Jump to content
IGNORED

Misleading Measurements


Recommended Posts

3 hours ago, March Audio said:

This will be a typical peak voltage of around 2.8 volts.

 

I know that you know this but for others*:  the RMS voltage (VRMS) is determined by multiplying the peak voltage value by 0.7071.

 

* I had to look this up :)

Link to post
Share on other sites
59 minutes ago, stereo coffee said:

 

 

-10 dBV Standard voltage reference level for consumer and some pro audio use (e.g. TASCAM), equal to 0.316 Vrms. (Tip: RCA connectors are a good indicator of units operating at -10 dBV levels.)

https://www.aes.org/par/d/#decibel

 

You do realize this means what everyone has been telling you it means, right? That it's simply a *standard reference level*?

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> wi-fi to router -> EtherREGEN -> microRendu -> USPCB -> ISO Regen (powered by LPS-1) -> USPCB -> Pro-Ject Pre Box S2 DAC -> Spectral DMC-12 & DMA-150 -> Vandersteen 3A Signature.

Link to post
Share on other sites
3 minutes ago, Jud said:

 

You do realize this means what everyone has been telling you it means, right? That it's simply a *standard reference level*?

Which happens to be used in the majority of audio equipment, in use today, ie those having RCA connectors.  

-10 dBV Standard voltage reference level for consumer and some pro audio use (e.g. TASCAM), equal to 0.316 Vrms. (Tip: RCA connectors are a good indicator of units operating at -10 dBV levels.)

Link to post
Share on other sites
6 hours ago, March Audio said:

Music usually has a low rms level compared to its peak, maybe a ratio of 1:5. 

Wow - this is news!  So the maximum volume level of the transient peaks in most music is about 5 times the level of the bulk of its content.  We should name that ratio because it seems so useful.  Let’s figure out a good name together.

 

Maybe there’s a way to measure and specify the two parameters in your 1:5 ratio.  The 1 is the general overall level of the main sonic content of music - that sounds kinda sorta like it could be the average level. The 5, which represents the level of the highest peak, is an instantaneous measurement we could call (drum roll please...) the peak.  How about calling it the average-to-peak ratio?

 

Hmmmm - I just noticed the curious coincidence that the ratio between -10 dBV and +6 dBV is in the general range of 1 to 5.  Perhaps, per rule #39 of Leroy Jethro Gibbs, there are no coincidences.  

Link to post
Share on other sites
2 hours ago, stereo coffee said:

Which happens to be used in the majority of audio equipment, in use today, ie those having RCA connectors.  

 

 

-10 dBV Standard voltage reference level for consumer and some pro audio use (e.g. TASCAM), equal to 0.316 Vrms. (Tip: RCA connectors are a good indicator of units operating at -10 dBV levels.)

With respect this is simply incorrect.

 

Well I can only suggest you take a test signal at 0dBfs and measure the voltage at the RCA output of a range of dacs to test if your opinion is correct.

 

You will find they are nearly all around 2 volts rms.

 

If you don't believe me please take a visit to Audio Science Review where there are tests of probably a hundred plus dacs that show this to be the case.

 

Can I ask if you have a dac or CD player and if so what make and model it is?

MARCH~audio
excellence in audio
www.marchaudio.com
 

Link to post
Share on other sites
47 minutes ago, bluesman said:

Wow - this is news!  So the maximum volume level of the transient peaks in most music is about 5 times the level of the bulk of its content.  We should name that ratio because it seems so useful.  Let’s figure out a good name together.

 

Maybe there’s a way to measure and specify the two parameters in your 1:5 ratio.  The 1 is the general overall level of the main sonic content of music - that sounds kinda sorta like it could be the average level. The 5, which represents the level of the highest peak, is an instantaneous measurement we could call (drum roll please...) the peak.  How about calling it the average-to-peak ratio?

 

Hmmmm - I just noticed the curious coincidence that the ratio between -10 dBV and +6 dBV is in the general range of 1 to 5.  Perhaps, per rule #39 of Leroy Jethro Gibbs, there are no coincidences.  

1:5 is very much a generalisation, it depends on the music.  I will post some examples later.

