Jump to content

For CD what does "0dBFS 2.3VRMS Signal Output" really mean?


Recommended Posts



On 21/02/2020 at 4:42 PM, tesla13BMW said:

@Ittaku thank you for that.  So if the speaker output is much the same for each of my sources and I have calculated the phono stage output voltage correctly then the cd output voltage must be similar to the calculated 700mV?

 

So reality is the 2VRMS CD standard for one is not a standard between cd player manufacturers nor is it of any use if one is trying to balance a gain structure from different sources as we do not know what output a CD is due to how it is recorded?

Exactly, the point is the potential of a CD players analog output stage to swing voltage into a given load, is always limited and controlled by the level that is recorded on the CD.  The players can state much higher figures as potential to swing into a given load, but it never happens because the CD itself limits attaining such high levels

 

You can begin to see the much lower level CD's really attain by allowing a computer program called audacity to sample a CD track and to display the level. Here is Vangelis 1492 track 12. The RMS scale value shown in light blue is not attributed as voltage, but is exceedingly close when checked with a multimeter playing the same track as a CD via the RCA output jacks on a CD player- which also indicates the flac file is the same level as the CD . The peak level is shown in dark blue.

 

We can also say the CD is well recorded in terms of the peaks never flattening and remaining within the allowed scale. 

Screenshot from 2020-02-20 13-30-08.png

Edited by stereo coffee
Link to comment
Share on other sites

2 hours ago, Telecine said:

Yes, I don’t think that the version with digital input had volume control.

Maybe the earlier version with digital in didn't have volume control but mine has both digital in and volume control.

Link to comment
Share on other sites



21 hours ago, Telecine said:

The output is the same from your line stage, regardless of the source feeding it, provided that the source output meets the minimum input sensitivity of your line stage. We don’t know what that is from the spec sheet.

@Telecine so I thought that all made sense.

 

But, my CD and phono have the same speaker output when the SUT is on 10 x.  However, if I go to 20 x on the SUT I get a volume lift on phono and the phono is louder than the CD.  This suggests to me that both signals were not above the input sensitivity of the pre amp and negates your reason for them both sounding the same when on 10 x.

 

I am still confused how the 2.3V from the CD sounds the same loudness as the 0.7V calculated for phono.

 

I wish @Zaphod Beeblebrox was still here as he seemed to explain these types of questions.

Link to comment
Share on other sites

15 hours ago, stereo coffee said:

Exactly, the point is the potential of a CD players analog output stage to swing voltage into a given load, is always limited and controlled by the level that is recorded on the CD.  The players can state much higher figures as potential to swing into a given load, but it never happens because the CD itself limits attaining such high levels

 

You can begin to see the much lower level CD's really attain by allowing a computer program called audacity to sample a CD track and to display the level. Here is Vangelis 1492 track 12. The RMS scale value shown in light blue is not attributed as voltage, but is exceedingly close when checked with a multimeter playing the same track as a CD via the RCA output jacks on a CD player- which also indicates the flac file is the same level as the CD . The peak level is shown in dark blue.

 

We can also say the CD is well recorded in terms of the peaks never flattening and remaining within the allowed scale. 

Screenshot from 2020-02-20 13-30-08.png

So if you cannot use a CD players stated output, what can you use to investigate the gain structure of an amplification system for CD? 

 

 

Link to comment
Share on other sites

2 hours ago, tesla13BMW said:

So if you cannot use a CD players stated output, what can you use to investigate the gain structure of an amplification system for CD? 

 

 

You can arrange various test signals such as sine waves at levels such as ~2.0 V to be recorded - then playing back from a CD player test linearity using a oscilloscope for deviation from ideal. 

 

The point I was making though is levels like ~2.0V are NOT available from commercially available CD's where  to allow for handling peaks and offering pleasant if not outstanding reproduction of music, levels are deliberately kept much lower. 

 

A interesting point to consider is if we use alien levels and signals for testing CD players, when we expect them to also following testing to then reproduce music, would we have been better using music and painstakingly examining linearity of say a single piano note  in the first place?

 

I think the latter, as the further we distance ourselves from our hearing and natural perception of musical instruments ,   the objective measurement using alien wave forms and high levels ,  loses all credibility to then apply to be capable of  comparing to what we hear. 

 

To demonstrate this - at low level - listen to the sine wave - then a few minutes later a piano note at natural level, which is a valid test and which is invalid ?   

 

  

Edited by stereo coffee
Link to comment
Share on other sites

17 hours ago, stereo coffee said:

To demonstrate this - at low level - listen to the sine wave - then a few minutes later a piano note at natural level, which is a valid test and which is invalid ?   

