Rodolfo,
I can only agree to your post. The discussion could be, what is an adequately low level to which the distortion should be beaten? It seems to me that if you don't see any signal artifacts (under the various conditions you noted) above the noise level, we have reached our target, provided the noise level is low enough. -100dB?
Inevitably, there will be those pointing out that theory predicts that components below the noise floor can be recovered, but I am not aware that HUMANS can do that.
There have been some impressive studies where humans could recognize extremely low level components at the noise level, notably by IRCAM in France. But, as always in these type of studies, it was with specifc single and multiple tone signals with a carefully selected level- and frequency ratio. In other words, fully artificial signals. I am not aware of any studies where listeners could recognise any artifacts even 80dB below output level in music and other non-correlated signal constructs.
Jan Didden
I can only agree to your post. The discussion could be, what is an adequately low level to which the distortion should be beaten? It seems to me that if you don't see any signal artifacts (under the various conditions you noted) above the noise level, we have reached our target, provided the noise level is low enough. -100dB?
Inevitably, there will be those pointing out that theory predicts that components below the noise floor can be recovered, but I am not aware that HUMANS can do that.
There have been some impressive studies where humans could recognize extremely low level components at the noise level, notably by IRCAM in France. But, as always in these type of studies, it was with specifc single and multiple tone signals with a carefully selected level- and frequency ratio. In other words, fully artificial signals. I am not aware of any studies where listeners could recognise any artifacts even 80dB below output level in music and other non-correlated signal constructs.
Jan Didden
janneman said:
No, the reason why there are those 618uV error triangular spikes is because the gain is not constant and experiments discontinuities, dips across the output current/voltage swing that require a great amount of drive signal (+/-9uA) in order to be compensated. In fact, I can't recognise any music or tone signal coming out of the LTP at all, I only see zero-current crossing events (where the O/L gain has an inherent strong discontinuity due to te nature of my circuit).
If I force the output stage to operate into class-A in a I/V region without discontinuities (constant 125dB O/L gain), I can no longer see any current swing in the output of the LTP
Also, the spikes are of approx 80uS width, so their frequency content is well above 10Khz, where the open loop gain starts to sharply roll-off at 9dB/oct (PSpice says thtat at 10Khz it has already decreased to 100dB or so). Of course, the circuit does its best to try to correct a dozen milivolts of 50, 100 or 200Khz error components in the output...
I'm considering ways to reduce those discontinuities.
janneman said:
Jan,
Probably there is no single answer to this.
On one hand, masking at the hearing process level is a well known phenomenom used for example for audio compression like MP3. What is an adequate threshold is therefore problematic to establish though undoubtedly it exists at least taking conservative (overkill) margins.
On the other hand, we do not all listen and perceive alike. There are naturally gifted persons, to which substantial training adds for a very high level of perception beyond what is the average threshold. Just like sports practice for example.
I place high bets on a better than 0.01% THD (across the full audio band and all load conditions) as a number representative of linearity such that for the *average* listener, there is not difference between original and amplified sound.
Rodolfo
Hi, Janneman,
I just experience this yesterday. I go to my friend's house. My hand is holding a portable SPL meter in his sound room. First the pink noise. Standard =90dB. I couldn't stand hearing the pink-noise in 90dB for 10minutes. I think putting people in 5 hours with 90dB pink noise will drive him nuts
But when I played some nice LP's, the time passes by, and the level is about 92-95dB, and my ear just enjoys it And 92-95dB just heard as "normal"
I know this is not related to amplifier, but is there anything "fatigue listening" for our ears?
I just experience this yesterday. I go to my friend's house. My hand is holding a portable SPL meter in his sound room. First the pink noise. Standard =90dB. I couldn't stand hearing the pink-noise in 90dB for 10minutes. I think putting people in 5 hours with 90dB pink noise will drive him nuts
But when I played some nice LP's, the time passes by, and the level is about 92-95dB, and my ear just enjoys it
And 92-95dB just heard as "normal"I know this is not related to amplifier, but is there anything "fatigue listening" for our ears?
lumanauw:
Pink noise and white noise are fatiguing because the average energy content is very high and is evenly spread across the audible frequency range. Our ear is much more sensitive in some frequency ranges than in others. Furthermore, the sensitivity is actually a dynamic parameter since some frequency ranges mask others and vice-versa, thus we require some energy in some frequency ranges to make other ranges pleasant. This means that carefully equalized pink noise will cause much less fatigue.