MARCH~audio
excellence in audio
www.marchaudio.com
 

Link to post
Share on other sites
2 hours ago, Jud said:

 

You do realize this means what everyone has been telling you it means, right? That it's simply a *standard reference level*?

 

2 hours ago, stereo coffee said:

Which happens to be used in the majority of audio equipment, in use today, ie those having RCA connectors.  

-10 dBV Standard voltage reference level for consumer and some pro audio use (e.g. TASCAM), equal to 0.316 Vrms. (Tip: RCA connectors are a good indicator of units operating at -10 dBV levels.)

 

What I think you may not be catching is that *a* standard reference level is not *the only* standard reference level.

One never knows, do one? - Fats Waller

The fairest thing we can experience is the mysterious. It is the fundamental emotion which stands at the cradle of true art and true science. - Einstein

Computer, Audirvana -> wi-fi to router -> EtherREGEN -> microRendu -> USPCB -> ISO Regen (powered by LPS-1) -> USPCB -> Pro-Ject Pre Box S2 DAC -> Spectral DMC-12 & DMA-150 -> Vandersteen 3A Signature.

Link to post
Share on other sites
27 minutes ago, March Audio said:

1:5 is very much a generalisation, it depends on the music.  I will post some examples later.

You don’t need to post examples. Of course it’s a generalization, just as “consumer line level” specs are generalizations.  And the 0.316 / 2 volt spec pair is a generalized compromise reflecting real world use.  But the concept is what’s important in this discussion.  Whether you call it usual, general, average, mean or something else, common use of a design spec for a common operating parameter that serves as a proxy for the amplitude of your music at your usual listening level helps to assure compatibility among multiple brands and designs of devices.  In this case, we’re discussing the general loudness of your music during playback when the device being measured is in the optimal part of its operating range.
 

You missed my point, which is that the concept of an average level vs a peak level is exactly what is defined by the -10 dBV / +6 dBV line level spec.  The quadratic mean (yes - mean!) of the audio signal passing through your system (measured as RMS voltage) when you play music is the “usual” level during playback at the point of measurement.  At line level outputs, it’s in the range of -10 dBV (0.316 vRMS).  As there is no mandated standard for this, line levels vary from design to design - but few consumer preamps ignore or vary very much from the 0.316 / 2 volt ratio because few amplifiers have input sensitivities far enough from the norm to be unusable with this range of input signal levels.  Yes, there are exceptions just as there are audiophiles who listen at average SPLs far above or below the usual range.

 

The proof of the pudding is that we don’t have to buy one preamp, DAC, amplifier etc for quiet listening, one for our usual level, and one for loud playback.  We don’t have a chamber music amplifier and a hard rock amplifier.  Conformity with the general line level range (defined by -10 dBV for usual / average / overall SPLs at the speakers driven by almost any amplifier and +6 dBV for the highest peak level at rated distortion and totally unclipped) makes most front end devices compatible with most power stages.

 

This is all based on one’s average listening level, which is measurable all along the signal chain from first analog input transducer (cartridge, tape head, DAC etc) through each interface all the way to the listening position (where it’s measured as SPL in acoustic dB).
 

Once the electrical signal is transduced into mechanical energy by the speakers, it becomes audible and we measure it as the average SPL coming from your speakers.  And there too, we use both average and peak measurements to define the operating range.

Link to post
Share on other sites
1 hour ago, pkane2001 said:

 

This is called crest factor, which is specifically peak-to-RMS ratio. 

 

Thanks, but I was being educationally ironic (my just-invented term for using sarcasm as a teaching aid). If I was a bit too sarcastic, thanks to my level of frustration over the refusal to accept that RMS voltage is an averaging measurement, I apologize.  
 

The crest factor is yet another example of the same concept applied at a different point in the signal path - it’s probably most useful to recording engineers, mastering labs etc because of the broad range of average-to-peak content in the sounds of individual instruments.  It can also serve as a ratio of sustained to transient content, eg it’s quite different for violins and percussive instruments.  So it’s yet another example of a useful parameter based on the relationship between peak and average content in the same complex signal.

