A Spectrogram is a visual representation of the the raw waveform audio.
If you recorded the same song playing back on an amp with Capacitor's A's installed vs Capacitor's B installed with a mic into a program like Audacity you could analyze or compare the the audio in visual form. This would make it 100% indisputable if there is or is not a difference in the tones/peaks of the waveform.
If you recorded the same song playing back on an amp with Capacitor's A's installed vs Capacitor's B installed with a mic into a program like Audacity you could analyze or compare the the audio in visual form. This would make it 100% indisputable if there is or is not a difference in the tones/peaks of the waveform.
"In the spectrogram view, the vertical axis displays frequency in Hertz, the horizontal axis represents time (just like the waveform display), and amplitude is represented by brightness."
https://www.izotope.com/en/learn/understanding-spectrograms.html#:~:text=In the spectrogram view, the,wave moving up in pitch.
So, what spectrograms don't show is the phase of each frequency. It turns out humans can hear phase to some extent and that extent is called the "Threshold of Audibility for Group Delay."
So, one thing that humans can hear is not shown in a spectrogram. Whatever made you think it shows everything?
Also, a spectrogram is a rather rough measurement. It is uncontroversial that under some conditions humans can hear very low level sounds, like at or around 1 part per 64,000 (or less). You can't tell a difference in brightness to one part in 64,000 looking at a dot on a spectrogram.
https://www.izotope.com/en/learn/understanding-spectrograms.html#:~:text=In the spectrogram view, the,wave moving up in pitch.
So, what spectrograms don't show is the phase of each frequency. It turns out humans can hear phase to some extent and that extent is called the "Threshold of Audibility for Group Delay."
So, one thing that humans can hear is not shown in a spectrogram. Whatever made you think it shows everything?
Also, a spectrogram is a rather rough measurement. It is uncontroversial that under some conditions humans can hear very low level sounds, like at or around 1 part per 64,000 (or less). You can't tell a difference in brightness to one part in 64,000 looking at a dot on a spectrogram.
Last edited:
Any audible differnces to do with capacitors I've read about claim "this one is warmer" and or "this one is brighter" and there for the tones will visually differ in a spectrogram.
As for your comment any sort of delay that is audible would be visually seen. (For example I have a # of IEM's here using various tech, some of them have a much faster recovery rate than others, meaning the diaphragm inside returns faster in some than others. What this affects is the "muddyness" or "warble" where the diaphragm is still recovering from generating one tone when the next tone is played this causes a fluttering/distorting on the new sound. The faster recovering diaphragms have a much clearer more precise sound.
Also this concept would be akin to using tight foam surround on a woofer vs loose surrounds. The tighter the foam the faster the recovery time (also too tight and you loose certain frequncies/tones) But even though these differences are slight they are audible and there for recordable and thus can be seen visually in an analyzer. I have some electrostatic IEM's as well which are so accurate there is no fluttering/muddyness whatsoever because all tones are generated by electrical pulses and not a physical moving diaphragm. The end result is they sound "transparent" as if there is no speaker/woofer... instead they sound more like you are just hearing the samples played inside your head in real time. In some instances these are so accurate that I can hear slight warbling of the encoding patterns of very high quality encoded audio sources - which I am not able to hear on any other form of speaker or headphone. But your claim about 64,000 cycles is going to be impossible to hear a slice of audio between this gap when 64,000 cycles per second is occuring. Your brain simply cannot process this small of a gap of time. However when playing back digital many soruces (keyboards etc) use samples that are encoded or digitally generated which can introduce audible algorythms (patterns heard in the tones).
Anything audibly heard difference wise (including delay) will be detectable by analyzing a recording of it visually, anything else "not detectable" would be internal processing of the mind. And altering capacitors is not going to alter how we might add a delay to processing certain tones. Tones are tones externally. The only phase change of the tones would be in your own perceptions.
As for your comment any sort of delay that is audible would be visually seen. (For example I have a # of IEM's here using various tech, some of them have a much faster recovery rate than others, meaning the diaphragm inside returns faster in some than others. What this affects is the "muddyness" or "warble" where the diaphragm is still recovering from generating one tone when the next tone is played this causes a fluttering/distorting on the new sound. The faster recovering diaphragms have a much clearer more precise sound.