I don't know how to link these facts to amplifiers, because I'm also very involved into speaker design and experimentation, and frankly speaking, to me all amplifiers are almost 99% neutral in comparison with what speakers (and room acoustics) can do and will do to the sound. I will never trust any statement about "listening to amplifiers" coming from people that does not have a clue about speaker design and room acoustics and can't distinguish by ear a 500hz tone from a 1Khz tone or a 2Khz one, or a notch filter, or a tweeter hooked with the wrong polarity...
I have never found a "pleasant" or "fatiguing" amplifier, but I have run into thousands of "pleasant" and "fatiguing" speakers, rooms and music programs (badly equalized and mixed). In my experience, listening fatigue is always related to improper acoustic energy density distributions. I always use an equalizer since it allows to make most music listenable at most volumes with most gear, by adjusting acoustic energy ratios to match ear/brain expected patterns.
Maybe I'm too busy perceiving +6dB/-12dB peaks and dips and standing waves from speakers and rooms, that will make your ears bleed if you don't correct them quickly, that I can't perceive 0.1% amplifier distortion products???
Pink noise and white noise are fatiguing because the average energy content is very high and is evenly spread across the audible frequency range. Our ear is much more sensitive in some frequency ranges than in others. Furthermore, the sensitivity is actually a dynamic parameter since some frequency ranges mask others and vice-versa, thus we require some energy in some frequency ranges to make other ranges pleasant. This means that carefully equalized pink noise will cause much less fatigue.
I don't know how to link these facts to amplifiers, because I'm also very involved into speaker design and experimentation, and frankly speaking, to me all amplifiers are almost 99% neutral in comparison with what speakers (and room acoustics) can do and will do to the sound. I will never trust any statement about "listening to amplifiers" coming from people that does not have a clue about speaker design and room acoustics and can't distinguish by ear a 500hz tone from a 1Khz tone or a 2Khz one, or a notch filter, or a tweeter hooked with the wrong polarity...
I have never found a "pleasant" or "fatiguing" amplifier, but I have run into thousands of "pleasant" and "fatiguing" speakers, rooms and music programs (badly equalized and mixed). In my experience, listening fatigue is always related to improper acoustic energy density distributions. I always use an equalizer since it allows to make most music listenable at most volumes with most gear, by adjusting acoustic energy ratios to match ear/brain expected patterns.
Maybe I'm too busy perceiving +6dB/-12dB peaks and dips and standing waves from speakers and rooms, that will make your ears bleed if you don't correct them quickly, that I can't perceive 0.1% amplifier distortion products???
Hi, EVA,
Yes, that phenomenon has nothing to do with amplifiers. I just ask, wheter there is a "fatigue listening" to our ears. It seems so to me. If it existed, how much contribution from an amplifier?
About room and speakers. Yes, they are the no.1 that should be considered when playing good quality audio. We cannot cheat nature, speaker placement and room condition is a must.
I've got a friend that crazy enough about audio here, he CAN equalize flat RTA response without equalizer. How? He adjusted the room.. He can add or reduce 2khz +/-3 or 6db, for example, by adjusting the room. Adjusting height of the ceiling (yes, he do that all the time), adjusting damping material, adjusting % of wall covering, etc. But took very long time to do it. It takes him 7 months to know how adjust 1 frequency point.
When Janneman is already in Bandung, I will take him to see/hear his system
Yes, that phenomenon has nothing to do with amplifiers. I just ask, wheter there is a "fatigue listening" to our ears. It seems so to me. If it existed, how much contribution from an amplifier?
About room and speakers. Yes, they are the no.1 that should be considered when playing good quality audio. We cannot cheat nature, speaker placement and room condition is a must.
I've got a friend that crazy enough about audio here, he CAN equalize flat RTA response without equalizer. How? He adjusted the room..
He can add or reduce 2khz +/-3 or 6db, for example, by adjusting the room. Adjusting height of the ceiling (yes, he do that all the time), adjusting damping material, adjusting % of wall covering, etc. But took very long time to do it. It takes him 7 months to know how adjust 1 frequency point. When Janneman is already in Bandung, I will take him to see/hear his system
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.