A few years back I had a test bench setup that used an integrated amp as the switching for several sound sources. I ran out of inputs, and cascaded an additional preamp made by the same manufacturer. One day I had occasion to make some routing changes, and found the same signal coming from the left and right channels into a matched set of speakers arranged on the bench as a stereo pair. The expected result would be a mono signal, with the typical tightly narrow sound source perception. BUT IT WAS WILDLY ARTIFICIAL STEREO SOUNDING! Now this wasn’t like a different pair of tuners on the same radio station or anything, it was a “y” splitter delivering the exact same source through two different paths. The low frequencies had not disappeared, so it wasn’t a gross phase misalignment like –180 or polarity inversion. All of the tone controls had their bypasses engaged. Essentially the only difference was a few feet of whatever cable was used in the patching, and that one channel passed directly into the integrated amp, and the other one first had to go through the preamp, and then follow basically the same route through the integrated amp as well. To confirm that it wasn’t speaker or placement related, I depressed the “mono” mode switch on the integrated amp. The sound field instantly collapsed into a very narrow mono window, and low frequencies were still present. The overall spectral balance was also largely preserved. The sound did not get bass shy, dull with HF roll-off nor hollow like with the “smile curve” so many people apply (to make up for waveform timing errors, among other things).
A number of years ago I did some calculations on the perception of sound arrival angles in the binaural system. My findings coincided with other articles on the upper range of frequencies involved in the perception of stereo image retrieval. The ear-brain timing response of human beings may be the most critical function for survival in the hearing subsystem of mammals. Perceiving accurately the location of a potential threat, perhaps while blinded by darkness, hiding or having it behind you, may make the difference between survival and oblivion. The almost unbelievable conclusion is that we have an internal response time as high as the 1Mhz range for arrival time differences when considering perception angle, phase and 10kHz sound source harmonics.
I propose a method for detecting audible differences in sequential parts of the reproduction signal chain. It involves branching using a system of known-to-be-similar components such as a split signal source, cabling, amplification and speakers/headphones. Level matching for equipment with a gain function is a given. Having a stereo/mono mode switch could be helpful. It is not the same thing as a DUT (device under test) bypass. However this arrangement can demonstrate the differences between two alternatives. The first is a solid monophonic perception experience through identical (actually always only NEARLY IDENTICAL) twin speakers of same manufacture and date, or dual headphone elements of the same design, and the parallel path through the singular component like the amplifier (or with the mono switch ON to make most of the path similar). The second is the phantom image that is generated when rather subtle changes in spectral response, noise floor, and phase timing are provided by the DUT on just one channel. The practitioner can make some allowance for propagation delay that is otherwise largely sonically neutral, by watching for a position shift in mono location, as opposed to a widening. The reason for my proposal is the originally ASTOUNDINGLY easy to detect change in image I experienced, and the fact that it could be corroborated with a scope (e.g. in X-Y mode). I was about 3ft from each speaker, and they were about 5 ft apart, so the orientation could be considered “near field”. Like headphones, which I have not tried yet but should give a similar effect, near field listening promotes the first arrival signal from the speakers to primary status over any room effects.
Very high quality equipment (not by price, but by design) subjected to such a test should maintain a solid mono image. A variation on the connections could simply diverge at the signal source and pass through alternate pieces of equipment. One would be a “reference” and the other the DUT. A phantom image would again reveal a difference, although not the source or reason for it. Null summing on an oscilloscope could show the difference signal between channels. Remember; don’t limit the data to the traditional audio bandwidth. While 1μs may be beyond the actual upper perceptual limit, it leaves one with the realization that the path a cable takes as it lays on the carpet or differences in PCB layout between channels during recording or playback may audibly effect the outcome of reproduced music.
One might be tempted to translate a perceived image difference into making direct inferences about specific sound quality in either the reference or DUT piece. That should be avoided. The obverse can be considered true, however. Little or no difference indicates “straight wire with or without gain” behavior by the DUT. Major easily detected differences provide the impetus to do more fact-based research on the reason. This could be especially useful to the “before and after” modification crowd. The relative merit of cables and other “external” mods might be more easy to sort from social effects. There are some issues this may not detect; one I can think of is microphonics, which need very specific equipment configurations to be highlighted (the obvious one is a turntable sitting on a speaker box).
I suggest some variation of this experiment to anyone who has been frustrated by the “all amplifiers are the same” argument and then shown the 20-20,000Hz flat line presented by the secular engineering community. In Science, there must always be a rational explanation of manifest phenomena. It is in fact the definitive canon of Science. The problems come when you look for an egg by picking up a rabbit, on a day that isn’t Easter. You have to know where you should be looking!
A number of years ago I did some calculations on the perception of sound arrival angles in the binaural system. My findings coincided with other articles on the upper range of frequencies involved in the perception of stereo image retrieval. The ear-brain timing response of human beings may be the most critical function for survival in the hearing subsystem of mammals. Perceiving accurately the location of a potential threat, perhaps while blinded by darkness, hiding or having it behind you, may make the difference between survival and oblivion. The almost unbelievable conclusion is that we have an internal response time as high as the 1Mhz range for arrival time differences when considering perception angle, phase and 10kHz sound source harmonics.