Link to post
Share on other sites
57 minutes ago, bluesman said:

You don’t need to post examples. Of course it’s a generalization, just as “consumer line level” specs are generalizations.  And the 0.316 / 2 volt spec pair is a generalized compromise reflecting real world use.  But the concept is what’s important in this discussion.  Whether you call it usual, general, average, mean or something else, common use of a design spec for a common operating parameter that serves as a proxy for the amplitude of your music at your usual listening level helps to assure compatibility among multiple brands and designs of devices.  In this case, we’re discussing the general loudness of your music during playback when the device being measured is in the optimal part of its operating range.
 

You missed my point, which is that the concept of an average level vs a peak level is exactly what is defined by the -10 dBV / +6 dBV line level spec.  The quadratic mean (yes - mean!) of the audio signal passing through your system (measured as RMS voltage) when you play music is the “usual” level during playback at the point of measurement.  At line level outputs, it’s in the range of -10 dBV (0.316 vRMS).  As there is no mandated standard for this, line levels vary from design to design - but few consumer preamps ignore or vary very much from the 0.316 / 2 volt ratio because few amplifiers have input sensitivities far enough from the norm to be unusable with this range of input signal levels.  Yes, there are exceptions just as there are audiophiles who listen at average SPLs far above or below the usual range.

 

The proof of the pudding is that we don’t have to buy one preamp, DAC, amplifier etc for quiet listening, one for our usual level, and one for loud playback.  We don’t have a chamber music amplifier and a hard rock amplifier.  Conformity with the general line level range (defined by -10 dBV for usual / average / overall SPLs at the speakers driven by almost any amplifier and +6 dBV for the highest peak level at rated distortion and totally unclipped) makes most front end devices compatible with most power stages.

 

This is all based on one’s average listening level, which is measurable all along the signal chain from first analog input transducer (cartridge, tape head, DAC etc) through each interface all the way to the listening position (where it’s measured as SPL in acoustic dB).
 

Once the electrical signal is transduced into mechanical energy by the speakers, it becomes audible and we measure it as the average SPL coming from your speakers.  And there too, we use both average and peak measurements to define the operating range.

Sorry but there is no "usual level" and there is no reason to relate it to 316mV.  You seem to be conflating "a level you may listen at" with a technically defined sensitivity.

 

The level required to playback at any particular volume will depend on multiple factors including pre / power amp gain and speaker sensitivity.

 

MARCH~audio
excellence in audio
www.marchaudio.com
 

Link to post
Share on other sites

The Quad 306 was mentioned earlier.  Typical of amp designs from that era (ie pre cd)  it had high sensitivity.

 

50 watt output into 8 ohms amp.

Input of 375mV required to drive it to 50 watts into 8 ohms.

50 watts into 8 ohms requires 20volts which is a gain of 34.5dB.

 

This old amp is close to conformance with the -10dBv convention.

 

Most modern power amps are only around 26dB gain (it does depend on their power output) because line levels have shifted to 2 volts.  Just do a bit of reading on amp spec pages to confirm this is the case.

 

MARCH~audio
excellence in audio
www.marchaudio.com
 

Link to post
Share on other sites
On 4/27/2021 at 7:15 PM, stereo coffee said:

You only need to attenuate consumer line level,  which is NOT 2v RMS its nominal 0.316v RMS  , You might be reading too many reviews that test equipment at that high level, that is unrelated to what is at a RCA socket when playing CD's in home audio systems. 

 

Why its bad is because your source component contains the very best opportunity of presenting music as it actually can be, adding anything in-between other than resistance to attenuate,  simply detracts from that opportunity.    

 

I believe the initial standard for the output of CD players was 2V.

 

Really, all of this comes down to proper gain-staging rather than any standards, i.e. the lowest gain that drives your amplifier to full output, such that you minimize attenuation in preceding components, either analog or digital.  Significant gain in a preamp that is only subsequently attenuated is suboptimal at best.

 

My DAC outputs at the professional standard.  My amplifier on its lowest gain setting (which minimizes noise, the goal of all of this) typically leads to -5 to -12 dbs of total system attenuation, loud listening it is close to 0 (if our listening was only at one level this would allow and be the ideal scenario).

