The Half Life of Expectation Bias
- Thread starter mep
- Start date
- Status
This thread started in a open-ended, philosophical manner, but, like so many others before it, somehow became personal. Is there another form of bias we should be discussing here?
Well Geardaddy, I started this thread to have an honest discussion about why I don't believe that expectation bias is something that has permanence. The thread I started pretty much ran its course and then devolved into why headphones are better than speakers and Tom telling everyone that there has been no advancement in 2 channel high end analog gear in the last 20 years and me getting accused of making up that statement by Dingus.
That would be a sight worth seeing. There's a lot primping (or is that pimping?) going on here.
i stand corrected and i concede the point.
here is the quote snippet;
i do agree there have been improvements in this regard, but i dont believe there is much that has been truly significant, or real advancements as the quote says. sure tooling, computing/software and other aspects of the development process have been greatly improved to provide better parts and components at lower costs, but i see those as advancements in manufacturing rather than advancements in 2-channel stereo.
take the Krell KSA-250 for instance. introduced over 20 years ago, today we have better ways of building it, but have the technological advancements in amplification given us anything significantly better in terms of sound quality? go back another ten years to the Krell KSA-100 and i think the same question has a lot of merit. we can do more of the same, but have we actually done better?
Dingus-I appreciate that you realize I was telling the truth and didn't just make this up. As far as your thoughts on the progression or lack thereof with regards to 2 channel analog playback, I disagree. You are discounting the improvements we have seen in resistors and capacitors and the overall design of power supplies. I owned a KSA-250 which you may or may not know as well as a KRC-HR preamp. I still think the KRC-HR is a great sounding preamp and so is the KSA-250 power amp. I haven't owned the ARC REF 5SE and the REF 75 long enough yet to make any definitive statements on which is the better sounding. I still have hundreds of hours to go before I reach the recommended 600 hour break-in.
I think if you take a look back at components made 20+ years ago, you will see the dynamic range and S/N ratio of modern high-end components has improved. I do think that the old Krell company (when Dan was still running it) was ahead of its time compared to some other companies.
I think if you take a look back at components made 20+ years ago, you will see the dynamic range and S/N ratio of modern high-end components has improved. I do think that the old Krell company (when Dan was still running it) was ahead of its time compared to some other companies.
Let's take the KSA-250. You should have a listen to Brett's new amplifier's that are based on the 250 and then get back to us.
Moreso, I suggest people take (like I have on occasion) some of what was highly regarded 20 years ago and put it up against the best of today's gear. Sorry, but it's not even close. So bad in fact, that one begins to doubt their hearing and how that equipment was even deemed acceptable back then. Biggest improvements? You name it. Transparency, performance at the frequency extremes, resolution, noise, soundstaging and imaging, micro (!) and macro-dynamics, timbre, etc. YMMV.
Let's take the KSA-250. You should have a listen to Brett's new amplifier's that are based on the 250 and then get back to us.
Moreso, I suggest people take (like I have on occasion) some of what was highly regarded 20 years ago and put it up against the best of today's gear. Sorry, but it's not even close. So bad in fact, that one begins to doubt their hearing and how that equipment was even deemed acceptable back then. Biggest improvements? You name it. Transparency, performance at the frequency extremes, resolution, noise, soundstaging and imaging, micro (!) and macro-dynamics, timbre, etc. YMMV.
in my experience, evenly capable amps can vary widely in the way they sound, but i've never been able to discern which produces the better sound. maybe i need to expand my audio horizons (and really, who couldnt?), and i think both of our opinions here could be fairly attributed to expectation bias, but the more important issue is knowing where our subjectivity takes over. in regards to improving fidelity, i dont really know if the science behind the gear has actually technologically moved forward in any truly significant manner since the early 80's, but i cant really tell that it has either.
Sorry but I don't buy into this expectation bias. The only thing I expect is that most gear will not perform up to snuff until properly broken in. Not only that, there's usually an inverse correlation between stuff that sounds good right out of the box and long term listening satisfaction.
i didnt say expectation bias was factor in shaping our opinions about amps or gear in general. i said it could be, in order to illuminate the bigger issue of subjectivity when we make our assessments of how something sounds.
Where's the aspirin !
I would be happy if Tom would learn how to use properly the quote feature of WBF. Otherwise I will need something stronger that aspirin.
I respectful disagree with your valuation of the improvements. As my opinion is supported mainly by subjective findings obtained through listening and reading about stereo reproduction, and most probably the systems and experiences I could refer do not mean anything to you, readers will have to rely on their experiences to appreciate the argument.
But I answered mainly to tell that IMHO we will never see real improvements looking at singe pieces of equipment - you must listen to properly assembled and optimized FULL systems in ADEQUATE rooms.
Can you please clarify what you would consider to be "truly significant or real advancements"?
One example I would cite in the bargain home theater receiver department is the incorporation of "Audysey" calibration to address frequency and time domain sonic abnormalities in a typical listening room when listening to multi-channel video / audio discs.
There are higher powered versions of this available for two channel end systems.
PS: Another would be the improvement in digital technology such as jitter reduction, oversampling, etc. that, IMHO, has resulted in major advances in sound quality versus what was available 20 years ago.
Last edited:
Oh, the righteous indignation. We have already traveled down that path before Gary. Teflon caps take a long time before they break in according to the companies that use them and the owners who have bought them. It's not like your new gear sounds horrible until the break in process is over.
