not pick a difference with foobar on my cheap computer headphones, looks like the main system is doing a better job
Note that I hid the results whilst doing the actual test. I had a peek at 7 (at which point I had 5 right) and then hid again until I had completed 10 iterations. I think that not knowing whether I am getting them right or wrong until the end is probably for me important.
Tony.
The difference I am picking up on is what I will call richness or fullness in the sound. I first listened to A and B in their entireity and for the ABX'ing I only listened to roughly the first 10 seconds.
One to me sounds like it has more depth, and perhaps slightly more reverb/decay, the other sounds slightly flatter. That's the best I can do with a subjective description. I think 7 out of 10 is probably good enough to try the sorting part of the test. I could not pick a difference with foobar on my cheap computer headphones, looks like the main system is doing a better job
Note that I hid the results whilst doing the actual test. I had a peek at 7 (at which point I had 5 right) and then hid again until I had completed 10 iterations. I think that not knowing whether I am getting them right or wrong until the end is probably for me important.
Tony.
Attachments
Great result Tony... and you know whats coming next , you gotta match 'em
It seems like everyone latches on to different aspects of the "sound" to find the differences, be it the overall "wash" of the sound or perhaps emphasising and concentrating on transients or hf detail.
Its the result that counts though.
, you gotta match 'emIt seems like everyone latches on to different aspects of the "sound" to find the differences, be it the overall "wash" of the sound or perhaps emphasising and concentrating on transients or hf detail.
Its the result that counts though.
Me and Frank have high confident in this test. Accuracy just about time/effort. For example, if I can do Foobar ABX 10/10 or even 6/6 it is a guarantee (for me not for others) that the difference is there.
What I have learned from these DBTs is that it is very easy to perceive differences. Choosing which is which is the difficult part. For example, it is very easy to perceive that one file is natural and the other is unnatural (for example). What is difficult is to remember how a natural should sound and how an unnatural should sound. This makes ABX result can be imperfect. But the AMAZING thing is, when the difference is not there, I really cannot pick differences!
SY, I joined this test because this is the only one that you think is better controlled. Next time I expect a more "useful" test such as COPPER versus SILVER, CFA versus VFA, Litz versus solid core, air core versus iron core,... Any issue out there that is hot or needs proof. Or another tube versus solid state to invalidate the Stereophile test
No, that's not what I mean. I need to repeat this as I think this is important (at least this is something new for me):
When we think there is differences between two files, even if we failed in identification of which is which (ABX), it is a good sign that the differences do exist. This is because perceiving differences is quite an accurate process. What inaccurate (difficult) is the process to remember the characteristics of the sound (which is which).
I have experienced this in previous ABX tests. For example, I tried to listen which one is natural from the singer's inhale in two files. Up to 5/5 I can remember which inhale is natural and which is not natural but for the next round I started to confuse which one was supposed to be the natural one. And it is easy to make mistake in this process.
I have my sorting done only in a few minutes, just playing each file in a minute and I could see if the file should belong to which group. But I wanted to test it with ABX. Originally I wanted to ABX each file with A and B. This is where I know that when the file is the same, I couldn't pick differences. But I tried to assume that the file could be somewhere between A and B so I need to find the closest character if any. I really didn't know what to pick because I couldn't fool myself that there was differences. But I picked just to see the result, which is way below my result when there is difference.
Another surprise (actually kgrlee already know this) is that, so far, what I can do in may be an hour to get 7/7 is actually no difference than what I can do in a minute when I can perceive the difference. From 12 files, only number 7 and 8 I couldn't decide from my quick sorting.
So, with such variability in the pb/rec system, which of the 14 files are actually the most representative of A and B ?.
Karl, I get the point that the 'signatures' of film vs electro should/ought/might shine through despite the system/connectors/pb/rec variations that you have inadvertently included.
On first inspection, it would seem that the system variation would serve to make differentiation of the film vs electro more difficult (less reliable).
However without knowledge of the nature of the system caused variation, the above observation is not certain, and the system caused masking may or may not be deleterious to the experiment.
Indeed, it seems that Frank and Jay have been able to 'hear through' the system variations and discriminate the two capacitor sounds.
I will endeavour to download the 12 cap sounds and the 12 opamp sounds, and sort them in the next few days.
No time now, I am off in a few hours to do load in and production on a local big outdoor festival....Pearl Jam and Snoop Dogg (aka Snoop Lion
) are headline acts...another couple ticked off the bucket list .As a last comment, the cap sounds are not all that distinct until one zooms into fine details like the 5/6 note piano trills (I used this detail also).
The X/Y opamp sonic differences are much more obvious.
Dan.
That is what we want to hear
Yes I have done several files. Not much time I can use because of I have guest (my mother) so I cannot play the music at night (headphone is fatiguing and thus decrease my concentration). Will send when I complete the ABX or when the time due.