 

Bill

Labels assigned by CA members: "Cogley's ML sock-puppet," "weaponizer of psychology," "ethically-challenged," "professionally dubious," "machismo," "lover of old westerns," "shill," "expert on ducks and imposters," "Janitor in Chief," "expert in Karate," "ML fanboi or employee," "Alabama Trump supporter with an NRA decal on the windshield of his car," sycophant

Link to post
Share on other sites
1 hour ago, March Audio said:

Sorry but there is no "usual level" and there is no reason to relate it to 316mV.  You seem to be conflating "a level you may listen at" with a technically defined sensitivity.

 

The level required to playback at any particular volume will depend on multiple factors including pre / power amp gain and speaker sensitivity.

 

I let the edit function time out - here’s the rest of the story:

 

The range and distribution of the SPLs in the first part of my response are the acoustic energy analogues of the same mathematical functions derived for line level voltages in the system while listening - so the usual range of listening levels is mirrored by the usual signal levels in the signal path during listening.  
 

A distribution function is defined by location, consistency, scale, shape, and skew.  Location is determined by measures of central tendency - mean, median and mode. Scale is the range of highest and lowest values.  The shape of a distribution like this is determined by its symmetry around the chosen measure of central tendency (skew) and the standard deviation of the value set (variability).

 

An audio engineer’s choice of line level at preamp outputs is based on the above.  Despite your assertion that there is no usual average SPL for audiophile listening, there is clearly an average range - and the choice of line levels near the top of the range for rated distortion and other performance specs (~0.316 and ~2 VRMS for average / usual / habitual / common / often used listening levels from most analog consumer power amplifier stages) is made to accommodate average use by the average buyer.

 

Link to post
Share on other sites
8 hours ago, March Audio said:

With respect this is simply incorrect.

 

Well I can only suggest you take a test signal at 0dBfs and measure the voltage at the RCA output of a range of dacs to test if your opinion is correct.

 

You will find they are nearly all around 2 volts rms.

 

If you don't believe me please take a visit to Audio Science Review where there are tests of probably a hundred plus dacs that show this to be the case.

 

Can I ask if you have a dac or CD player and if so what make and model it is?

You have identified precisely why misleading measurements persist, particularly in the forum you mention. 

here is the Audio Engineering Society    https://www.aes.org/par/d/#decibel    definition of what you refer to: 

 

0 dBFS A digital audio reference level equal to "Full Scale." Used in specifying A/D and D/A audio data converters. Full scale refers to the maximum peak voltage level possible before "digital clipping," or digital overload (see overs) of the data converter. The Full Scale value is fixed by the internal data converter design, and varies from model to model. [According to standards people, there's supposed to be a space between "dB" and "FS" — yeah, right, like that's gonna happen.]

 

The same level ability though is not occurring at the RCA outputs of CD players, when playing CD's, because they always comply with consumer line level 0.316V RMS  

 

The peak level of 0.316V RMS is  0.447V RMS  , and the Peak to Peak of 0.316V RMS is 0.894V RMS 

 

You can see compliance to those levels here, RMS is in light blue , and peak in dark blue. 

 

 

 

 

 

Screenshot from 2021-02-21 17-04-20.png

Link to post
Share on other sites
8 minutes ago, stereo coffee said:

You have identified precisely why misleading measurements persist, particularly in the forum you mention. 

here is the Audio Engineering Society    https://www.aes.org/par/d/#decibel    definition of what you refer to: 

 

0 dBFS A digital audio reference level equal to "Full Scale." Used in specifying A/D and D/A audio data converters. Full scale refers to the maximum peak voltage level possible before "digital clipping," or digital overload (see overs) of the data converter. The Full Scale value is fixed by the internal data converter design, and varies from model to model. [According to standards people, there's supposed to be a space between "dB" and "FS" — yeah, right, like that's gonna happen.]

 

The same level ability though is not occurring at the RCA outputs of CD players, when playing CD's, because they always comply with consumer line level 0.316V RMS  

 

The peak level of 0.316V RMS is  0.447V RMS  , and the Peak to Peak of 0.316V RMS is 0.894V RMS 

 

You can see compliance to those levels here, RMS is in light blue , and peak in dark blue. 

