[MrHaelscheir]

Hello. In response to WaveTheory's recent "Cables Matter" videos, claiming such things as cables being able to "smooth out peaky treble" or some cables having innate "peaky treble", and their original four similarly named videos covering the possible science behind the audibility of cable differences, I have gone ahead and taken a few measurements with my in-ear microphones obtained from Earfish (I will be reporting on my findings with binaural headtracking soon), whether or not one sees benefits to the using of real human ears. At best, these measurement conditions should reveal practical frequency response variations caused by the mere act of slightly moving your head or jaw.

In short, the "cable science" videos explain fundamentals reasonably, making very clear to all of us what sound is (unfortunately never mentioning magnitude or phase versus frequency response; take some time to convince yourself that "sound" is a collection of longitudinal air vibrations that can be divided into individual superimposed sine waves arriving with varying phases, and that audio systems alter this in various ways, whereby true transparency aims to leave this frequency content untouched and not introduce any new sine waves that don't belong (nonlinear distortion)), then in the third and fourth parts mainly point out miniscule phase distortions caused by cable velocity factors (without ever mentioning the word "phase") as potentially causing time domain waveform changes detectable by human pattern recognition facilities (I have many issues with this).

This report will proceed to demonstrate the lack of frequency response differences between the measured cables mind between balanced and unbalanced that cannot be attributed to minor movements of the headphones or skin, and that phase variations due to head movements disturbing the headphones are most likely to create significant phase distortions past 1 kHz far surpassing any that could be introduced by cables, such not considering literature on the inaudibility of phase distortions in most audio.

Disclaimer: Note that I am not discounting the reality of these subjective cable experiences for users, but rather the cause or the belief that these subjective effects are physically innate to the cables rather than consistent perceptual distortions molded by knowledge of what cable one is listening to. As I had discussed at length in https://audiosciencereview.com/foru...rent-are-that-many-confused.9245/post-1522578 (post #5,428), if those subjective effects are consistent to you within your practical, sighted listening conditions, then all power to you to enjoy your gear, but just know that there is currently no science that should suggest that these effects are remotely guaranteed to be perceived by others and that such investments are more effective than properly done EQ and other digital processing.

2023-11-12 update: Please disregard all theorizing below regarding the phase response being extremely sensitive to the slightest perturbations to the headphones' or in-ear microphones' positioning. After acquiring a proper audio interface to ensure proper sample synchronization, as can be seen in https://www.head-fi.org/threads/con...bles-mattering-for-audio.970430/post-17816683 (post #8), the phase response consistency has been drastically improved, whereby the phase responses between these cables though these respective headphones are effectively identical.

Methodology:

In https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152 and the following), you will see my acquisition of a pair of in-ear microphones from Earfish and EQed frequency response, phase, and distortion measurements taken for the Meze Elite and Arya Stealth. Below is the full measurement methodology for this report using that equipment:

• I am seated on my computer chair before my computer desk and measurement equipment. The ambient noise level was between 35 to 40 dBA.
• The in-ear microphones are inserted as deeply as possible with the capsules facing the ear canal entrance such that the ear canal resonance is excluded from the frequency response measurement; the memory foam mounts also protect the eardrums from the loud volumes needed for high-SNR measurements.
• The included Axagon USB HQ Audio Mini Adapter is mounted to the back of a headband (meant for HRTF measurements), the in-ear microphones wires being held taut between the headband and the head to prevent the seating of the in-ear microphones from being disturbed during and between measurements. The wires held or pulled up so that they are pressed only by the upper part of the headphone earpads.
• Room EQ Wizard (REW) is used to facilitate the frequency response and distortion measurements, the Axagon adapter and FiiO K9 Pro ESS DAC/amp being selected as the input and output devices in non-exclusive mode, all EQ applications being disabled. The Windows Mixer volume settings for both devices are set to 100%.
• The DAC/amp is set to low gain.
• The cables are connected directly to the DAC/amp's respective outputs with all other ports unoccupied.
• The "Measure" button on the upper left of REW is clicked, the measurement length (in samples) set to 4M to maximum single-repetition noise reduction, and the sample rate set to 192 kHz purely to reduce the time duration of the measurement.
• "Check levels" in the "Measure" window is clicked and played for each channel to facilitate volume matching such that switching between measurements in REW best shows the identicality.
• "Start" in the "Measure" window is clicked and my right hand placed on top of the DAC/amp as I have found that this somehow reduces the noise floor by around 10 dBFS in the "Check levels" sound level log on the lower left such that vibrations from my heartbeat are visible. I keep my head and body still for the duration of the measurement with light, regular breathing, avoiding twitching my ears.
• The "Controls" button on the upper right of the measurement viewing window is clicked and "Unwrap Phase" clicked, the phase graph then positioned with the +-360 buttons for convenient viewing (these settings may or not be carried over to your own instance of REW, and you might also have to click "Limits" on the upper right and "Apply Settings" for the default to obtain the intended viewing scale).
• Cables are switched with the headphones still on the head, trying to disturb their position as minimally as possible.

Observations:

The full REW .mdat files containing the measurements can be found at https://drive.google.com/file/d/1XG3cA9byn0p8BHBqYiOx4wF8s5DS8-KV/view?usp=share_link (the file size is too large to attach here) with comments explaining each measurement.

Only left channel measurement images are shared in this post since the right in-ear microphone was found to have bass and top octave roll-off compared to the left in-ear microphone which was found to be more accurate, roughly confirmed against outdoor on-axis measurements of a Genelec 8341A calibrated with GLM.

This report assumes an understanding of how to read magnitude and phase versus frequency response graphs as well as REW distortion graphs, all of which can be learned online. Consider frequency response graphs as summarizing the "wave superposition" described by WaveTheory in their "cable science" videos.

The below graphs show the magnitude versus frequency response on the top and the unwrapped phase versus frequency response on the bottom.