 

It depends on what you are testing. One of the most fundamental contributions to signal theory, brought to you by esteemed mathematicians such as Fourier and Laplace, is that all possible waveforms can be constructed from a superposition of harmonic oscillations (sine waves). Further, the transmission characteristics of a linear system can be completely described by its frequency/phase response or its impulse response (the two are complementary and can be transformed into each other, but the former is easier to measure). Hence, the study of how systems respond to sine waves tells you what you need to know about them, and lets you predict how they will respond to arbitrary signals (within their range of linear response).

 

As for testing CD player output levels, reference 0dB 1kHz tracks are commonly found on test CDs. Play that and measure the output level of the CD player. Done. You can then use that to find an amplifier that matches in terms of sensitivity (the required input level to reach the specified output power into 8 ohms) and/or gain (the voltage amplification factor between input and output).

  • Like 1
Link to comment
Share on other sites



The two are though false ,

 

What makes easier to test, is not necessarily then properly testing the system, rather it has appealed to another sensibility, that of comprehending different information unrelated to music before you.  You are led to believe the test signal and its verification, can equally apply to hearing music, which is entirely false.

 

The second. Indeed yes might be found on test CD's but once again it is unrelated to the complexities of reproducing music. It is similarly false to assume test signals at reference 1khz 0db have relevance to matching to other equipment, simply because such levels and that form of test, do not occur in music CD's 

 

Here is what a 1Khz Sine wave looks like2018895926_Screenshotfrom2020-02-2414-14-21.thumb.png.f943978b30bf7415468c1ccd52fe7e7a.png

 

 

Rather you need a reference recording of I would suggest a few minutes on a piano, piano happens to contain some of the most complex ability of interpretation. 

 

Firstly make effort to hear piano being played well in a live environment. The environment is critical too because unlike test signals there is added information being presented to you - namely reverberation, decay, harmonic structure, the construction of the piano itself, the way the artist approaches and plays the instrument.    In a few weeks you should be able to hear distinct differences between a Steinway     https://www.steinway.com/   and a Bosendorfer https://www.boesendorfer.com/en/   as well as differences between artists those who play them well. 

 

And here is piece of piano music by Liz storyScreenshot from 2020-02-24 14-16-09.png

 

 

The added information qualifies audio equipment as being faithful in ability, whereas without reverberation decay harmonic structure etc, we are being provided limited information like the first image, that is quite alien to our hearing ability, and being asked to make false assumptions of ability to reproduce music, without that much needed information.  

 

A handful of pianists for you to hear on CD are Jon Balke -  Book of Velocities , Anat Fort - A long story ,  George Winston - December,  Liz Story - Unaccountable Effect

 

For actual testing of a CD player you will need a reference recording of one of these and audacity will do to begin with, as it displays in graphical form the piano recording - that similarly has captured reverberation, decay and harmonic structure. From the output of the CD player you then will need a line in recording to audacity to compare the CD player to the first  graphical representation.  You can also unlike test signals that are alien in comparison ,subjectively assess each by listening.  

Edited by stereo coffee
Link to comment
Share on other sites

1 hour ago, stereo coffee said:

What makes easier to test, is not necessarily then properly testing the system, rather it has appealed to another sensibility, that of comprehending different information unrelated to music before you.  You are led to believe the test signal and its verification, can equally apply to hearing music, which is entirely false.

 

It doesn't apply to hearing music (that happens in your brain), but it applies to transmitting it as an electronic signal.

 

1 hour ago, stereo coffee said:

The second. Indeed yes might be found on test CD's but once again it is unrelated to the complexities of reproducing music. It is similarly false to assume test signals at reference 1khz 0db have relevance to matching to other equipment, simply because such levels and that form of test, do not occur in music CD's 

 

Whatever occurs on music CDs is (once converted to analog) a superposition of sine waves at various levels. That's the entire complexity of it. The same holds for any random noise off the street. Our individual subjective response to music, or any kind of artistic noise, has nothing to do with how electronic systems transmit the signals.

 

When level matching the output of one component to the input of the next we try to maximise signal level in order to raise above analog noise as much as possible, while at the same time avoiding to overload the input and cause clipping. Hence, it is useful to consider the maximum possible output level of the source and the sensitivity and gain of the amplifier. I'm well aware that piano notes on CDs rarely come close to the maximum level, and that classical CDs in general are mastered well below 0dB. However, I do have some recordings that actually hit 0dB (or get very close), those will have to be catered for as well when matching up sources and amps.

Link to comment
Share on other sites

37 minutes ago, Steffen said:

 

It doesn't apply to hearing music (that happens in your brain), but it applies to transmitting it as an electronic signal.