I propose a method for detecting audible differences in sequential parts of the reproduction signal chain. It involves branching using a system of known-to-be-similar components such as a split signal source, cabling, amplification and speakers/headphones. Level matching for equipment with a gain function is a given. Having a stereo/mono mode switch could be helpful. It is not the same thing as a DUT (device under test) bypass. However this arrangement can demonstrate the differences between two alternatives. The first is a solid monophonic perception experience through identical (actually always only NEARLY IDENTICAL) twin speakers of same manufacture and date, or dual headphone elements of the same design, and the parallel path through the singular component like the amplifier (or with the mono switch ON to make most of the path similar). The second is the phantom image that is generated when rather subtle changes in spectral response, noise floor, and phase timing are provided by the DUT on just one channel. The practitioner can make some allowance for propagation delay that is otherwise largely sonically neutral, by watching for a position shift in mono location, as opposed to a widening. The reason for my proposal is the originally ASTOUNDINGLY easy to detect change in image I experienced, and the fact that it could be corroborated with a scope (e.g. in X-Y mode). I was about 3ft from each speaker, and they were about 5 ft apart, so the orientation could be considered “near field”. Like headphones, which I have not tried yet but should give a similar effect, near field listening promotes the first arrival signal from the speakers to primary status over any room effects.
Very high quality equipment (not by price, but by design) subjected to such a test should maintain a solid mono image. A variation on the connections could simply diverge at the signal source and pass through alternate pieces of equipment. One would be a “reference” and the other the DUT. A phantom image would again reveal a difference, although not the source or reason for it. Null summing on an oscilloscope could show the difference signal between channels. Remember; don’t limit the data to the traditional audio bandwidth. While 1μs may be beyond the actual upper perceptual limit, it leaves one with the realization that the path a cable takes as it lays on the carpet or differences in PCB layout between channels during recording or playback may audibly effect the outcome of reproduced music.
One might be tempted to translate a perceived image difference into making direct inferences about specific sound quality in either the reference or DUT piece. That should be avoided. The obverse can be considered true, however. Little or no difference indicates “straight wire with or without gain” behavior by the DUT. Major easily detected differences provide the impetus to do more fact-based research on the reason. This could be especially useful to the “before and after” modification crowd. The relative merit of cables and other “external” mods might be more easy to sort from social effects. There are some issues this may not detect; one I can think of is microphonics, which need very specific equipment configurations to be highlighted (the obvious one is a turntable sitting on a speaker box).
I suggest some variation of this experiment to anyone who has been frustrated by the “all amplifiers are the same” argument and then shown the 20-20,000Hz flat line presented by the secular engineering community. In Science, there must always be a rational explanation of manifest phenomena. It is in fact the definitive canon of Science. The problems come when you look for an egg by picking up a rabbit, on a day that isn’t Easter. You have to know where you should be looking!
Exactly, Westerp! I tried to acknowledge that a certain mixing occurs when using the downstream amplifier's mode switch, as compared to a true bypass of the DUT. But in mono mode, the paths through the 2nd amp are largely similar. In this case the gear was some vintage Technics gear. The integrated was the SU-V9 and the preamp was the SU-A6. The date of manufacture, parts and internal layout/design engineers/philosophy were all related. However, the extra trip for just one channel through the preamp made highly audible changes.
http://www.vintagetechnics.co.uk/controlpower/sea5sua6.htm
The very strong phantom image indicated to me that the SU-A6 is horribly colored in its transmission of a signal from its line level inputs to its main outputs. The equipment has been struck and is in storage. I am tempted to reconnect it to see if its record out has similar or lesser coloration. While the switches and pots have been cleaned, it is a 25 year old unit. I hold the suspicion that the more crap a signal has to struggle through, the worse its micro-detail.
The reason the above method could be so powerful as a subjective tool is the concept of biological leverage. I have read of cat, gerbil and chicken brain dissection and live stimulus studies of the bilateral interaural cascade effect. Under normal chemical messenger release at a synapse (gap between axon of one neuron and the surface of another neuron), most nerve pathways in humans run at a surprisingly languid rate of around 40hz, or 25ms per synapse. I'm sure you've heard about recommended follow distances in traffic on the freeway. This is based on the velocity of the cars and the typical cognitive & motor response rate of 200ms, or an 8-10 neuron sequence for nerve impulses to convert what the eye sees into the foot pressure on the break pedal. This long time propagation would never work for stereo sound localization. As it turns out, many mammals have a special pair of structures that I remember are termed "medial superior olives". Imagine a rope ladder suspended from two points in space of equal height, but independently adjustable. Now place the equivalent of a dinner plate horizontally set upon one of the rungs and a bowling ball set onto the plate. A relatively small change in the height of one of the ropes making the ladder sides will cause the plate to tilt a correspondingly small angle. The bowling ball, being barely stable, readily rolls to the direction of the lower rope without much delay. It is this type of "signal" mammals use to determine the direction of a sound. The response time, in comparison to other sensory activity, is near instantaneous. There are some frequency dependent effects, and I don't pretend to understand the exact neural structures involved. The research in this area seems fairly recent.
http://www.ionchannels.org/showabstract.php?pmid=17690144&redirect=yes&terms=Medial+Superior+Olive
My suggestion is to harness this effect of localization sensitivity to essentially take a look at a musical signal using a electrochemical mental trick, a sort of "side view" that will amplify the perception of normally subtle audible differences to make them discernible. In the case that they are easier to perceive, their existence can be verified by a wider audience, including the technically trained and inclined. This opens the opportunity to explore the nature and causes of many psychoacoustical and euphonic phenomena. It also would allow the dismissal of other "effects" that are not discernible with the more "sensitive" method.
A perfect example would be to use twin lengths of different cable designs. Start with a "y" made of PCB and connectors with a geometric mirror image. Run the two cables in the same physical orientation including curves, etc. Operate the stereo mono switch and see if the sound field changes. Since I propose this is far more sensitive than "acoustic memory" or trying to judge the effects through abx or "social expectation testing", such experiments can be either vindication or negation for certain comparisons.
The primary purpose of Science can't be taken to spoil anyone's fun. It IS FUN! Never be afraid to look at the truth, whatever it may be...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.