If, as is customary in these types of tests, we pull bb because of possible disclosure and bias, the three remaining are clearly indistinguishable. That is in itself is also an interesting and usefull result matching the measurements.
Jan
Jan
The difference between bb and aa was a lot in my system (with speakers -I did not try with the headphones).
These years I have been eliminating different bottlenecks: basically I have been eliminating noise and RFI / EMI interferences (network and atmosphere) and almost incredible 3 Vdc !!!! from my very awful 230 Vac / 50 Hz electrical network.
I can understand that people do not differentiate between bb and dd, but not in the rest of comparatives.
These years I have been eliminating different bottlenecks: basically I have been eliminating noise and RFI / EMI interferences (network and atmosphere) and almost incredible 3 Vdc !!!! from my very awful 230 Vac / 50 Hz electrical network.
I can understand that people do not differentiate between bb and dd, but not in the rest of comparatives.
Last edited:
So you have good ears and a good system! 🙂
The measurable difference is also biggest between aa and bb, and audio diffmaker difference as well, though it is noise only and buried-in-noise mains components.
The input filter does two things: it degrades sound quality and it increase stability. So it is not fair for a stable amp to have high input filter when the unstabke one needs it.
Imo, to be fair, each opamp should have their own circuit. Lm4562 and Opa2134 needs comp cap (across fb resistor) Tl072 does not.
Trust me, is my tweaked system and not my ears.
The same bottlenecks affecting my main system - kidnaped by the television and the family. With same solutions and the sound much improved too.
The same bottlenecks affecting my main system - kidnaped by the television and the family. With same solutions and the sound much improved too.
Nonsense. Take a calculator and calculate -3dB frequency of the RC. Look at the frequency response measurements with wideband noise that I posted in my page.
BTW, no FB resistor there. You guys should at least check the page where I described and shown the test circuit. That is why the page was created.
Last edited:
Dont forget that whatever the circuits, LM4562 has the tendency or statistics to win the competition.
Most ears will appreciate the sound of LM4562 but imho, better ears will not easily prefer it.
I checked your cct. I was saying that i treat each opamp differently. To handle issue with fast opamps i use cap across fb resistor, not input filter or a cap across differential inputs.
Scott asked how bass sound can be different when there is sufficient gain and the tesponse is flat near dc. I didnt answer because when an expert ask something like that, most probably he is not ready to hear the answer. Bass is not perceived with the existence of only LF. Existence of HF and how they are time related change the perceived sound of bass. Latfet amps have very good HF performance and i think lower damping than BJT. I think those are reasons for the unique bass character.
BTW, any plan to post different clips with the same opamps to see how 'valid' is the result of the test?
Don't worry I can take it, how those 8-legs store some of the signal to change the time relationships is a mystery to me too. Maybe the medical and scientific community should go back to discrete, or better yet, tube design.
Last edited:
Well if the statement shows clueless of the difference between comp cap and input roll-off, and then comment on a circuit that was not used, you know it's going downhill. Fast.
Anyway, kudo's to PMA for taking the trouble to set this up. I hope all the pundits here realize that this is not about someone's personal preference. Someone may feel happy because his preferred opamp 'won', but realize that for anyone that voted for your favorite, someone else voted against it.
Excluding bb for earlier stated reasons, it's about equal, within statistical errors, between the other three. What that means is that there is no audible difference between those three. You may not like it but that's how it is.
What I do hope is that a similar (or the same) test can be done with much higher signal levels to exercise large signal linearity and slew rate. That should also be interesting.
Also hope that people going to participate in any follow up test will be able to refrain from trying to influence others before the end of the test. Intellectual honesty and all that.
Jan
Another big opamp listening test
LME49720+Shield+ALPS Potentiometer stereo preamplifier kit | eBay
Maybe this is a cheap and good preamp:
LME49720+Shield+ALPS Potentiometer stereo preamplifier kit | eBay
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
Yes, it can be done. At some realistic level, like 2 - 3 Vrms. And in 24-bits resolution, to prevent criticism, though I am sure it would bring nothing in this current test. If there are people interested in a new test, I can do it. At least I am able to assure correct technical conditions 😉
By the way, from the O2 and ODAC author:
Op Amp Measurements
-> https://nwavguy.blogspot.com.es/2011/08/op-amp-measurements.html
Op Amp Measurements
-> https://nwavguy.blogspot.com.es/2011/08/op-amp-measurements.html
I find this a remarkable result and just think for a moment the implications of this.
What Pavel is saying (and the tests back this up) are that nothing in the audio signal even remotely pushed the limits of the 1458, at least at these levels.
I've played with those chips before and recall that they are truly lousy when for example fed a 10 or 20kHz squarewave of even a few volts.
The other toss of the coin is that I also know how lousy the squarewave performance of 16 bit 44.1kHz audio is (and of course it can't be any other way).
I suspect the limitations would have shown more had the amplitude been significantly higher though.
Did you read Samuel's original, it's free. IMO his common mode test is unrealistic (for bi-polars) and is akin to exaggerating an effect in order to measure it. IIRC 100k (or even 10k) source resistance is absurd for low noise applications. Notice who wins 😀.
- Status
- Not open for further replies.