Meze Elite:

Here, we see a comparison between two balanced cables of different tiers within the same brand. The below measurements targeted 105 dB SPL uncalibrated on the "Measure" window's "Check levels" output, corresponding with around 96.5 dBA when placing an A-weighted SPL meter with wind shield lightly against the inner driver grill of the right driver unsealed and facing outward.

Magnitude and phase versus frequency responses:

Figure 1: Frequency response for Meze Elite hybrid pads with Meze OFC 99.95% Copper Standard Cable Balanced 4 pin XLR Cable​

Figure 2: Frequency response for Meze Elite hybrid pads with Meze Mini XLR to 4.4 mm Balanced Silver Plated PCUHD Copper Premium Cable​

At this point, you will have to just trust that I definitely did swap the cables and correctly kept track of the measurements, since the frequency responses are nigh identical (within the 20 Hz to 20 kHz audio band; I would highly doubt that ultrasonic performance matters, especially when your DAC already filters the ultrasonic components out per the Nyquist-Shannon sampling theorem). When switching between the premium and stock measurements for the same channel when viewing with REW, you will see minor variations of the sort you would see with the same cable when slightly disturbing the headphone's or in-ear microphones' positioning. So how are you to know if the differences that you hear are due to the cable or due to the headphones being positioned differently, especially in the case that you remove the headphones first to change the cable? Regardless, what is clear is that the 99.95% copper cable certainly does not (measurably) smooth out treble peaks mind the 6 kHz peak, and neither does the act of plating the copper with silver increase the brightness, both effects being achievable for free with EQ. The only reason I own this premium upgrade cable is that my preferred 4.4 mm termination was not available for the stock cable, and that it feels nice and is a joy to look at every day as is the Tungsten version of the headphone itself.

The stock cable's measured magnitude versus frequency response was maybe a fraction of a dB lower than its premium cable counterpart's, though whatever factors in the cable or DAC/amp may have contributed to this, I doubt that this would be enough to cause perception of massive increases in "soundstage" or "resolution" among other claimed phenomena.

You may notice that the phase responses past 1 kHz show quite noticeable differences (not considering the literature regarding the inaudibility of phase distortions for certain audio; in short, the cochlea performs the equivalent of a Fourier transform such that neurons are only excited by individual narrow frequency bands, your brain this primarily hearing the magnitude response sans phase information; see https://www.phon.ucl.ac.uk/courses/spsci/AUDL4007/12.pdf which explains the entire auditory path), which surprised me at first, though again, I thought it common knowledge that these are largely inaudible in these conditions (except maybe for the quality and decay of sharp transients). After conducting more measurements as included and annotated in the full REW measurements, I found that even for the same cable while sitting as still as possible, small perturbations from my unavoidable slight head movements or ear twitches between measurements were likely the cause of these significant changes in phase response past 1 kHz, unless it is rather an error of my measurement system. Otherwise, when looking through the various measurements, it could be figured that any differences in phase response between the two cables were comparable to the differences that could be measured from the same cable due to head perturbations, there also being some measurements where the phase response for both cables are quite similar. Thus, there is likely no significant difference in phase response between these cables.

Distortion comparison:

Figure 3: Distortion measurement for Meze Elite hybrid pads with Meze OFC 99.95% Copper Standard Cable Balanced 4 pin XLR Cable​

Figure 3: Distortion measurement for Meze Elite hybrid pads with Meze Mini XLR to 4.4 mm Balanced Silver Plated PCUHD Copper Premium Cable​

As can be seen, the distortion profiles and levels between the two cables when volume-matched are largely the same. I am unable to compare measurements for electrical noise rejection.

In subjective sighted comparisons, I personally have never noticed any tonal differences or any sounding "wider" or "more detailed" than the other.

HiFiMan Arya Stealth:

Here, we see a comparison of a stock unbalanced cable with a balanced cable from a different brand.

Figure 1: Frequency response for HiFiMan Arya Stealth with fabric sleeve stock (crystalline silver and/or copper) cable, unbalanced and volume-matched

Figure 2: Frequency response for HiFiMan Arya Stealth with Meze Mono 3.5 mm to 4.4 mm Balanced Silver-Plated Copper Upgrade Cable
​

Again, the frequency responses are virtually identical (within the audio band). Likewise, the phase responses are similar, and as you would see in the full REW measurements, can vary noticeably for the same cable. Likewise, in practical listening, mere head movements will likely induce high-frequency magnitude and phase response differences greater than what a properly designed cable would induce. At least when directly A/Bing an unbalanced cable with a balanced cable without volume-matching (or worst case "volume-matching" until the preferred cable sounds "better"), there should be a clear difference in volume which would obviously have major subjective effects. Otherwise, the only reason I have this cable was because it was what the shop offered when I asked for a balanced cable purely for the purpose of taking advantage of the DAC/amp's power and noise capabilities through its balanced topology.

Figure 3: Distortion measurement for HiFiMan Arya Stealth with fabric sleeve stock (crystalline silver and/or copper) cable, unbalanced and volume-matched

Figure 4: Distortion measurement for HiFiMan Arya Stealth with Meze Mono 3.5 mm to 4.4 mm Balanced Silver-Plated Copper Upgrade Cable​

Again, the distortion profiles between the unbalanced and balanced cables when volume-matched are largely the same.

