• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Master Thread: Are measurements Everything or Nothing?

Awsmone

Member
Joined
Jan 29, 2019
Messages
68
Likes
67
That's where the ego stroking comes in. They believe that sighted listening can be used as a near infinitely sensitive measuring instrument leaving all scientific discoveries in the dust. The only thing that beats it in precision is the positive customer feedback :facepalm:
Yes, the is the same as the philosophy of flat earth, is that all observations have to be validated by the “sighted” observations of the flat earther and reports by others, including experiments are just dogma, and part of a cabal to keep the masses ignorant of the true nature of the world
 

Shadrach

Addicted to Fun and Learning
Joined
Feb 24, 2019
Messages
662
Likes
947
Where is the post? I was commenting on your idea to base your purchases on reviews/measurements it’s will be interesting to read:)
 

captainbeefheart

Senior Member
Joined
Apr 18, 2022
Messages
360
Likes
446
I'm going to be straight up honest that I probably will not read all 127 pages of this discussion so please forgive me if what I say has already been said.

There is one test that is sure way to tell you if two different amps sound the same (even with completely different technologies) and that's a null test. Great part about the null test is you can even use music to do the test to satisfy people that argue we don't listen to sine waves, which of course they are wrong because any waveform you can think of or make is nothing more than the sum of multiple sine waves of different frequencies summed together. Thanks Mr. Fourier. I feel if two different amp pass the null test then they will sound identical. I don't think it's sufficient to say all devices that have a SINAD of say -100db or whatever will all sound the same.

Subjective is when you prefer a particular change in performance, for example amplifiers with rather high distortion, let's accept it's low order harmonic dominant. For purely objective buyers that will not entertain the idea of liking an amplifier knowing it might not test perfect where someone with more of an open mind may not care if it's "perfect" so long as they enjoy the outcome. I personally don't want 30% THD at 35Hz like some tube amps test because although the saturated core of the output transformer creates nonlinear second harmonic distortion and to many sounds "fuller". Combine this with a high output impedance where the resonance at bass frequencies of the speakers are emphasized and mistaken as "big bass" and to me that's just not what I want even if it sounds okay. For low order harmonics I'm totally okay with 1% THD at full power because I know I'm listening to less than 1 watt 99% of the time and if I do require higher power I feel speaker distortion at these higher sound pressure levels will almost certainly be much higher than 1%, try up to 10% or higher at low frequencies. So int he grand scheme of the signal chain the transducer is always going to be the largest offender of distortion. My attitude is what's the point of my amp being .0001% THD when my speakers are going to never come close to this.

As an engineer I need to be objective to meet design goals but then my design goals can be argued as subjective. I'll take a little distortion over instability any day hands down. How many tests do you see of reviewers adding in some capacitance to the dummy load and shooting 10kHz square waves into the amp? None. I'll take an unconditionally stable amp with .5%THD over one that's marginally stable with .001% THD.

I think us engineers need to be a little more open to an amp that may not test perfectly but still sounds great, yes there are such amps. Lots of old amps once had variable damping controls but it's for some reason not popular any more. My guess we just feel over damped is good enough and modern speakers are designed with the notion the source will have very low output impedance. The speaker cable will always quash ridiculous claimed high DF amplifiers, the manufacturers take this measurement from inside the amplifier which should tell you something right there. I like to sense current through the speaker load and add a current feedback loop that is adjustable between positive feedback and negative feedback. I have found many times that amps I sold to people with this function the customer really enjoys being able to control this and they seem to be able to find the sweet spot that just works the best with that particular speaker. They are doing this adjustment by ear so it can be considered subjective since the user doesn't know any data about the changes being made, they just use their ear to find the spot they prefer the best. Same for EQ etc.. we know the response is no longer flat but if the end result is you like the sound then who cares if fidelity to the original recording is lost?
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
I'm going to be straight up honest that I probably will not read all 127 pages of this discussion so please forgive me if what I say has already been said.

There is one test that is sure way to tell you if two different amps sound the same (even with completely different technologies) and that's a null test. Great part about the null test is you can even use music to do the test to satisfy people that argue we don't listen to sine waves, which of course they are wrong because any waveform you can think of or make is nothing more than the sum of multiple sine waves of different frequencies summed together. Thanks Mr. Fourier. I feel if two different amp pass the null test then they will sound identical. I don't think it's sufficient to say all devices that have a SINAD of say -100db or whatever will all sound the same.