 

 

 

 

 

Screenshot from 2021-02-21 17-04-20.png

Sorry too many RMS's there, Vp being the term for Voltage peak, and Vpp being peak to peak 

Link to post
Share on other sites
6 hours ago, bluesman said:

I let the edit function time out - here’s the rest of the story:

 

The range and distribution of the SPLs in the first part of my response are the acoustic energy analogues of the same mathematical functions derived for line level voltages in the system while listening - so the usual range of listening levels is mirrored by the usual signal levels in the signal path during listening.  
 

A distribution function is defined by location, consistency, scale, shape, and skew.  Location is determined by measures of central tendency - mean, median and mode. Scale is the range of highest and lowest values.  The shape of a distribution like this is determined by its symmetry around the chosen measure of central tendency (skew) and the standard deviation of the value set (variability).

 

An audio engineer’s choice of line level at preamp outputs is based on the above.  Despite your assertion that there is no usual average SPL for audiophile listening, there is clearly an average range - and the choice of line levels near the top of the range for rated distortion and other performance specs (~0.316 and ~2 VRMS for average / usual / habitual / common / often used listening levels from most analog consumer power amplifier stages) is made to accommodate average use by the average buyer.

 

You have actually confirmed what I have said.

 

I didn't say there was no average listening level, although obviously it varies dramatically, I was referring to signal source level, whichvis what this conversation is about.

 

Question for you.

 

Assuming a desired listening level of 85dB spl, calculate the signal source level when:

 

1. Speaker sensitivity is 84dB/2.83v/m and amplifier gain is 24dB

 

2. Speaker sensitivity is 92dB/2.83v/m and amplifier gain is 30dB.

 

Domestic set ups can easily fall within this range.

MARCH~audio
excellence in audio
www.marchaudio.com
 

Link to post
Share on other sites

Not quite sure what the fuss is about ... simply put, there is a huge range of subjective average levels with the recordings themselves - I've had systems which completely ran out of gain with some CDs of mine, it was impossible to lift the level above, say, what a TV feels like; yet that same system could have my ears ringing after 5 minutes of a modern pop recording, at a volume setting quite a bit of a way down.

 

Worrying about differences of a few dB in some part of the chain is completely meaningless in the face of this - what one wants is the capability of a huge range of gain settings, without hearing any anomalies at either ends of the range.

Frank

 

http://artofaudioconjuring.blogspot.com/

 

 

Over and out.

.

 

Link to post
Share on other sites
18 minutes ago, fas42 said:

Not quite sure what the fuss is about ... simply put, there is a huge range of subjective average levels with the recordings themselves - I've had systems which completely ran out of gain with some CDs of mine, it was impossible to lift the level above, say, what a TV feels like; yet that same system could have my ears ringing after 5 minutes of a modern pop recording, at a volume setting quite a bit of a way down.

 

Worrying about differences of a few dB in some part of the chain is completely meaningless in the face of this - what one wants is the capability of a huge range of gain settings, without hearing any anomalies at either ends of the range.

The fuss is IMO about levels we find available on CD's DVD's blurays etc . The levels being measured with equipment bear no relationship at all,  being absurdly 6.33 times higher. 

 

Mike Oldfields Amarok is the loudest CD I own, but it still has peaks that accommodate consumer line level.    Hence the effort I consider is to start seeing reviews of equipment that match the variety of media we enjoy. We are hopefully collectively now able to see the fabrication of measuring equipment at 0DBfs that is NOT consumer line level we enjoy from our CD's DVD's etc which is 0.316V RMS. Yes sure the equipment can reproduce such level with test CD's , but few of us would enjoy or have in our collections sine waves at 0DBfs, rather what we have is CD's DVD's etc with music,  that rigidly sticks to consumer line level.  

 

Below is Mike Oldfields Amarok, we can see the level gets to extend to the absolute peak to peak value of consumer line level, which is  0.894V, Vpp , and as can be seen explains visually why it is the loudest CD I own    

 

The secondary fuss is about sensitivity of power amps also matching to consumer line level. If this is done we can then just use resistance attenuation in between, and get in theory as close as possible to what our source has with capability at its output RCA.   We should instead be talking about ideal shunt and series resistances - as  Ike Willis said in Frank Zappa's Thing fish ... moving the project forward 

Screenshot from 2021-04-30 12-12-37.png

Link to post
Share on other sites
2 hours ago, stereo coffee said:

You have identified precisely why misleading measurements persist, particularly in the forum you mention. 

here is the Audio Engineering Society    https://www.aes.org/par/d/#decibel    definition of what you refer to: 

 

0 dBFS A digital audio reference level equal to "Full Scale." Used in specifying A/D and D/A audio data converters. Full scale refers to the maximum peak voltage level possible before "digital clipping," or digital overload (see overs) of the data converter. The Full Scale value is fixed by the internal data converter design, and varies from model to model. [According to standards people, there's supposed to be a space between "dB" and "FS" — yeah, right, like that's gonna happen.]