Regarding cable phase distortion:

WaveTheory's videos did well to explain the fundamentals of sound, and made a reasonable argument on the invalidity of the term "bias" or "placebo" when the signals are in fact measurably different, though I would vie that "confounding variable" is perhaps the term that should be used. The problem is that they predicated this and the entire "human pattern recognition" or "auditory camouflage" argument on velocity factor measurements which they failed to identify [[rather did not describe as for pedagogical reasons]] as causing phase distortions which should not be audible--actually, looking at https://www.audiosciencereview.com/...se-distortion-shift-matter-in-audio-no.24026/, the research referenced was specifically for the audibility within reverberant rooms through speakers, whereby I am not sure of similar research for headphones (Amir does mention headphones), but again, I have shown that merely moving your head or jaw would surely cause phase differences greater than what most cables could induce; then we would need to discuss the science for determining how these could possibly cause large perceptual distortions of magnitude response mind imaging and soundstage if at all. The assumption was that the time domain differences in waveforms from these phase distortions would be processed by the brains as changes in patterns that could be detected through extended listening, but again per https://www.phon.ucl.ac.uk/courses/spsci/AUDL4007/12.pdf, the "pattern" the brain is processing through the nerves of the cochlea is rather more like a spectrogram containing magnitude information excluding phase information. Now, perhaps there are mechanisms through which phase distortions do cause the cochlea to report alterations to this "spectrogram" (e.g. supposed hair cell sensitivity to phase inversion), but again, as shown, head movements or ear twitches on their own may induce these phase differences.

Phase response of some other systems:

One interesting thing. Behold the Sennheiser HE-1's ultra-smooth frequency response (measured during a Bay Bloor Radio listening event), but substantially decreasing phase response (wrapped) which IIRC implies that it isn't minimum-phase:

Figure 9: Sennheiser HE-1 magnitude and phase with respect to frequency​

Perhaps this monotonically downward slope may be related to the HE-1 as an electrostatic headphone being a capacitive load, or phase distortions inherent to electrostatic headphone amplifiers. Subjectively, I would argue that I only "heard" the tonality and that I preferred the sound and presentation of my EQing results from https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152).

Finally, here is my left in-ear measurement while 1.5 m in front of a GLM-calibrated Genelec 8341A (very neutral and excellent directivity active studio monitor) in my backyard with my head directed 30 degrees to the left (left ear thus facing away from the speaker):

Figure 10: Left ear HRTF when facing 30 degrees left​

This by chance shows a phase profile much more similar to my headphone measurements while the others show more phase "extension" and peaks past 7 kHz, some measurements from the other channel or configurations showing more greatly decreasing phase profiles.

Conclusion:

It has been shown that at least these cables for these mostly flat-impedance headphones have effectively identical magnitude frequency responses, minor differences being attributable to minor movements of the headphones when one moves one's head or switches the cables, the treble phase response appearing to be the most greatly affected by these slight perturbations. Furthermore, these motional perturbations cause greater changes for the same cable than the cables themselves possibly could within the audio band. Likewise, when volume-matched, no appreciable differences were measured between a stock unbalanced cable and a slightly more premium balanced cable from a different brand.

Regardless, the point is that though individuals who report differences do perceive them, assessment of causation must not forget that the system includes both the knowledge of the cable and the actual signals reaching one's ears, whereby in the equation of ab + c = d, when b (knowledge of the cable) and d (subjective sound) change, and coefficient a ("susceptibility") for the individual happens to be large enough, so long as those changes for them are consistent in their listening environment, the individual need not be offended by the notion that c (the audio signal) is in fact demonstrably unchanging or that in the grand scheme of the '+' operator of human hearing, such tiny changes in c are inaudible. Otherwise, perhaps indeed, companies had ought to be held accountable when they claim changes in c that demonstrably do not exist in the audio band.

Given this, I will continue to advocate that if you want to achieve both clearly measurable and reliably audible changes and improvements, this can be best achieved with EQ for which the only cost is the time spent on learning how to use it effectively. But this requires lifting any bias that EQ degrades audio, whereby as I have shown in https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152) and my other EQ posts, EQ can measurably eliminate peaks, cleaning up the sound, correct tonal balance for excellent and transparent richness, clarity, and naturalness, and in the case of minimum-phase EQ even flatten the phase response.

 

[gregorio]

It has been shown that at least these cables for these mostly flat-impedance headphones have effectively identical magnitude frequency responses …

You’re preaching to the converted here. Differences in cables were measured here years ago and in the science and comms and pro-audio worlds decades earlier. Everyone agrees, the magnitude of differences is well below audibility. The only ones who disagree are the relatively tiny handful of audiophiles suckered by the marketing. There are no audiophile cables in the pro-audio/sound world, the engineers would just laugh at the marketing.

There are some rare exceptions, very specific cases of low and variable impedance IEMs where cable gauge can make an audible difference, although that’s not an argument for audiophile cables, just a slightly bigger gauge than typical in some rare instances.

G

 

[MrHaelscheir]

You’re preaching to the converted here. Differences in cables were measured here years ago and in the science and comms and pro-audio worlds decades earlier. Everyone agrees, the magnitude of differences is well below audibility. The only ones who disagree are the relatively tiny handful of audiophiles suckered by the marketing. There are no audiophile cables in the pro-audio/sound world, the engineers would just laugh at the marketing.

There are some rare exceptions, very specific cases of low and variable impedance IEMs where cable gauge can make an audible difference, although that’s not an argument for audiophile cables, just a slightly bigger gauge than typical in some rare instances.

G

I am aware. The point of posting it here was to host the response to WaveTheory's videos, and should anyone not in the loop (surely we've come across threads on this sub-forum with believers having long debates with the locals) come across this more recent data using "real human ears".

 

[gregorio]

The point of posting it here was to host the response to WaveTheory's videos, and should anyone not in the loop (surely we've come across threads on this sub-forum with believers having long debates with the locals) come across this more recent data using "real human ears".

Yep, I didn’t mean to come across as critical of your post. On the contrary, I “Liked” it. Unfortunately though, objective tests and facts simply don’t work with the “believers”, engineers and others have tried for nigh on half a century.

I have to say, although I’ve only watched a few bits of the WaveTheory’s videos you posted, it’s those sorts of audiophile charlatans I hate the most. They are deliberately perverting the science, lying and misleading the average consumer/audiophile, just for their own avarice.

G

 

[bigshot]

I don't think evidence has any impact on people who are deeply invested in snake oil. Changing their mind involves convincing them to admit they were wrong, and they won't allow themselves to do that. They'll go to their death swearing down is up.

