I participated in one of the listening tests which was done with op-amps here a while back (maybe late last year). IIRC, it was a test of a high quality op amp in a non-inverting gain of 1 vs bypass test. THD plots showed no difference between the two modes.
I have moderate hearing loss (Sensory Neural loss from flying in the NAVY) and have a bad notch from 4K to 6K in both ears with the right well below -40dB wrt normal listening. With my hearing aids, I roll off before 10KHz.
Even with my hearing loss, in the hearing test I was able to have a confidence of greater than 80% that I was hearing the difference when running the test with foobar.
I don't like taking such tests because of the intense concentration it takes to make a sufficient number of passes for statistical validity.
In casual listening I would never hear the difference.
I have moderate hearing loss (Sensory Neural loss from flying in the NAVY) and have a bad notch from 4K to 6K in both ears with the right well below -40dB wrt normal listening. With my hearing aids, I roll off before 10KHz.
Even with my hearing loss, in the hearing test I was able to have a confidence of greater than 80% that I was hearing the difference when running the test with foobar.
I don't like taking such tests because of the intense concentration it takes to make a sufficient number of passes for statistical validity.
In casual listening I would never hear the difference.
I believe that you guys may hear some difference between good amplifiers, but then do not blame it on high-order distortion of up-to-date solid state amplifiers. It is an utter nonsense. There are hundreds of another, more likely reasons of sound difference, non-linear distortion being the last one.
Well, PMA has apparently solved all the world's power amp distortion problems, but I am still having trouble. For example, this is the distortion plot of a prototype power amp outputting 8V into 4 ohms. That is all it is trying to do! Note the higher order distortion compared to the lower order distortion.
Attachments
Last edited by a moderator:
I have no little doubt that even very small changes can be detected under certain conditions....type, level, harmonic, freq., etc. But it iS hard. It can take me all day going back and forth (with someone else doing the changing or not) listening to the same track of music to begin to detect a difference and even a pattern. But not expect to hear it in a quick A-b or Abx. get nothing. Such levels of distortions certainly are nit-picking and only of interest to perfectionist. Much has been made of it for years.... But it does appear that enough accumulation in a system has greater detectability..... like many coupling caps vs none.
My point is, we will not know what our limit of detectability is with distortion as high as they are in most speakers. Detectability which certain 'depends' on a number of factors... BUT, still the non-linear distortions of most any type is very high in speakers (included is compression). Cant we do better by now? There are a very few that are suitable in most ALL areas. Very few. With those you are more able to detect differences.
THx-RNMarsh
My point is, we will not know what our limit of detectability is with distortion as high as they are in most speakers. Detectability which certain 'depends' on a number of factors... BUT, still the non-linear distortions of most any type is very high in speakers (included is compression). Cant we do better by now? There are a very few that are suitable in most ALL areas. Very few. With those you are more able to detect differences.
THx-RNMarsh
Last edited:
If you have ever studied robotic vision you will find put quickly that using to cameras with not give you depth perception. In fact the third dimension is earned and processed through the brain. Example: babies knowing over their bottle as they try to grab for it.
So is the third dimension in hearing. It is learned. Example: throw a ball for a dog, if it lands behind him, he can hear it land but cannot tell where it landed, like humans can - even though they have vastly superior hearing.
Barring hearing damage, old age or brain inefficiencies, everyone is capable learning to hear at the highest level. Anyone with truly exceptional hearing would have to be in a higher class than 5%. More like the number of elite athletes in the US divided by the population.
So is the third dimension in hearing. It is learned. Example: throw a ball for a dog, if it lands behind him, he can hear it land but cannot tell where it landed, like humans can - even though they have vastly superior hearing.
Barring hearing damage, old age or brain inefficiencies, everyone is capable learning to hear at the highest level. Anyone with truly exceptional hearing would have to be in a higher class than 5%. More like the number of elite athletes in the US divided by the population.
Yes, but a prototype is just that - something that is to be tested, but not yet necessarily complete.
You can't compare that to something fully finished that you are happy to let out into the wild....
I just got some proto pcb antennas in today. One works well, the other doesn't quite cut it. But I needed to actually test the real thing, not a model, to be sure.
Graham
You can't compare that to something fully finished that you are happy to let out into the wild....
I just got some proto pcb antennas in today. One works well, the other doesn't quite cut it. But I needed to actually test the real thing, not a model, to be sure.
Graham
Gpauk is antenna guy, not hardware development. Different worlds, different rules. Physical layer vs. schematic.
Either the projects are very easy or very difficult (normally not so for medium difficulty), iterations are usually planned as mandatory, especially if RnD process is highly formalized sometimes at insane levels, effectively preventing Gpauk to be the release engineer.
Either the projects are very easy or very difficult (normally not so for medium difficulty), iterations are usually planned as mandatory, especially if RnD process is highly formalized sometimes at insane levels, effectively preventing Gpauk to be the release engineer.
I am only trying help Canyoncruz understand Graham's position on "prototypes".
But I dare to correct Graham: normal is, that a good prototyping work (no production tradeoffs and manufacturing being done by designer himself) to be always much better than the final product released
But I dare to correct Graham: normal is, that a good prototyping work (no production tradeoffs and manufacturing being done by designer himself) to be always much better than the final product released
Last edited:
Ah ok. I have to disagree. A pick and place machine will always do a better assembly job than a designer. That's why a good lab tech is worth their money as designers are often lousy at precision soldering.
Now clearly hand crafted boutique products and volume production are worlds apart, but a couple of spins is not abnormal. Annoying, but not abnormal.
Now clearly hand crafted boutique products and volume production are worlds apart, but a couple of spins is not abnormal. Annoying, but not abnormal.
- Status
- Not open for further replies.