You can also view it as an exageration to the point where no one could deny that it would be measurable and/or audible.
Yes, taste, and ears sensitivity to hear things that will affect preference.
Look at the opamp blind tests on this site. Conduct another similar test. I believe the LM4562 will be the winner. I think this is because its strength is easy to perceive (resolution & dynamics) while the weaknesses are only perceivable by more sensitive and trained ears...
This should beg a BIG question... But people are too busy with scratching the surface...
The above member, prefers 4558 over 4562 and 5532, isn't that a surprise 😀
(In my experience, 4558 is only good enough when the power supply is +/-5V, lacking competition may be)
Okay, what about an exaggeration to the point where 50% of people would not deny it would be measurable and/or audible?
(I mostly tend to find absolutes uninteresting, except maybe to identify endpoints of a continuum. The interesting stuff is usually somewhere in between.)
Yes, you are right. But remember, I see the case from different perspective (where I understand the effect of the "control").
Sure. Just like the above where tonitonitoni prefers 4558 over 4562 and 5532. It is not a proof that people hear things that do not exist. It is because they have different taste/preference and they have different ability to hear things that will affect their preference.
I'm not surprised at all if someone prefer the old 4558 when compared to 4562 or 5532. But if compared to TL072, OPA2134, LF412, etc, I will be very surprised (because I cannot see how it is possible, unless he hears things that do not exist, or broken circuit).
Last edited:
Yes. In this case, I will challenge them to do 8/8 with FoobarABX to prove it, and I believe they cannot pass, unless there is a gross control, such as "level not matched"...
For me, level difference of even LESS than 0.1dB has an effect on preference (the louder will have the benefit)...
In two extremely similar circuits, I have used 1943/5200 output's and TIP35/6 outputs. I have also used MJE340/50 and 4793/1837 drivers with TIP35/6 outputs. The different combinations do feel like they sound different to me. The 1943/5200 I immediately didn't quite like. Maybe I'm mad, but they sound accurate and fast, but lacking in depth. The TIP35/6's sound sweet and heavy. As far as drivers, MJE340/50 sound slower and less accurate than 4793/1837. I've found that 4793/1837 in combination with TIP35/6 sounds the best to me. It's almost as if the TIP's give the feeling that there's always more power, and the accuracy comes from the 4793/1837's. The 5200/1943's sound like you're not sure if there's more power available.
Now, I'm really not one to write what seems to be utter nonsense like this, but these are my first impressions when I've built an amp, and I hear it for the first time. And honestly, I probably couldn't tell the difference in an AB test, but I'll do it one day. But again, in my experience, I can understand how some may be able to recognize different op amps by their sound.
Commendable at least that you're admitting they're assertions rather than as claimed previously, explanations 😎 Some progress.
No, that's not nonsense. That is how it is (the sound).
Of course, if you "roll" transistors in the same circuit, the performance of the circuit will change, just like when you roll opamps in fixed circuit. So from that POV, the sound of the circuit is not the sound of the transistors.
When you change the transistors, many parameters of the amplifier will change also, hence the sound...
I like to check my transistors to find their "sweet-spot". Then I try to design a circuit that will make use of these sweet spots of all the transistors in the circuit.
I tried the feather test last night. Could not tell the difference as hard as I tried (my wife did the test on me). It was the same when I swapped LM7910 for SA5534 in my Symphony pre a few years ago and listened carefully through a cheap pair of $350 headphones. No difference.
After the feather test last night, we did the hammer test.
I am writing this from the emergency room of the local hospital.
This is more a psychological thing than technical (I'm not arguing on the long-term memory issue)
Of course, I understand your point and I agree, to certain extent. You can say this to the majority of people:
(as if no exclusion is used in the statement). But I know it wont apply to me.
What is the objective of the test? To know your thresholds??
If you want to sell your pre commercially and you need to know the subjective quality between LM7910 and SA5534, you can set a listening panel consisting of people with better ears than yours. With less effort (and money), you can create the clips and post them here.
Now you are talking closer to how to start thinking about how to measure what you hear. If you can record the sound of different opamps and preserve what you hear in the recording, we would be getting a step closer to measure-ability.
For example, much of what I find or have found objectionable about a lot of digital sound is in HF. If that sound could be captured at 88.2 kHz, then sharply band limited in the range 10-20 kHz, then it could be played back at 44.1 sample rate which would shift the spectrum down to where possibly anybody could hear it. Something like that, experiment and find some way to highlight the differences. If we don't have instrumentation to measure it, at least make it hearable by more people who might then have some ideas about measurement approaches.
Last edited:
I am sure my thresholds are within the range for someone of my age. My 25 and 28 year old sons (both very good musicians and one very involved in recording music, sound systems at thd time, with degree in that field) also struggled to hear a difference.
But Jay, clearly you are at another level in these matters. I salute you and trust your DNA will propagate through the human genome in centuries to come so we can all hear these differences as easily as you (in well designed and engineered gain stages of course).
I don't listen to opamps. I listen to working circuits. I never ended up using either the 49710 or the 5534 in the final design for reasons other than what they sounded like.
LM7910 looks to me like a 3 terminal reg. In an opamp socket with no difference heard? Expectation bias I think can be the only possible explanation.
To prove they can't hear a difference? You lost me. Max thinks FoobarABX is fubar so you folks with these talents don't seem to agree again. There are dozens of op-amp rolling threads across the many forums there is no consensus lots of different winners and losers. You're getting close to, "they have to be deaf or liars", we've been there before.
Last edited:
- Status
- Not open for further replies.