 

[WaveTheory]

Hello. In response to WaveTheory's recent "Cables Matter" videos, claiming such things as cables being able to "smooth out peaky treble" or some cables having innate "peaky treble", and their original four similarly named videos covering the possible science behind the audibility of cable differences, I have gone ahead and taken a few measurements with my in-ear microphones obtained from Earfish (I will be reporting on my findings with binaural headtracking soon), whether or not one sees benefits to the using real human ears. At best, these measurement conditions should reveal practical frequency response variations caused by the mere act of slightly moving your head.

In short, the "cable science" videos explain fundamentals reasonably, making very clear to all of us what sound is (unfortunately never mentioning magnitude or phase versus frequency response; take some time to convince yourself that "sound" is a collection of longitudinal air vibrations that can be divided into individual superimposed sine waves arriving with varying phases, and that audio systems alter this in various ways, whereby true transparency aims to leave this frequency content untouched and not introduce any new sine waves that don't belong (nonlinear distortion)), then in the third and fourth parts mainly point out miniscule phase distortions caused by cable velocity factors (without ever mentioning the word "phase") as potentially causing time domain waveform changes detectable by human pattern recognition facilities (I have many issues with this).

This report will proceed to demonstrate the lack of frequency response differences between the measured cables mind between balanced and unbalanced that cannot be attributed to minor movements of the headphones or skin, and that phase variations due to head movements disturbing the headphones are most likely to create significant phase distortions past 1 kHz far surpassing any that could be introduced by cables, such not considering literature on the inaudibility of phase distortions in most audio.

Disclaimer: Note that I am not discounting the reality of these subjective cable experiences for users, but rather the cause or the belief that these subjective effects are physically innate to the cables rather than consistent perceptual distortions molded by knowledge of what cable one is listening to. As I had discussed at length in https://audiosciencereview.com/foru...rent-are-that-many-confused.9245/post-1522578 (post #5,428), if those subjective effects are consistent to you within your practical, sighted listening conditions, then all power to you to enjoy your gear, but just know that there is currently no science that should suggest that these effects are remotely guaranteed to be perceived by others and that such investments are more effective than properly done EQ and other digital processing.

Methodology:

In https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152 and the following), you will see my acquisition of a pair of in-ear microphones from Earfish and EQed frequency response, phase, and distortion measurements taken for the Meze Elite and Arya Stealth. Below is the full measurement methodology for this report using that equipment:

• I am seated on my computer chair before my computer desk and measurement equipment. The ambient noise level was between 35 to 40 dBA.
• The in-ear microphones are inserted as deeply as possible with the capsules facing the ear canal entrance such that the ear canal resonance is excluded from the frequency response measurement; the memory foam mounts also protect the eardrums from the loud volumes needed for high-SNR measurements.
• The included Axagon USB HQ Audio Mini Adapter is mounted to the back of a headband (meant for HRTF measurements), the in-ear microphones wires being held taut between the headband and the head to prevent the seating of the in-ear microphones from being disturbed during and between measurements. The wires held or pulled up so that they are pressed only by the upper part of the headphone earpads.
• Room EQ Wizard (REW) is used to facilitate the frequency response and distortion measurements, the Axagon adapter and FiiO K9 Pro ESS DAC/amp being selected as the input and output devices in non-exclusive mode, all EQ applications being disabled. The Windows Mixer volume settings for both devices are set to 100%.
• The DAC/amp is set to low gain.
• The cables are connected directly to the DAC/amp's respective outputs with all other ports unoccupied.
• The "Measure" button on the upper left of REW is clicked, the measurement length (in samples) set to 4M to maximum single-repetition noise reduction, and the sample rate set to 192 kHz purely to reduce the time duration of the measurement.
• "Check levels" in the "Measure" window is clicked and played for each channel to facilitate volume matching such that switching between measurements in REW best shows the identicality.
• "Start" in the "Measure" window is clicked and my right hand placed on top of the DAC/amp as I have found that this somehow reduces the noise floor by around 10 dBFS in the "Check levels" sound level log on the lower left such that vibrations from my heartbeat are visible. I keep my head and body still for the duration of the measurement with light, regular breathing, avoiding twitching my ears.
• The "Controls" button on the upper right of the measurement viewing window is clicked and "Unwrap Phase" clicked, the phase graph then positioned with the +-360 buttons for convenient viewing (these settings may or not be carried over to your own instance of REW, and you might also have to click "Limits" on the upper right and "Apply Settings" for the default to obtain the intended viewing scale).
• Cables are switched with the headphones still on the head, trying to disturb their position as minimally as possible.

Observations:

The full REW .mdat files containing the measurements can be found at https://drive.google.com/file/d/1XG3cA9byn0p8BHBqYiOx4wF8s5DS8-KV/view?usp=share_link (the file size is too large to attach here) with comments explaining each measurement.

Only left channel measurement images are shared in this post since the right in-ear microphone was found to have bass and top octave roll-off compared to the left in-ear microphone which was found to be more accurate, roughly confirmed against outdoor on-axis measurements of a Genelec 8341A calibrated with GLM.

This report assumes an understanding of how to read magnitude and phase versus frequency response graphs as well as REW distortion graphs, all of which can be learned online. Consider frequency response graphs as summarizing the "wave superposition" described by WaveTheory in their "cable science" videos.

The below graphs show the magnitude versus frequency response on the top and the unwrapped phase versus frequency response on the bottom.

Meze Elite:
Here, we see a comparison between two balanced cables of different tiers within the same brand. The below measurements targeted 105 dB SPL uncalibrated on the "Measure" window's "Check levels" output, corresponding with around 96.5 dBA when placing an A-weighted SPL meter with wind shield lightly against the inner driver grill of the right driver unsealed and facing outward.