Subjective is when you prefer a particular change in performance, for example amplifiers with rather high distortion, let's accept it's low order harmonic dominant. For purely objective buyers that will not entertain the idea of liking an amplifier knowing it might not test perfect where someone with more of an open mind may not care if it's "perfect" so long as they enjoy the outcome. I personally don't want 30% THD at 35Hz like some tube amps test because although the saturated core of the output transformer creates nonlinear second harmonic distortion and to many sounds "fuller". Combine this with a high output impedance where the resonance at bass frequencies of the speakers are emphasized and mistaken as "big bass" and to me that's just not what I want even if it sounds okay. For low order harmonics I'm totally okay with 1% THD at full power because I know I'm listening to less than 1 watt 99% of the time and if I do require higher power I feel speaker distortion at these higher sound pressure levels will almost certainly be much higher than 1%, try up to 10% or higher at low frequencies. So int he grand scheme of the signal chain the transducer is always going to be the largest offender of distortion. My attitude is what's the point of my amp being .0001% THD when my speakers are going to never come close to this.

As an engineer I need to be objective to meet design goals but then my design goals can be argued as subjective. I'll take a little distortion over instability any day hands down. How many tests do you see of reviewers adding in some capacitance to the dummy load and shooting 10kHz square waves into the amp? None. I'll take an unconditionally stable amp with .5%THD over one that's marginally stable with .001% THD.

I think us engineers need to be a little more open to an amp that may not test perfectly but still sounds great, yes there are such amps. Lots of old amps once had variable damping controls but it's for some reason not popular any more. My guess we just feel over damped is good enough and modern speakers are designed with the notion the source will have very low output impedance. The speaker cable will always quash ridiculous claimed high DF amplifiers, the manufacturers take this measurement from inside the amplifier which should tell you something right there. I like to sense current through the speaker load and add a current feedback loop that is adjustable between positive feedback and negative feedback. I have found many times that amps I sold to people with this function the customer really enjoys being able to control this and they seem to be able to find the sweet spot that just works the best with that particular speaker. They are doing this adjustment by ear so it can be considered subjective since the user doesn't know any data about the changes being made, they just use their ear to find the spot they prefer the best. Same for EQ etc.. we know the response is no longer flat but if the end result is you like the sound then who cares if fidelity to the original recording is lost?
Software that might be useful to you if you have not run across it is member Pkane's Deltawave. Free nulling software. He also has Multi-tone you see people using and a few others.

Here is a thread where he first started it.

And here you can download it and his other software.
 

captainbeefheart

Senior Member
Joined
Apr 18, 2022
Messages
360
Likes
446
Software that might be useful to you if you have not run across it is member Pkane's Deltawave. Free nulling software. He also has Multi-tone you see people using and a few others.

Here is a thread where he first started it.

And here you can download it and his other software.

Thanks! I have never tried a software null tester, depending on performance it may have saved me a lot of time. I copied Ethan Winer's circuit shown in the Audioxpress article and it's good to -110db which is good enough for me. My audiophile friends were not happy when I made it but was so proud of myself bringing it to audio hangouts where they would drink beer and bring their new "toys" for everyone to audition. We do sources, amps etc.... but lots of the guys bring their cables and other snake oil crap which was the main reason behind building the null tester. I could quickly show them in terms they can clearly understand that once there is a null there is no difference electrically; i.e. zero difference in sound. Some thick headed types that believe "we don't know how to measure all audio qualities" don't understand the concept but for many deprogrammed them from the myths and lies and ended up saving a lot of money and being a lot happier. I find the ones that believe in voodoo get hung up on always trying to find the right "synergy" not just between components but also cables. I don't see how anyone would ever be happy going about their systems that way, there is just too much trial and error especially human perception error that cannot be trusted.
 

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,521
Likes
37,050
Thanks! I have never tried a software null tester, depending on performance it may have saved me a lot of time. I copied Ethan Winer's circuit shown in the Audioxpress article and it's good to -110db which is good enough for me. My audiophile friends were not happy when I made it but was so proud of myself bringing it to audio hangouts where they would drink beer and bring their new "toys" for everyone to audition. We do sources, amps etc.... but lots of the guys bring their cables and other snake oil crap which was the main reason behind building the null tester. I could quickly show them in terms they can clearly understand that once there is a null there is no difference electrically; i.e. zero difference in sound. Some thick headed types that believe "we don't know how to measure all audio qualities" don't understand the concept but for many deprogrammed them from the myths and lies and ended up saving a lot of money and being a lot happier. I find the ones that believe in voodoo get hung up on always trying to find the right "synergy" not just between components but also cables. I don't see how anyone would ever be happy going about their systems that way, there is just too much trial and error especially human perception error that cannot be trusted.
Comparing with a device like Ethan's is probably the best way to null test. The software lets you record digitally the output of devices and then compare to two files at a later time. It includes the ability to synchronize the two files in time even if the clocks were running at different speeds etc.
 