Also this concept would be akin to using tight foam surround on a woofer vs loose surrounds. The tighter the foam the faster the recovery time (also too tight and you loose certain frequncies/tones) But even though these differences are slight they are audible and there for recordable and thus can be seen visually in an analyzer. I have some electrostatic IEM's as well which are so accurate there is no fluttering/muddyness whatsoever because all tones are generated by electrical pulses and not a physical moving diaphragm. The end result is they sound "transparent" as if there is no speaker/woofer... instead they sound more like you are just hearing the samples played inside your head in real time. In some instances these are so accurate that I can hear slight warbling of the encoding patterns of very high quality encoded audio sources - which I am not able to hear on any other form of speaker or headphone. But your claim about 64,000 cycles is going to be impossible to hear a slice of audio between this gap when 64,000 cycles per second is occuring. Your brain simply cannot process this small of a gap of time. However when playing back digital many soruces (keyboards etc) use samples that are encoded or digitally generated which can introduce audible algorythms (patterns heard in the tones).
Anything audibly heard difference wise (including delay) will be detectable by analyzing a recording of it visually, anything else "not detectable" would be internal processing of the mind. And altering capacitors is not going to alter how we might add a delay to processing certain tones. Tones are tones externally. The only phase change of the tones would be in your own perceptions.
Do you know what group delay is? Doesn't sound like it. May I ask about your technical background? Are you an EE?
What does any of that matter to real world? The men who created the manual's you read from had no background or credentials - they were just men who made observations and wrote it all down.
A dog whistle makes a sound that you can not hear... yet it's a recordable tone that can be seen on a visual spectrum. There for if a spectrum can see audio beyond your hearing capabilities - then whatever "group delay" is will be detectable on a spectral image IF it is indeed audible.
In audio if there is any audible difference then it WILL be detectable, as technology has the ability to record and analyze frequencies and tones far beyond your ear's ability. To claim otherwise is pure ego.
A dog whistle makes a sound that you can not hear... yet it's a recordable tone that can be seen on a visual spectrum. There for if a spectrum can see audio beyond your hearing capabilities - then whatever "group delay" is will be detectable on a spectral image IF it is indeed audible.
In audio if there is any audible difference then it WILL be detectable, as technology has the ability to record and analyze frequencies and tones far beyond your ear's ability. To claim otherwise is pure ego.
B.S. You don't know what you are talking about. Seriously. Claims like the above which you are making have no basis in science. What I want to know is where did you get these ideas, did you think them up yourself? I mean you don't even seem to understand the difference between frequency and phase.
Last edited:
It seems like you are trying to claim that an audible sound that is an "analog vibration" which is detectable and recordable and measurable... can make tones/frequencies that are not measurable or detectable other than in ones own ears? That is non-sense.
Sound tones are sound tones they either differ or they don't.
And if you are trying to claim that plotting sounds through a DSP or digital source vrs an analog source are going to cause an audible difference then yes at LOW resolutions sure. But in this day and age a recording spectrum can be literally scaled to more and more precision based on bit rate and sample rates. Any variations between these "points" on a "digital waveform" are not going to make much difference if any at all... as we are talking the distances between the plotted points are going to be so minuscule there is so little room for the line to curve between said points.... at such high resolutions it will be nearly impossible to detect audibly, however it could still be detectable with recording and analyzing it visually by using a higher bit rates/sample rates than the source. So to see the missing tones in a 16Bit 44Khz sample you could record it at 24Bit 96Khz to see all the details. Recording 16bit 44Khz at 16bit 44khz you will be limited by the resolution and won't be able to see beyond it. There for if you are claiming recording digitally to view the audible differnces between capactors tones is not possible because the resoultion is not precise enough... then all you need to do is increase the resultion and sample rate higher than the source to detect the differnces.
(Besides most sources like a digital keyboard or drum machine or sampler use digitally generated tones... so we are limited to the resolutions of these sources in the first place.)