To try to bring the thread back to the original post, a number of points spring to mind:
- Sometimes expectation bias & preference gets defined as psychoacoustics which I think is worth teasing out. I see it as follows which I believe is supported by the science of psychoacoustics itself "The way I use the term psychoacoustics is to describe the function of the transducer in the ear and the processing applied by the brain in interpreting that." Some people seem to equate this term with all things that are variable & unreliable in what we hear - expectation bias, preference, etc.
- This is an important distinction because psychoacoustics explains how we hear & make sense of the real world through our sense of hearing. It is a study of the workings of the perception we call hearing & works roughly equivalently in the majority of humans & has it's equivalence in animals also.
- The most interesting area of investigation is the field of auditory scene analysis which is where I believe most of the disagreement about measurements occurs. The old quip "we can hear things which aren't measurable" Vs "we can measure all that the ear can hear" is the distillation of this difference. What is being spoken about in the first quote is the soundwaves impinging on the ear across time & it's post-processing by the brain leading to "near real-time" perception of hearing. What is being referred to in the second quote is simply the waveform that impinges on the ear.
- Some of the things auditory scene analysis tries to explain are the techniques we inherently use to follow a conversation in a noise environment or how we follow one instrumental line in an orchestral performance. Essentially how we determine what sounds we should group together in the stream of waves hitting the ears to achieve the above phenomena. It appears that these are not learned phenomena as they are there in young babies (or at least the rudiments of them are there). It also appears that they are found in the animal kingdom. So our perception of hearing is wired to be able to achieve these sorts of phenomena.
- Now this scientific area of study has not yet been able to explain how we do this - as usual there are various conflicting theories.
- the even more tricky bit is that 2 channel stereo playback is not how we are presented with sound in the real world but it seems to have ticked enough psychoacoustic cues that we are fooled into the perception of it as a 3 dimensional sound scape reminiscent of our real-world auditory expectation of such an event. So what we are trying to achieve in our playback systems are even more of the cues needed to appear to be even more "realistic sounding" but without knowing exactly what these cues are.
Is it any wonder that the final arbiter is listening rather than measurements - until we have a theory that explains these psychoacoustic techniques, we are far from being able to measure them.
So, I believe what the op is saying (& what the question was) is that over a longer timeframe of listening to a piece of audio equipment, the majority will have a tendency to fall back on our "natural" hearing perceptions & use this as the basis for our judgement. I also believe that this is the best way to overcome any expectation bias which we might have initially brought to the table. I would concede that it does depend to some extent on how deeply ingrained the bias is & also will depend on the biggest overriding bias in all of this - whether one "allows" oneself to trust one's hearing - I contend that without this trust, the psychoacoustic higher processes will be disturbed & the ability to use "natural" hearing will affect our evaluation.
- Sometimes expectation bias & preference gets defined as psychoacoustics which I think is worth teasing out. I see it as follows which I believe is supported by the science of psychoacoustics itself "The way I use the term psychoacoustics is to describe the function of the transducer in the ear and the processing applied by the brain in interpreting that." Some people seem to equate this term with all things that are variable & unreliable in what we hear - expectation bias, preference, etc.
- This is an important distinction because psychoacoustics explains how we hear & make sense of the real world through our sense of hearing. It is a study of the workings of the perception we call hearing & works roughly equivalently in the majority of humans & has it's equivalence in animals also.
- The most interesting area of investigation is the field of auditory scene analysis which is where I believe most of the disagreement about measurements occurs. The old quip "we can hear things which aren't measurable" Vs "we can measure all that the ear can hear" is the distillation of this difference. What is being spoken about in the first quote is the soundwaves impinging on the ear across time & it's post-processing by the brain leading to "near real-time" perception of hearing. What is being referred to in the second quote is simply the waveform that impinges on the ear.
- Some of the things auditory scene analysis tries to explain are the techniques we inherently use to follow a conversation in a noise environment or how we follow one instrumental line in an orchestral performance. Essentially how we determine what sounds we should group together in the stream of waves hitting the ears to achieve the above phenomena. It appears that these are not learned phenomena as they are there in young babies (or at least the rudiments of them are there). It also appears that they are found in the animal kingdom. So our perception of hearing is wired to be able to achieve these sorts of phenomena.
- Now this scientific area of study has not yet been able to explain how we do this - as usual there are various conflicting theories.
- the even more tricky bit is that 2 channel stereo playback is not how we are presented with sound in the real world but it seems to have ticked enough psychoacoustic cues that we are fooled into the perception of it as a 3 dimensional sound scape reminiscent of our real-world auditory expectation of such an event. So what we are trying to achieve in our playback systems are even more of the cues needed to appear to be even more "realistic sounding" but without knowing exactly what these cues are.
Is it any wonder that the final arbiter is listening rather than measurements - until we have a theory that explains these psychoacoustic techniques, we are far from being able to measure them.
So, I believe what the op is saying (& what the question was) is that over a longer timeframe of listening to a piece of audio equipment, the majority will have a tendency to fall back on our "natural" hearing perceptions & use this as the basis for our judgement. I also believe that this is the best way to overcome any expectation bias which we might have initially brought to the table. I would concede that it does depend to some extent on how deeply ingrained the bias is & also will depend on the biggest overriding bias in all of this - whether one "allows" oneself to trust one's hearing - I contend that without this trust, the psychoacoustic higher processes will be disturbed & the ability to use "natural" hearing will affect our evaluation.
- Status
Similar threads