Magnitude and phase versus frequency responses:

Figure 1: Frequency response for Meze Elite hybrid pads with Meze OFC 99.95% Copper Standard Cable Balanced 4 pin XLR Cable​

Figure 2: Frequency response for Meze Elite hybrid pads with Meze Mini XLR to 4.4 mm Balanced Silver Plated PCUHD Copper Premium Cable​

At this point, you will have to just trust that I definitely did swap the cables and correctly kept track of the measurements, since the frequency responses are nigh identical (within the 20 Hz to 20 kHz audio band; I would highly doubt that ultrasonic performance matters, especially when your DAC already filters the ultrasonic components out per the Nyquist-Shannon sampling theorem). When switching between the premium and stock measurements for the same channel when viewing with REW, you will see minor variations of the sort you would see with the same cable when slightly disturbing the headphone's or in-ear microphones' positioning. So how are you to know if the differences that you hear are due to the cable or due to the headphones being positioned differently, especially in the case that you remove the headphones first to change the cable? Regardless, what is clear is that the 99.95% copper cable certainly does not (measurably) smooth out treble peaks mind the 6 kHz peak, and neither does the act of plating the copper with silver increase the brightness, both effects being achievable for free with EQ. The only reason I own this premium upgrade cable is that my preferred 4.4 mm termination was not available for the stock cable, and that it feels nice and is a joy to look at every day as is the Tungsten version of the headphone itself.

The stock cable's measured magnitude versus frequency response was maybe a fraction of a dB lower than its premium cable counterpart's, though whatever factors in the cable or DAC/amp may have contributed to this, I doubt that this would be enough to cause perception of massive increases in "soundstage" or "resolution" among other claimed phenomena.

You may notice that the phase responses past 1 kHz show quite noticeable differences (not considering the literature regarding the inaudibility of phase distortions for certain audio; in short, the cochlea performs the equivalent of a Fourier transform such that neurons are only excited by individual narrow frequency bands, your brain this primarily hearing the magnitude response sans phase information; see https://www.phon.ucl.ac.uk/courses/spsci/AUDL4007/12.pdf which explains the entire auditory path), which surprised me at first, though again, I thought it common knowledge that these are largely inaudible in these conditions (except maybe for the quality and decay of sharp transients). After conducting more measurements as included and annotated in the full REW measurements, I found that even for the same cable while sitting as still as possible, small perturbations from my unavoidable slight head movements or ear twitches between measurements were likely the cause of these significant changes in phase response past 1 kHz, unless it is rather an error of my measurement system. Otherwise, when looking through the various measurements, it could be figured that any differences in phase response between the two cables were comparable to the differences that could be measured from the same cable due to head perturbations, there also being some measurements where the phase response for both cables are quite similar. Thus, there is likely no significant difference in phase response between these cables.

Distortion comparison:

Figure 3: Distortion measurement for Meze Elite hybrid pads with Meze OFC 99.95% Copper Standard Cable Balanced 4 pin XLR Cable​

Figure 3: Distortion measurement for Meze Elite hybrid pads with Meze Mini XLR to 4.4 mm Balanced Silver Plated PCUHD Copper Premium Cable​

As can be seen, the distortion profiles and levels between the two cables when volume-matched are largely the same. I am unable to compare measurements for electrical noise rejection.

In subjective sighted comparisons, I personally have never noticed any tonal differences or any sounding "wider" or "more detailed" than the other.

HiFiMan Arya Stealth:

Here, we see a comparison of a stock unbalanced cable with a balanced cable from a different brand.

Figure 1: Frequency response for HiFiMan Arya Stealth with fabric sleeve stock (crystalline silver and/or copper) cable, unbalanced and volume-matched


Figure 2: Frequency response for HiFiMan Arya Stealth with Meze Mono 3.5 mm to 4.4 mm Balanced Silver-Plated Copper Upgrade Cable
​

Again, the frequency responses are virtually identical (within the audio band). Likewise, the phase responses are similar, and as you would see in the full REW measurements, can vary noticeably for the same cable. Likewise, in practical listening, mere head movements will likely induce high-frequency magnitude and phase response differences greater than what a properly designed cable would induce. At least when directly A/Bing an unbalanced cable with a balanced cable, there should be a clear difference in volume which would obviously had major subjective effects. Otherwise, the only reason I have this cable was because it was what the shop offered when I asked for a balanced cable purely for the purpose of taking advantage of the DAC/amp's power and noise capabilities through its balanced topology.

Figure 3: Distortion measurement for HiFiMan Arya Stealth with fabric sleeve stock (crystalline silver and/or copper) cable, unbalanced and volume-matched


Figure 4: Distortion measurement for HiFiMan Arya Stealth with Meze Mono 3.5 mm to 4.4 mm Balanced Silver-Plated Copper Upgrade Cable​

Again, the distortion profiles between the unbalanced and balanced cables when volume-matched are largely the same.

Regarding cable phase distortion:
WaveTheory's videos did well to explain the fundamentals of sound, and made a reasonable argument on the invalidity of the term "bias" or "placebo" when the signals are in fact measurably different, though I would vie that "confounding variable" is perhaps the term that should be used. The problem is that they predicated this and the entire "human pattern recognition" or "auditory camouflage" argument on velocity factor measurements which they failed to identify as causing phase distortions which should not be audible--actually, looking at https://www.audiosciencereview.com/...se-distortion-shift-matter-in-audio-no.24026/, the research referenced was specifically for the audibility within reverberant rooms through speakers, whereby I am not sure of similar research for headphones (Amir does mention headphones), but again, I have shown that merely moving your head or jaw would surely cause phase differences greater than what most cables could induce; then we would need to discuss the science for determining how these could possibly cause large perceptual distortions of magnitude response mind imaging and soundstage if at all. The assumption was that the time domain differences in waveforms from these phase distortions would be processed by the brains as changes in patterns that could be detected through extended listening, but again per https://www.phon.ucl.ac.uk/courses/spsci/AUDL4007/12.pdf, the "pattern" the brain is processing through the nerves of the cochlea is rather more like a spectrogram containing magnitude information excluding phase information. Now, perhaps there are mechanisms through which phase distortions do cause the cochlea to report alterations to this "spectrogram" (e.g. supposed hair cell sensitivity to phase inversion), but again, as shown, head movements or ear twitches on their own may induce these phase differences.