Sound tones are sound tones they either differ or they don't.
And if you are trying to claim that plotting sounds through a DSP or digital source vrs an analog source are going to cause an audible difference then yes at LOW resolutions sure. But in this day and age a recording spectrum can be literally scaled to more and more precision based on bit rate and sample rates. Any variations between these "points" on a "digital waveform" are not going to make much difference if any at all... as we are talking the distances between the plotted points are going to be so minuscule there is so little room for the line to curve between said points.... at such high resolutions it will be nearly impossible to detect audibly, however it could still be detectable with recording and analyzing it visually by using a higher bit rates/sample rates than the source. So to see the missing tones in a 16Bit 44Khz sample you could record it at 24Bit 96Khz to see all the details. Recording 16bit 44Khz at 16bit 44khz you will be limited by the resolution and won't be able to see beyond it. There for if you are claiming recording digitally to view the audible differnces between capactors tones is not possible because the resoultion is not precise enough... then all you need to do is increase the resultion and sample rate higher than the source to detect the differnces.
(Besides most sources like a digital keyboard or drum machine or sampler use digitally generated tones... so we are limited to the resolutions of these sources in the first place.)
What is the difference between frequency and phase? Do you know? If you don't know that, and especially if you don't want to know about it, then you may not find much value in being around here. Virtually everyone else here will know more than you.
There is nothing wrong with being a newbie if you are here to learn. However if you come here with a chip on your shoulder thinking you already know it all, and yet you are generally ignorant about very basic audio technicalities, then you probably shouldn't expect people to take you very seriously.
There is nothing wrong with being a newbie if you are here to learn. However if you come here with a chip on your shoulder thinking you already know it all, and yet you are generally ignorant about very basic audio technicalities, then you probably shouldn't expect people to take you very seriously.
Last edited:
I don't care to be taken seriously or not... it seems you did not even read my comments properly, instead you are fixated on egotistical belief's over "degrees" as having more value than logical statements - which is non-sense.
Again I'll state the obvious logic.
If there is an audible difference then it WILL BE detectable on a recording. This is because in this day & age our recording technology has the ability to see/analyze frequencies, and subtle differences so minute it far surpasses your ability to detect it audibly. (such as a dog whistle).
But again your trying to claim just because there is "something there" that you can detect it. Which is like saying you can hear a dog whistle because it's "there"... but only a machine can "hear and see" such a frequencies... there for it surpasses human ability to detect.
The limiting factors here that alter "what you hear" is TIME... and SCALE you simply are not able to process such small slices of either. Human hearing is fixed to this current scale and time. The gaps between points on a plotted digital sample are on such a small scale you're not going to hear it because at this scale the length or gap is so minute that the speaker or headphone itself will counter it when it creates an analog waveform in the air. Speakers are just not that precise... to respond at this sort of resolution. But a speaker moves in an analog (in and out) it is not a digital precise instrument. And yes I understand you might try to claim it's accumulative... these slight variations alter the overall "tone" but if it does THEN this will be detectable since our recording tech now FAR surpasses our own ability to hear/detect it.... again because of SCALE.
A digital recording of an analog waveform can be plotted so precise at this point it is litterally "dot to dot" meaning there is no "dot -line- dot" where a straight line is drawn to fill in the gaps of an analog curve. Also the slight curves lost at this level are not going to be heard on analog speakers and headphones which are limited to their motion/movement and SPEED based on this pyshical time and space. Again these slight differnces are going to be limited by scale of time, down to millionths of a second... speakers are not going to be able to recreate such subtle diffrences because of the physical time/scale they are limited to.... yet we can use a computer to zoom into sections of audible time we cannot detect. But again speakers are limited to how precise they can be, based on the physcail speed of which electricty can flow, magnetic fields, and the speed of air compression etc. No matter how much you might "detect" a diffrence with "the math theory" it's not going to be aubdible at this TIME/SCALE we are limited to. The ONLY way to hear such minute detail at the levels you speak of would be if you could slow time down and alter the speed of vibrations and motion etc.