BTW, the opamp test can be useful if the quality is better, at least equal to previous tests. In this test I couldn't pick up my preference because both files have their own weaknesses (and I have never accepted any of such weaknesses). Did you still use the Marantz SACD player or a less good equipment?
So, with such variability in the pb/rec system, which of the 14 files are actually the most representative of A and B ?.
Karl, I get the point that the 'signatures' of film vs electro should/ought/might shine through despite the system/connectors/pb/rec variations that you have inadvertently included.
On first inspection, it would seem that the system variation would serve to make differentiation of the film vs electro more difficult (less reliable).
However without knowledge of the nature of the system caused variation, the above observation is not certain, and the system caused masking may or may not be deleterious to the experiment.
Indeed, it seems that Frank and Jay have been able to 'hear through' the system variations and discriminate the two capacitor sounds.
I will endeavour to download the 12 cap sounds and the 12 opamp sounds, and sort them in the next few days.
No time now, I am off in a few hours to do load in and production on a local big outdoor festival....Pearl Jam and Snoop Dogg (aka Snoop Lion
As a last comment, the cap sounds are not all that distinct until one zooms into fine details like the 5/6 note piano trills (I used this detail also).
The X/Y opamp sonic differences are much more obvious.
Dan.
That would be great Dan and I'd be particularly interested in your impressions of the opamp test. If you can detect a difference on that one... which do you prefer if you had to have one or the other ? That's personal preference, no right and wrong answer to that one. Answers on a postcard (or a pm will do
) You mentioning that the differences stand out on the opamp files I think shows the tests have real validity. That test is where bigger differences should appear, and it seems maybe they do... perhaps adding more weight to the validity of the cap tests. The way you all pick the differences out is interesting. With the earlier tests some weeks back I found just listening to perhaps 4 or 5 seconds of the same part of the track could clinch it for me... but they were just ABX tests, not sorting tests as this one is.
I'm just trying to collate the results that are in and set them out so that it looks good and can be easily assimilated. Quite a job but worth it I hope.
Enjoy the gig...
Sorry Karl, I had to comment...
How does natural inhaling sound and who was inhaling when you first formed the memory sample? Was it when your girlfriend gasped for air, while sleeping, taking a good breath to get the bowels moving while sitting on the can. Could you tell the difference between a man and a woman breathing or between your girlfriend, your wife or your mom?
This is creepy stuff!
Last edited:
Sorry Karl, I had to comment...
How does natural inhaling sound and who was inhaling when you first formed the memory sample? Was it when your girlfriend gasped for air, while sleeping, taking a good breath to get the bowels moving while sitting on the can. Could you tell the difference between a man and a woman breathing or between your girlfriend, your wife or your mom?
Be careful with your comments Nico... There are many intelligence people here. You risk yourself to look stupid (because your comments may show your not understanding of something and/or your lack of intelligence). Of course it doesn't matter if you (like me) don't care or never tried to look smart like some others
Just in case it matters to you... 
I am flabbergasted, you are capable of identifying which hand the player used by listening alone, but you cannot say whether it is five or six notes!
This is absolutely stunning stuff, can you tell what the musician was thinking while he was playing?
Are you using expensive audiophile equipment?
Nico - it can be regarded as common knowledge that higher notes are played with the right hand and lower with the left hand on a piano.... - no? Have you listened to the track?
Compare and listen as much as you like and in whatever orders you like
I would suggest using the two reference files to begin with, and then, if and when you think you can hear a difference, try and see if you can match the others in sound to either reference.
Thanks for giving these a listen too... the more results the better.
Sorry I am from Africa, we hit tree trunks with clubs while dancing around the fire and it does not matter which hand you use, really.
Anyway if it is so obvious then why mention it as if it is a characteristic of the sound. These two recordings are amazing though Mooly. The file sizes are exactly the same to the bit. But we know they are different because they have different CRC.
I tried the same this afternoon just to prove a bit of conversion theory to myself. I recorded the same streamed audio twice without changing a single setting and low and behold these two files did not look the same (as pavel found errors) nor present the same CRC simply because there is no ADC in this world that can convert streamed analog into digital with zero errors.
So besides the two files A & B there also exist conversion errors A' and B' which could contribute as much difference as the capacitors in them sounding differently.
Pavel, Sy, Frank, Karl what have you to say about conversion errors causing differences in perception.
Hello Nico,
these errors depend on DAC and ADC used, which is probably no surprise. We have to spend as much effort as possible to minimize such errors. We can do it by a choice of converters used and also by using 24-bit conversion rather than 16 bit or 14 bit conversion.
Last - these errors are measurable.
Regards,
Pavel
I haven't seen evidence that there actually is a difference in perception. Lots of creativity, though.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.