What is the scientific value knowing they are different?
"More scientific' = 'Less than scientific' - Blind comparisons are not going to give us a better transistor, car or airplane. Amplifiers can be added to the list.
Sure, nobody ask you to do other than what you want, as others are also free to do as they see fit. There is nothing wrong to seek greater entertainment, audio reproduction is no dialysis.
This is a very good point as well
As I mentioned earlier, the "holographic image" is a second priority effect for me, although it's fun and it's highly appreciated by many listeners. The main observation is that the low-global-feedback amplifiers provide more precise localization of sound sources in horizontal axis than the ones with high amounts of global feedback. The statement "More global feedback make things better" has to be seen as "More global feedback make SOME things better".
Discussion about the reasons for this kind of behavior requires far more research. Good material for the beginning is referenced by XRK.
Many experienced listeners who auditioned the "holographic image", were saying "How can I make my system sounding this way?"
That's the most important appreciation of a sound reproduction arrangement from my point of view
Discussion about the reasons for this kind of behavior requires far more research. Good material for the beginning is referenced by XRK.
Many experienced listeners who auditioned the "holographic image", were saying "How can I make my system sounding this way?"
That's the most important appreciation of a sound reproduction arrangement from my point of view
We already know they are different. The issue is whether they sound different. This is useful information. If they sound the same then by definition one cannot sound better than the other.spladski said:
Silly comparison. Better transistors, cars or aeroplanes can be measured. Better sound can only be measured if first we determine what sounds better, which means finding what actually sounds different.
There is a post by Plasnu on H2 thread that may be relevant to the subject.
This is a circular argument. Amp A sounds better than amp B and then you change the speakers and amp B sounds better than amp A.
Last edited:
This doesnt make sense. How does a phase shift in an amp at 100khz change anything in a signal that only has content to 20khz? It doesnt, so why would you hear any difference. Your taking pinnea reflection data and extrapolating to amp phase shift? Quite a stretch. Has any one done a blind test on the audibility of this? It would be very easy to add phase shift for a test. And you would think the smallest amount of wow and flutter would totaly ruin a soundstage. Not to mention moving your head a few millimeters.
Last edited:
Why is larger better? Wouldnt you want to hear the same sound stage as the guys in the mix? There are many ways to make it larger, move your speakers further apart, dosnt mean its right.
About global feedback, I just asked John Siau about his Benchmark AHB2 amplifier which has amazing numbers in the measurements that Amirm has made. Waiting for the answer.
Review and Measurements of Benchmark AHB2 Amp | Audio Science Review (ASR) Forum
Last edited:
When you ask about "global feedback", the first question I have is at what frequency, because the open loop gain falls at 6dB/octave above the dominant pole. So I will assume you mean at frequencies below the dominant pole. For frequencies above the dominant pole, the input, diff-amp, has to work harder and harder to maintain a flat output. So depending on the ultimate bandwidth, the input may clip, which is what we used to call TIM, or slew induced distortion.
My point is that while global feedback is associated with the problem, it is not the actual cause. There are several ways to improve performance without changing the amount of feedback.
However this is 2019 and we have 40MHz bjt power transistors as well as MOSFETs, so there should not be any more slew problems unless you get crazy with overly complicated circuits. And the days of component sound systems is past except for a few "audiophiles". Most of the classic amplifier discussions become mute in a world of computer based entertainment and class-D audio power.
My point is that while global feedback is associated with the problem, it is not the actual cause. There are several ways to improve performance without changing the amount of feedback.
However this is 2019 and we have 40MHz bjt power transistors as well as MOSFETs, so there should not be any more slew problems unless you get crazy with overly complicated circuits. And the days of component sound systems is past except for a few "audiophiles". Most of the classic amplifier discussions become mute in a world of computer based entertainment and class-D audio power.
There are three contenders of which two were named here by a fellow i know:Off topic: I am curious about this topology that can give output voltage swings higher than usual Op-amp power supply rails.
Op-amp front end with an unusual OPS that needs floating PSU.
Is this a worthwhile topology for high quality sound ?
https://www.diyaudio.com/community/threads/class-ap-amplification.262821/#post-4077413
The third was once found on Rutger.nl site that once had the ssa30 and ssa40 projects.I have no idea why it dissapeared as it had clear audiophile biases and that was the only project I found measuring the actual distortions when driving a pair of QUAD ESL 63 proving it was a very serious choice for driving tough loads.I have the whole site saved on one of my laptops fortunately.
Most probably we'll never have proper discussions about the QSC brand on diyaudio although there are already at least 3 decades of serious QSC dominance in the stage amp market.Way too many people would argue in favor of a collector output screwing phase response although these amplifiers are driving the most difficult loads on the market backing the largest back emf potential .
A few months ago i watched a Nelson Pass interview where he was praising all sort of "audiophile " companies , all of the expensive and fully packed with electronic components ones, none of the genious ones .There are 19 years since i discovered qsc and i stayed firm on my beliefs that Peter Quilter's design will remain as the most ingenious design in audio amplifiers even though we'll never be able to dump the other Peter's Quad 405 as the only other real competitor for this title.We can find way more interesting designs with better specs, but none will offer qsc simplicity and reliability while having just the right offer.Besides they used medium quality components but nothing will stop you simulating a qsc design with the best transistors , CRD and op-amps for the job.
Because anything else is biased, its the only unbiased test of auditory system because there is no a-priori knowledge and no judgement involved. Its the only way to gain verifyable knowledge of the capabilities of the auditory system (per individual and en-masse), and its still only statistical so requires a lot of work to get good data.
Science in general has to be extra careful to avoid expectation bias - but usually the measuring equipment is automated and not subject to subjectivity, but the human auditory system is not automated and highly suggestible - just check out the McGurk effect to see how completely malleable and up-for-grabs it is - much more so than vision for instance.
The human auditory memory is also very short term, a few seconds, so good XY testing allows each subject to switch between X and Y frequently as often as they like. Trying to compare a sound you heard 5 minutes ago to a sound you are currently listening to is usually a fruitless pursuit if the differences are subtle, at the threshold of perception, which is where XY comparisons excel.
Knowing things are different is not useful without a reference, which is a third object. I don't design with blind AB. Designs are sighted and performance technical and auditory is correlated with measurements. Being involved in design is about exploration, therefore expectation bias should not be involved. Expectation bias is for punters who have paid $$ for one product and $$$ for another.