If you wish to continue to disagree then you have a problem.
Again I'll state the obvious logic.
If there is an audible difference then it WILL BE detectable on a recording. This is because in this day & age our recording technology has the ability to see/analyze frequencies, and subtle differences so minute it far surpasses your ability to detect it audibly. (such as a dog whistle).
But again your trying to claim just because there is "something there" that you can detect it. Which is like saying you can hear a dog whistle because it's "there"... but only a machine can "hear and see" such a frequencies... there for it surpasses human ability to detect.
The limiting factors here that alter "what you hear" is TIME... and SCALE you simply are not able to process such small slices of either. Human hearing is fixed to this current scale and time. The gaps between points on a plotted digital sample are on such a small scale you're not going to hear it because at this scale the length or gap is so minute that the speaker or headphone itself will counter it when it creates an analog waveform in the air. Speakers are just not that precise... to respond at this sort of resolution. But a speaker moves in an analog (in and out) it is not a digital precise instrument. And yes I understand you might try to claim it's accumulative... these slight variations alter the overall "tone" but if it does THEN this will be detectable since our recording tech now FAR surpasses our own ability to hear/detect it.... again because of SCALE.
A digital recording of an analog waveform can be plotted so precise at this point it is litterally "dot to dot" meaning there is no "dot -line- dot" where a straight line is drawn to fill in the gaps of an analog curve. Also the slight curves lost at this level are not going to be heard on analog speakers and headphones which are limited to their motion/movement and SPEED based on this pyshical time and space. Again these slight differnces are going to be limited by scale of time, down to millionths of a second... speakers are not going to be able to recreate such subtle diffrences because of the physical time/scale they are limited to.... yet we can use a computer to zoom into sections of audible time we cannot detect. But again speakers are limited to how precise they can be, based on the physcail speed of which electricty can flow, magnetic fields, and the speed of air compression etc. No matter how much you might "detect" a diffrence with "the math theory" it's not going to be aubdible at this TIME/SCALE we are limited to. The ONLY way to hear such minute detail at the levels you speak of would be if you could slow time down and alter the speed of vibrations and motion etc.
If you wish to continue to disagree then you have a problem.
Last edited:
The late physicist Richard Feynman had a term for the sorts of ideas you are proposing: "Cargo Cult Science."
https://www.aier.org/article/do-not-trust-the-cargo-cult-science/
Cargo Cult Science may seem logical to the people doing it, but it is still a misinformed waste of time (or worse).
Since you seem to have no interest in, nor understanding of, the science of audio, I will leave off here. We'll see how other people react to what you have to say.
https://www.aier.org/article/do-not-trust-the-cargo-cult-science/
Cargo Cult Science may seem logical to the people doing it, but it is still a misinformed waste of time (or worse).
Since you seem to have no interest in, nor understanding of, the science of audio, I will leave off here. We'll see how other people react to what you have to say.
Last edited:
It seems you have no interest in understanding the other factors of the "scientific world".
We are subject to the limits of this physical world and so are speakers. This places limits of what you can hear or can't hear regardless of it being mathematically provable. The only way to hear more detail would be if you could slow time down and alter air resistance to be lower... to effectivly speed up the ability of the speaker to create such subtle differnces.
Because again recording tech is able to detect frequencies far beyond what the human ear can hear. And so if there is an audible difference to the human ear (from using different capacitors) - then it MUST be detectable because a recording can scale to levels faaar beyond our ability to process (in this time that our mind is limited to.) Even then as I stated speakers are limited to speed/movement and recovery time based on physical limits. Anything beyond this limit is going beyond the physical limitations of physics. A speaker can only move so fast back and froth that it's not going to be able to "create" such slight details. (Again which would be mathmatically there in theory - but limited to the phyics of time, space, resistance of air etc)
Again a digital recording can be scaled to such precision there are no gaps between plotted points on the waveform (at any physically detectable levels). Sure you can keep scaling in closer and close to be more accurate mathmatically BUT a physical speaker is NEVER going to be capable of being THAT precise (because of air compression rates and speed which the speaker can physically move and recover etc).... and even if it was able to move faster and more accurately your own brain is limited to this scale of time and ability to process this moment as this moment... that is a limit you cannot surpass. The math might surpass these limits... but the limits of the physical world LIMIT how precise the math you can "hear" can be. Again if it's not audibly detectable on a recording at very high resolutions on an analyzer then it's not going to be audible no matter how much you argue. The fact is if it's making an audible differnce then that is something that can be recorded (since recording capabilities far surpass our abilites). If you detect no differnces in the recording and you still believe you hear a difference then that is you imagining it.