Phase response of some other systems:

One interesting thing. Behold the Sennheiser HE-1's ultra-smooth frequency response (measured during a Bay Bloor Radio listening event), but substantially decreasing phase response (wrapped) which IIRC implies that it isn't minimum-phase:

Figure 9: Sennheiser HE-1 magnitude and phase with respect to frequency​

Perhaps this monotonically downward slope may be related to the HE-1 as an electrostatic headphone being a capacitive load, or phase distortions inherent to electrostatic headphone amplifiers. Subjectively, I would argue that I only "heard" the tonality and that I preferred the sound and presentation of my EQing results from https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152).

Finally, here is my left in-ear measurement while 1.5 m in front of a GLM-calibrated Genelec 8341A (very neutral and excellent directivity active studio monitor) in my backyard with my head directed 30 degrees to the left (left ear thus facing away from the speaker):

Figure 10: Left ear HRTF when facing 30 degrees left​

This by chance shows a phase profile much more similar to my headphone measurements while the others show more phase "extension" and peaks past 7 kHz, some measurements from the other channel or configurations showing more greatly decreasing phase profiles.

Conclusion:
It has been shown that at least these cables for these mostly flat-impedance headphones have effectively identical magnitude frequency responses, minor differences being attributable to minor movements of the headphones when one moves one's head or switches the cables, the treble phase response appearing to be the most greatly affected by these slight perturbations. Furthermore, these motional perturbations cause greater changes for the same cable than the cables themselves possibly could within the audio band. Likewise, when volume-matched, no appreciable differences were measured between a stock unbalanced cable and a slightly more premium balanced cable from a different brand.

Regardless, the point is that though individuals who report differences do perceive them, assessment of causation must not forget that the system includes both the knowledge of the cable and the actual signals reaching one's ears, whereby in the equation of ab + c = d, when b (knowledge of the cable) and d (subjective sound) change, and coefficient a ("susceptibility") for the individual happens to be large enough, so long as those changes for them are consistent in their listening environment, the individual need not be offended by the notion that c (the audio signal) is in fact demonstrably unchanging or that in the grand scheme of the '+' operator of human hearing, such tiny changes in c are inaudible. Otherwise, perhaps indeed, companies had ought to be held accountable when they claim changes in c that demonstrably do not exist in the audio band.

Given this, I will continue to advocate that if you want to achieve both clearly measurable and reliably audible changes and improvements, this can be best achieved with EQ for which the only cost is the time spent on learning how to use it effectively. But this requires lifting any bias that EQ degrades audio, whereby as I have shown in https://www.head-fi.org/threads/mez...eadphone-official-thread.959445/post-17743502 (post #5,152) and my other EQ posts, EQ can measurably eliminate peaks, cleaning up the sound, correct tonal balance for excellent and transparent richness, clarity, and naturalness, and in the case of minimum-phase EQ even flatten the phase response.

Found it! I came looking for this after seeing your comment on YT. Thanks for putting this work in. I appreciate your efforts. I have not had a chance to go over this in detail yet but rest assured I will take the time to analyze your argument and data.

A quick note of response on one objection you bring up... You are 100% correct that I did not use the term phase distortion (and some other terms) in my videos. I didn't see using that verbiage as being helpful. The videos are meant for a lay audience. The underlying concepts are more important than the vocabulary. So that was a pedagogical decision that I made for (hopefully) increased clarity to a non technical audience.

I'll be back when I've given your work it's due. Thanks for doing this and thanks for letting me know about it.

 

[gregorio]

The videos are meant for a lay audience. The underlying concepts are more important than the vocabulary.

And that is what is so objectionable about the videos!!

Clearly you are not presenting yourself as a lay person, in fact quite the opposite, as an extremely highly qualified (PhD) scientist. Although I frequently skipped ahead, the vast majority of what you stated is entirely accurate and indicates at least a good understanding of the basics of what science is and the physics of sound/signals. Unfortunately though, there is also a fundamental falsehood/omission, upon which nearly all of your part 4 video relies and upon which your conclusions entirely rely! You’ve apparently been “looking and looking and looking” for data and apparently missed almost an entire field of science, the field that actually deals with sound perception (!) and a great wealth of data going back over a century. Strangely enough you do actually mention some data you’ve found in the field of psychoacoustics, EG. A JND (just noticeable difference) but then falsely state there’s no data suggesting “how much a pattern has to change before we can perceive a difference in the pattern”, which is not only completely false but is even a self contradiction. What do you think a JND is, and how do you think they’re tested? Have you really never heard of DBT’s or seen any of the countless thousands of published studies employing them?

Your assertions in this regard indicates either astonishing ignorance or you do actually know but are deliberately misinforming/misleading your “lay audience”. If it’s the former, you’re violating your own definition of science, because you are not only ignoring but actually contradicting a very significant “body of evidence”, which would be almost unbelievable for someone claiming to actually be a scientist (with a PhD) and if it’s the latter, that’s unforgivable!

So, in answer to your question (upon which your conclusion entirely relies), “How do you know these electrical differences are too tiny to hear?” - The answer is: A great wealth of many decades of hard, reliable engineering and scientific evidence, a lack of any reliable evidence whatsoever to the contrary and every single time a cable believer has been tested using a controlled (double blind) test, their claims of being able to hear such differences magically vanish (again going back many decades).

Given the wealth of reliable data/facts, how could any rational person, let alone someone claiming to be a scientist, come to any conclusion other than there is no audible differences between cables? (Obviously this assertion is conditional but I wouldn’t have to explain that to an actual scientist).