Awsmone

Member
Joined
Jan 29, 2019
Messages
68
Likes
67
Any digital audio workstation (DAW)will allow you to do this ( Null test)

link below

 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,223
Likes
17,799
Location
Netherlands
Any digital audio workstation (DAW)will allow you to do this ( Null test)

link below

That all nice and good for test signals that you copy and invert, for two real recordings thought, it’s quite a bit more complex. You’ll need to align the two recordings perfectly for a good result, that means sub sample accuracy. With a DAW that is generally not easy to do, especially by hand.
 

Awsmone

Member
Joined
Jan 29, 2019
Messages
68
Likes
67
That all nice and good for test signals that you copy and invert, for two real recordings thought, it’s quite a bit more complex. You’ll need to align the two recordings perfectly for a good result, that means sub sample accuracy. With a DAW that is generally not easy to do, especially by hand.
Another option is to insert a sample-accurate delay plugin into one of the stereo pairs, play both pairs and adjust the delay on the fly until both streams are perfectly sync. With a bit of training this can be done within a minute. When sync is far off you hear double sources, like an echo. When sync is close, the mix becomes colored (comb filtering) and when sync is perfect (within one sample) the mix output will be silence or very low level, depending on how identical the two sources are., u are correct they need to be in sync
 

chych7

Active Member
Joined
Aug 28, 2020
Messages
276
Likes
422
No. The people who spend more money on cables don't do so because they have noise issues. They do so to get "darker backgrounds, better soundstage, more analog sound, etc." Paradoxically, some of the fancy cables I have tested are worse than generic ones when it comes to noise pick up.

I wish you applied this same logic to amplifiers and DACs. Lots of people believe they are getting "darker backgrounds, better soundstage" etc with various components reviewed here, and often use your reviews and measurement to support their judgement. But in reality, electrical measurements don't directly translate to what is audible, and is a big missing piece of information that this site could address, but doesn't.
 

captainbeefheart

Senior Member
Joined
Apr 18, 2022
Messages
360
Likes
446
But in reality, electrical measurements don't directly translate to what is audible, and is a big missing piece of information that this site could address, but doesn't.

What?:oops:

Why don't electrical measurements directly translate to what's audible? Feel free to educate us on information you have that engineers don't know.


Would you agree or disagree that two different sounding signals will have different electrical properties? If it's not electrical properties changing the sound then what is it?

I have heard this argument from Paul McGowan and his shills but nobody can ever expand upon this theory with any evidence or anything besides the bizarre idea from people that don't understand the physics.
 

chych7

Active Member
Joined
Aug 28, 2020
Messages
276
Likes
422
What?:oops:

Why don't electrical measurements directly translate to what's audible? Feel free to educate us on information you have that engineers don't know.


Would you agree or disagree that two different sounding signals will have different electrical properties? If it's not electrical properties changing the sound then what is it?

I have heard this argument from Paul McGowan and his shills but nobody can ever expand upon this theory with any evidence or anything besides the bizarre idea from people that don't understand the physics.

Yes I agree two different sounding signals would have different electrical properties (I did NOT say otherwise, you flipped the order of my logic). The corollary is not necessarily true, which is what I was getting at, because we are humans, not audio precision analyzers. For example - electrical measurements can show bandwidth differences between amps cause roll-off at higher frequencies (>15 kHz). Well, most adults have degraded hearing above 15 kHz, and discerning ~0.5-1 dB changes up there is questionable (I certainly can't). Then you have the whole topic of whether SINAD matters and if it's audible (I guess if you were a robot, you'd care about 100 vs 120 dB SINAD). I'd like to see a clearer link between electrical measurements and what's actually audible; not just to the point of detectable, but to the point of whether it actually matters or not in terms of subjective quality. This would be helpful in understanding if one really needs to spend $$ on a high end amp/DAC/whatnot or not for audio improvements, or whether the improvements are "academic", and result in no actually audible changes (for humans adults, at least). The world of audio is full of snake oil and false promises; science could cut through that if applied thoroughly. I feel as if the reviews on this site fail to take that last step on electronics. At least for speakers/headphone reviews, there's a stronger link to the subjective experience.
 