 

The same level ability though is not occurring at the RCA outputs of CD players, when playing CD's, because they always comply with consumer line level 0.316V RMS  

 

The peak level of 0.316V RMS is  0.447V RMS  , and the Peak to Peak of 0.316V RMS is 0.894V RMS 

 

You can see compliance to those levels here, RMS is in light blue , and peak in dark blue. 

 

 

 

 

 

Screenshot from 2021-02-21 17-04-20.png

 

With respect there is nothing misleading going on here, it is simply a misunderstanding on your behalf.

 

I will do my best to explain.

 

You have posted the correct definition of a full scale signal and yes it is exactly what s happening at a CD player or dac RCA output.  However I would go as far to say that no modern cd players or dacs have a line level output of 316mV.

 

Im afraid your plot does not demonstrate compliance.  It does not show that its voltage on the verticle scale and you are using music as a test test signal.

 

You simply cannot use music as a test signal to assess the line level output of a device.  You will get completely different results if you used say a quite piano concerto that was only recorded up to -10dBFS compared to a highly compressed death metal track that constantly hitting 0dBFS.

 

You *have* to use a consistent signal at a defined level; a sine wave at 0dBFS.

 

Here is a Gustard X16 that happens to be on my bench playing a 0dBFS 1kHz signal.  Output from the RCA is 2 volts RMS / 5.8v pk to pk.

 

20210430_084444.thumb.jpg.b6e62be51f9e9872e723b78302703384.jpg20210430_084453.thumb.jpg.5173d6e111a43dd6dd378896db1d5026.jpg

 

MARCH~audio
excellence in audio
www.marchaudio.com
 

Link to post
Share on other sites
1 minute ago, March Audio said:

 

With respect there is nothing misleading going on here, it is simply a misunderstanding on your behalf.

 

I will do my best to explain.

 

You have posted the correct definition of a full scale signal and yes it is exactly what s happening at a CD player or dac RCA output.  However I would go as far to say that no modern cd players or dacs have a line level output of 316mV.

 

Im afraid your plot does not demonstrate compliance.  It does not show that its voltage on the verticle scale and you are using music as a test test signal.

 

You simply cannot use music as a test signal to assess the line level output of a device.  You will get completely different results if you used say a quite piano concerto that was only recorded up to -10dBFS compared to a highly compressed death metal track that constantly hitting 0dBFS.

 

You *have* to use a consistent signal at a defined level; a sine wave at 0dBFS.

 

Here is a Gustard X16 that happens to be on my bench playing a 0dBFS 1kHz signal.  Output from the RCA is 2 volts RMS / 5.8v pk to pk.

 

20210430_084444.thumb.jpg.b6e62be51f9e9872e723b78302703384.jpg20210430_084453.thumb.jpg.5173d6e111a43dd6dd378896db1d5026.jpg

 

How did it sound ?  Try measuring  a music CD instead,  few of us enjoy or would have in our collections 1khz sine waves at 0DBfs, ....you should then find compliance to consumer line level.   

Link to post
Share on other sites

It sounded just fine.  You certainly will have recordings that hit 0dBFS.

 

Here is Amorok, I didnt search out the loudest segment just grabbed a small section around 6 minutes in.

 

767mV rms, 5.11 mV pk to pk

 

I could easily find louder recordings than this.

 

20210430_090433.thumb.jpg.23434dbcb828ffe6d8661d8fd15fbf55.jpg20210430_090439.thumb.jpg.70f21f53cb149a6e187981f08977f192.jpg

 

 

MARCH~audio
excellence in audio
www.marchaudio.com
 

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×
×
  • Create New...