G

 

[bigshot]

Audiophiles know every detail of the specs of their equipment, and can recite the numbers by heart. But when it comes to the specs of human ears, they don’t know much at all, and they assume that if we can measure a difference, then ears can probably hear it. They double down on best case scenarios and golden ear ego boosts, pushing the numbers we can supposedly hear out to absurd levels. The truth is that even inexpensive cables, DACs and amps are capable of achieving audible transparency. You would have a hard time finding one that is discernibly different in controlled listening tests. Only the most pathologically specialized stuff, like NOS DACs and non standard impedances might sound different.

 

[MrHaelscheir]

And that is what is so objectionable about the videos!!

Clearly you are not presenting yourself as a lay person, in fact quite the opposite, as an extremely highly qualified (PhD) scientist. Although I frequently skipped ahead, the vast majority of what you stated is entirely accurate and indicates at least a good understanding of the basics of what science is and the physics of sound/signals. Unfortunately though, there is also a fundamental falsehood/omission, upon which nearly all of your part 4 video relies and upon which your conclusions entirely rely! You’ve apparently been “looking and looking and looking” for data and apparently missed almost an entire field of science, the field that actually deals with sound perception (!) and a great wealth of data going back over a century. Strangely enough you do actually mention some data you’ve found in the field of psychoacoustics, EG. A JND (just noticeable difference) but then falsely state there’s no data suggesting “how much a pattern has to change before we can perceive a difference in the pattern”, which is not only completely false but is even a self contradiction. What do you think a JND is, and how do you think they’re tested? Have you really never heard of DBT’s or seen any of the countless thousands of published studies employing them?

Your assertions in this regard indicates either astonishing ignorance or you do actually know but are deliberately misinforming/misleading your “lay audience”. If it’s the former, you’re violating your own definition of science, because you are not only ignoring but actually contradicting a very significant “body of evidence”, which would be almost unbelievable for someone claiming to actually be a scientist (with a PhD) and if it’s the latter, that’s unforgivable!

So, in answer to your question (upon which your conclusion entirely relies), “How do you know these electrical differences are too tiny to hear?” - The answer is: A great wealth of many decades of hard, reliable engineering and scientific evidence, a lack of any reliable evidence whatsoever to the contrary and every single time a cable believer has been tested using a controlled (double blind) test, their claims of being able to hear such differences magically vanish (again going back many decades).

Given the wealth of reliable data/facts, how could any rational person, let alone someone claiming to be a scientist, come to any conclusion other than there is no audible differences between cables? (Obviously this assertion is conditional but I wouldn’t have to explain that to an actual scientist).

G

I understand the frustration, but I would still lean in favour of respectfulness and Hanlon's razor in this case with his actually bothering to seek out my post and if one assumes truth to the narrative of his early on having not expected to hear differences in his gear, his then doing his best within his respective fields to seek out the possible science behind his experiences, being only human and unfortunately heading down the wrong path to that truth. I feel that excessive confrontation rather than respectful explanation of and directing to facts has only pushed people away. At best, per my original post, the goal is to respect these individuals' experiences which in the "ab + c = d" analogy do exist for a > 0 (sighted test), our then striving to convince them that there is nothing wrong with c not actually encountering appreciable changes and hence d not changing when a = 0 (DBT). Science aims for reproducibility, whereby though this discovery regarding c needn't detract from their continued enjoyment of their gear in sighted conditions, there should be understanding as to why there is no guarantee of the reproducibility of their experiences with others and how there are more objectively effective ways to achieve the same effects.

 

[bigshot]

The lesson isn’t the results of your test, it’s the fact that they’re based on a test. There are plenty of resources for information besides subjective impressions in service of high end audio dealers. The first place he should start is joining the AES where he can do some due diligence studying up on what’s audible and what isn’t.

 

[gregorio]

I understand the frustration, but I would still lean in favour of respectfulness and Hanlon's razor in this case with his actually bothering to seek out my post and if one assumes truth to the narrative of his early on having not expected to hear differences in his gear, his then doing his best within his respective fields to seek out the possible science behind his experiences, being only human and unfortunately heading down the wrong path to that truth.

I agree we should lean in favour of respectfulness, and in daily life I do. Here though, with audiophiles, the only difference being respectful ever seems to make is just to delay the inevitable point at which the questioning of their unquestionable belief in audiophile myth causes them to lash out.

Hanlon’s Razor is certainly a default position generally but in this case, can we really assume stupidity or ignorance of the most basic science/scientific principles in someone who not only claims a double PhD but specific expertise in neuroscience/cognition? Incredulously, he actually employs probably the most stereotypical of all audiophile fallacies: “Even my wife heard it”, who apparently is also a scientist with a double PhD but also doesn’t know even the most basic scientific procedures that any degree undergrad would be failed for not knowing. It’s virtually impossible to conclude anything other than he‘s deliberately lying, either about his qualifications/expertise, about his conclusion or both.

G

 

[WaveTheory]

Hello. In response to WaveTheory's recent "Cables Matter" videos, claiming such things as cables being able to "smooth out peaky treble" or some cables having innate "peaky treble", and their original four similarly named videos covering the possible science behind the audibility of cable differences, I have gone ahead and taken a few measurements with my in-ear microphones obtained from Earfish (I will be reporting on my findings with binaural headtracking soon), whether or not one sees benefits to the using of real human ears. At best, these measurement conditions should reveal practical frequency response variations caused by the mere act of slightly moving your head or jaw....

Poking around on the REW website a bit, I haven't been able to confirm how the REW software suite performs the frequency-response test. I assume it is similar to how the FR sweep tests from miniDSP and others works, but I want to confirm. Could you give me a description, as step by step as possible, how exactly the magnitude portion of the FR sweep test is performed? I especially want to focus on what is actually being measured.