Take care.
We are subject to the limits of this physical world and so are speakers. This places limits of what you can hear or can't hear regardless of it being mathematically provable. The only way to hear more detail would be if you could slow time down and alter air resistance to be lower... to effectivly speed up the ability of the speaker to create such subtle differnces.
Because again recording tech is able to detect frequencies far beyond what the human ear can hear. And so if there is an audible difference to the human ear (from using different capacitors) - then it MUST be detectable because a recording can scale to levels faaar beyond our ability to process (in this time that our mind is limited to.) Even then as I stated speakers are limited to speed/movement and recovery time based on physical limits. Anything beyond this limit is going beyond the physical limitations of physics. A speaker can only move so fast back and froth that it's not going to be able to "create" such slight details. (Again which would be mathmatically there in theory - but limited to the phyics of time, space, resistance of air etc)
Again a digital recording can be scaled to such precision there are no gaps between plotted points on the waveform (at any physically detectable levels). Sure you can keep scaling in closer and close to be more accurate mathmatically BUT a physical speaker is NEVER going to be capable of being THAT precise (because of air compression rates and speed which the speaker can physically move and recover etc).... and even if it was able to move faster and more accurately your own brain is limited to this scale of time and ability to process this moment as this moment... that is a limit you cannot surpass. The math might surpass these limits... but the limits of the physical world LIMIT how precise the math you can "hear" can be. Again if it's not audibly detectable on a recording at very high resolutions on an analyzer then it's not going to be audible no matter how much you argue. The fact is if it's making an audible differnce then that is something that can be recorded (since recording capabilities far surpass our abilites). If you detect no differnces in the recording and you still believe you hear a difference then that is you imagining it.
Take care.
The problems we have is how small, possibly not probably measurable effects have on the audio we here and what we are susceptible to, it's not easy or manufacturers would solely focus on measurements and pay no attention to the unnecessary listening.
I've A-B'd many similar components in numerous situations and found demonstrable audio differences, especially capacitors. There is a presumption that the 'science' behind the A-B methodology is flawed, the sub conscious bias plays a part and the human is the problem not the measurements. The problem lies in what to measure and how this correlates to what we hear, what matters and what doesn't. If you look hard enough you will find differences in measurements, but these sometimes matter and sometimes don't.
If the 'proofs' mattered enough and the budgets were significant I am sure we would know more about which specific areas of phase were important for example and which were less so. How dynamic and transient responses were best measured and when making a compromise what to focus on etc, etc.
Here's an obvious one - voice recognition, look how complex this problem has been, lots and lots of money spent here. When we can recognise subtleties in voices in an instant and I've never seen a simple spectrum analyser show this easily, you need a very clever trained expert to know what to look for and even then....
I've A-B'd many similar components in numerous situations and found demonstrable audio differences, especially capacitors. There is a presumption that the 'science' behind the A-B methodology is flawed, the sub conscious bias plays a part and the human is the problem not the measurements. The problem lies in what to measure and how this correlates to what we hear, what matters and what doesn't. If you look hard enough you will find differences in measurements, but these sometimes matter and sometimes don't.
If the 'proofs' mattered enough and the budgets were significant I am sure we would know more about which specific areas of phase were important for example and which were less so. How dynamic and transient responses were best measured and when making a compromise what to focus on etc, etc.
Here's an obvious one - voice recognition, look how complex this problem has been, lots and lots of money spent here. When we can recognise subtleties in voices in an instant and I've never seen a simple spectrum analyser show this easily, you need a very clever trained expert to know what to look for and even then....