Last edited:

captainbeefheart

Senior Member
Joined
Apr 18, 2022
Messages
360
Likes
446
Yes I agree two different sounding signals would have different electrical properties (I did NOT say otherwise, you flipped the order of my logic). The corollary is not necessarily true, which is what I was getting at, because we are humans, not audio precision analyzers. For example - electrical measurements can show bandwidth differences between amps cause roll-off at higher frequencies (>15 kHz). Well, most adults have degraded hearing above 15 kHz, and discerning ~0.5-1 dB changes up there is questionable (I certainly can't). Then you have the whole topic of whether SINAD matters and if it's audible (I guess if you were a robot, you'd care about 100 vs 120 dB SINAD). I'd like to see a clearer link between electrical measurements and what's actually audible; not just to the point of detectable, but to the point of whether it actually matters or not in terms of subjective quality. This would be helpful in understanding if one really needs to spend $$ on a high end amp/DAC/whatnot or not for audio improvements, or whether the improvements are "academic", and result in no actually audible changes (for humans adults, at least). The world of audio is full of snake oil and false promises; science could cut through that if applied thoroughly. I feel as if the reviews on this site fail to take that last step on electronics. At least for speakers/headphone reviews, there's a stronger link to the subjective experience.

I totally understand now and I am right there with you on your thinking.

I'd say we have a pretty good understanding to which electrical properties are correlated to our sound experience. It's not black and white but we can certainly make some decent connections between what we measure and what we hear. Your example of bandwidth is an easy one, some humans are lucky to hear up around 20kHz especially at younger ages yet I doubt anything above 15kHz is going to be a huge difference to the average human being, especially older folks. So we use 20kHz as a metric to make sure the few that can hear up there are satisfied. Distortion is another metric we understand very well, we know lower order harmonics especially the second harmonic are fairly benign, I never treat THD as a sole metric, it needs to be broken down to the distortion content. So for .5% THD that is second harmonic dominant will sound different than another amp that also has .5% THD that is third harmonic dominant or higher. SINAD is fine if we are talking <-100db but when we get say near -80db or higher it will make a big difference whether the offender is the second harmonic vs higher harmonics or worse noise is the offender causing the bad SINAD score. One major aspect I feel that gets overlooked is stability, marginally unstable amps regardless of the SINAD score can have audible issues vs another amp that may test worse on the SINAD scale yet is unconditionally stable. Another one is output impedance, SINAD doesn't tell you anything how the amp will control the motor it's driving or how the output signal will track a lumpy impedance. So testing into a static resistive load doesn't tell you everything you need to know. Overload recovery is another aspect, yes we typically size for adequate headroom to not run into this problem but there can be low powered amplifiers that have good overload recovery and sound just fine vs another low powered amplifier that has horrible overload recovery. Another huge aspect people lose sight of is speaker distortion, us engineers know transducers are just the hardest to get accurate. What's the point of .0001% THD if the speaker has 10% THD at 96db? Is the amp that has .1% THD really going to be that much worse vs the .0001% THD when the speaker distortion grossly dominates the bigger picture. The louder the sound pressure levels out of the speaker the worse the distortion gets by a huge margin, so who cares if the amp is .0001%THD at 100 watts? To me I believe the first watt is the most important because that's where the speakers will be at their lowest distortion range and the amp needs to be most accurate here.

There are other things but I certainly do not put much weight into SINAD for amplifiers, for DAC's yes I want one that has great performance, but I do believe anything lower than -110db just isn't going to be much of a difference and it's purely academic at this point. We are not reading an MRI so we don't need this extremely high precision for audio equipment, to a certain point good enough is just that, good enough. For me I do not think I will gain any audible benefit upgrading from my Topping DAC that was $120 and good down to -114db, why pay 5x the price to get to -120db when I can't hear it? Yes it's academic and you have bragging rights that the better device is in fact technically better but in the grand scheme of things the difference between the two is not really a factor for my listening quality.

I apologize if I misunderstood you at first, I thought you were saying something completely different but come to find out we are both thinking the same things.
 
Last edited:

OWC

Active Member
Joined
Apr 14, 2019
Messages
204
Likes
154
I have been reading through this thread.
But to begin with, I find it a rather strange question to begin with.

It's a bit like, yes you can cook a meal by just randomly throwing things in a pan and hope for the best that it doesn't taste awful.
The more often you do that, the more you get a sense of what works and what not.
(which on itself can also be considered as some extremely simple form of science on itself)
Or you can to a course or even study for it to understand why things work and why not for making a great meal.

Speaker design is no different.
Even more so when you work on a professional level (like I do), you just simply have to justify your decisions.
Plus you want certain methods that a repeatable and standardized.

Fact is, that our brains play on us by definition, so things like placebo effects will always be at play.
Which also means it can be very tricky to trust our ears (since we hear or don't hear things that are/aren't there)
On the other hand, correlating taste with just plain measurements can be tricky.
 
Top Bottom