@gregorio, clearly you are frustrated and it's also true I didn't lead with a lit review on my YT series. Again, pedagogical decision for the bulk of what I perceive the YT audience to be. We can agree to disagree on if that was the right move. Regardless, be careful. Before we start attacking personal and professional integrity and/or qualifications we should make sure we are being as thorough as we are asking anyone else to be. My read on the state of the literature around audibility of phase distortions is summed up well by Moller et. al 2007 (https://vbn.aau.dk/ws/files/227876517/2007_M_ller_et_al_AES_Journal.pdf):

"[Sect. 0.3] The human sensitivity to phase has been the topic of many studies, and the question has generated heated discussions in the professional community and among enthusiastic audio amateurs. Helmholtz [2, pp. 124–127] is often quoted as claiming that the ear is insensitive to phase. It is true that this was his main conclusion, but he carefully pointed out limitations in his experiments, and later studies have shown that there are audible effects connected to phase aspects of signals. Some of the effects are subtle and occur only with certain signals and under certain well controlled and ideal listening conditions. On the other hand, there is no doubt that audible effects of phase do exist—they are real and repeatable, and in some cases they are clearly audible to everyone."

The paragraph immediately following the quoted (the full pdf is available in the link I provided) includes numerous citations defending those claims.

The above indicates that the question of phase distortion audibility is anything but settled. Furthermore, additional papers and AES conference presentations provide more published results showing that there are in fact cases where phase distortions or corrections to phase distortions are audible. A sampling:

1. Greenfield & Hawksford (AES Convention proceedings paper, 1990): https://www.researchgate.net/profil...udibility-of-Loudspeaker-Phase-Distortion.pdf
2. Schgor (2022) - also links phase audibility to ear anatomy and auditory system physiology https://www.politesi.polimi.it/bitstream/10589/186927/6/Tesi_Schgor.pdf
3. Johansen (2006) - "The human ear is not only a frequency analyser but is also sensitive to the temporal structure, e.g. periodic events." https://citeseerx.ist.psu.edu/docum...&doi=2c239f31e59616e9162a86a3409054989ba89963

I'll also drop a quote from Floyd Toole in response to a well done lit review (as of Spring 2005) done by Audioholics writer Mark Sanfilipo (https://www.audioholics.com/room-acoustics/human-hearing-phase-distortion-audibility-part-2):

"As was mentioned, our sensitivity to the timbral changes was very much dependent on the Q, or bandwidth of the phenomenon - with much lower 'thresholds' being found for wider bandwidth spectral changes. Interesting, because the narrower the bandwidth, the more the resonant system rings. Except, perhaps, at very low frequencies, it seems that we do not exhibit a primary response to the time-domain (ringing) problems but instead to the spectral features."

There are some big words in there but what Toole is essentially saying is that human hearing becomes more sensitive to the timbral effects of phase distortions as the range of frequencies present in the signal increases. ie, the larger the number of hertz between the lowest and highest pitched sounds becomes, the easier it is for us to hear if things start shifting around in the time domain (as is the case with phase distortions). If you go to the website and read the full quote you'll see that Toole names music genres in which it's easier to hear these effects than it is in others. The genres he names frequently have very wide and complex frequency spectra.

There are more, but these citations and the citations they include should make it pretty clear that random-Wikipedia-editor (as of 25 Oct 2023) nailed it:

https://en.wikipedia.org/wiki/Phase...tortion,typical sound systems remains debated. - not bad for a wiki statement.

I will circle back to the issues at hand in this thread. Please allow me some time. These are big concepts that take time to assemble into something resembling coherence. I'm a full time professor, have a family, and am still doing reviews of gear on YT.

More to come. Best to all

 

[bigshot]

The level of phase error in modern home audio equipment falls far below the threshold of audibility. Even if it was borderline present, as I understand it, it affects the very highest range of frequencies first. That is the least important part of the audible spectrum. You might have been able to hear it in cheap graphic equalizers back in the 1970s (along with spill and grounding noise), but not today. It isn't worth worrying about.

Too much of audiophilia nervosa involves long debates over stuff that just doesn't matter in everyday use. I imagine some people love endless discussions of "what if", but I find them tedious. I just want practical advice on how to make my home audio system sound better. I don't care about stuff I flat out can't hear.

 

[WaveTheory]

The level of phase error in modern home audio equipment falls far below the threshold of audibility. You might have been able to hear it in cheap graphic equalizers back in the 1970s, but not today. It isn't worth worrying about.

That's a big claim to make in light of the lack of well defined thresholds for phase audibility in various contexts in the literature combined with the staggering range of music on this planet and the nigh limitless combinations of gear that can be assembled to play it back. Can we really guarantee that in all of those combinations that we'll never hit a situation where the effects are audible? As a trained scientist, I am NOT comfortable making a claim like yours. We do not know enough to make such a sweeping statement about all of recorded music and audio gear regardless of gear age.

I'll also postulate that the question of "isn't worth worrying about" is a very subjective term. If you're not worried about it, great. But there is enough uncertainty out there combined with enough scientific evidence to suggest that there may very well be things to worry about for some.

 

[bigshot]

Can you point to a particular piece of equipment that has audible levels of phase error? It's best to identify a problem before you start thinking up ways to fix it. Audiophile nonsense is all about creating solutions to problems that have never been observed and documented and probably don't even exist. There's an old saying, "If it ain't broke, don't fix it."

I think you've skipped the observation step and jumped straight to theorizing. When you do that, you unmoor yourself from practical application and float into Brobnigagian "what ifs". This is sound science, not theoretical science. We have a physical goal in mind- high fidelity sound for the purposes of home entertainment. The things that matter are the ones that affect the sound of a CD of Mozart playing on our home stereo, not numbers on a page. It's a lot smarter to address observable problems first. There are plenty of those that don't get discussed much.

The plain truth is that modern digital home audio components have solved many of the problems to a degree where they have been pushed below the threshold of audibility. But home audio still isn't perfect. There are still ways to significantly improve the sound of music in the home... and that isn't just pushing the decimal points further and further back.