 

 

Whatever occurs on music CDs is (once converted to analog) a superposition of sine waves at various levels. That's the entire complexity of it. The same holds for any random noise off the street. Our individual subjective response to music, or any kind of artistic noise, has nothing to do with how electronic systems transmit the signals.

 

When level matching the output of one component to the input of the next we try to maximise signal level in order to raise above analog noise as much as possible, while at the same time avoiding to overload the input and cause clipping. Hence, it is useful to consider the maximum possible output level of the source and the sensitivity and gain of the amplifier. I'm well aware that piano notes on CDs rarely come close to the maximum level, and that classical CDs in general are mastered well below 0dB. However, I do have some recordings that actually hit 0dB (or get very close), those will have to be catered for as well when matching up sources and amps.

You are being asked to accept a different interpretation in this case sine waves - but at the same time being asked to apply that different interpretation once you are happy with it, to.... wait for it....  be somehow applying to music ?   No, test signals will always be test signals which to the smallest fraction possible cannot account for, and likely never will account for, the complexities within music.

 

It is better to learn the sound of instruments and all of the intricacies like harmonic structure reverberation etc,  by going to live music recitals - and never consider test signals as properly capable of testing audio equipment.

 

A proper testing facility if piano is the instrument chosen, would have one permanently available to measure, and compare against, a bit like this:  http://www.classicspeakerpages.net/IP.Board/index.php?/topic/869-live-versus-recorded-truth-in-listening/

 

As for level, with enough refinement to the electronics used for testing, we may see peaks captured in the recording much higher than the display reveals. Partnering equipment you use should have as little change to the audio available from the source, as is possible, prior to amplification, in order to pass on the same attributes in the recording.

 

Insight into the utmost care needed to initially capture dynamics and then attempt their replay,  is described here:      https://warehousesound.com/r/dbxTYPEIV.pdf   which shows replay equipment, thankfully is behind the capability of some recordings. It would be a worry, if it was the other way around. 

Edited by stereo coffee
Link to comment
Share on other sites

25 minutes ago, stereo coffee said:

You are being asked to accept a different interpretation in this case sine waves - but at the same time being asked to apply that different interpretation once you are happy with it, to.... wait for it....  be somehow applying to music ?   No, test signals will always be test signals which to the smallest fraction possible cannot account for, and likely never will account for, the complexities within music.

I'm not sure what you're saying, that there is some secret sauce hidden in the electronic signal, besides the physical reality?

26 minutes ago, stereo coffee said:

It is better to learn the sound of instruments and all of the intricacies like harmonic structure reverberation etc,  by going to live music recitals - and never consider test signals as properly capable of testing audio equipment.

I'm a great supporter of live music, especially of the acoustic kind (i.e. no loudspeakers), but this cannot replace component testing, let alone level matching between components (as per the OP's question). You can't listen to an individual component, only to an entire system from source to speaker/room. To get insight into how individual components affect the signal along the way you need to use test signals and compare inputs and outputs.

Link to comment
Share on other sites

30 minutes ago, Steffen said:

recordings that actually hit 0dB (or get very close)

I can add some data for that

 

Of my 970 non-classical CDs (pop,folk,rock,prog,metal.....)

859 have a peak value within 1dB of 0dBFS (89%)

Another 87 have peak value between -2dB and -1dB

ie 98% are within 2dB of 0dBFS

 

209 of the 970 have flat topped peaks ie clipped (22%)

 

It is fair to say that a CD will be peaked at or near maximum, and stands a good chance of having a clipped piece.

 

[I haven't bothered to do the sums on my 219 classical CDs except to say that every one I looked at just now had tracks within 2dB of 0dBFS.  However other tracks on the same CD were lower as one would expect, ie -3 or -4dB]

 

 

-------------------

 

Anyway, back to the original question ...

 

Time to measure some voltages.

 

 

 

 

  • Like 1
Link to comment
Share on other sites



23 minutes ago, aechmea said:

I can add some data for that

 

Of my 970 non-classical CDs (pop,folk,rock,prog,metal.....)

859 have a peak value within 1dB of 0dBFS (89%)

Another 87 have peak value between -2dB and -1dB

ie 98% are within 2dB of 0dBFS

 

209 of the 970 have flat topped peaks ie clipped (22%)

 

It is fair to say that a CD will be peaked at or near maximum, and stands a good chance of having a clipped piece.

But peak tells you nothing about perceived loudness which is more the RMS A weighted sound. Virtually all digital releases are designed to use as close to zero as possible for their loudest peaks, but have wildly different average values which are the perceived loudness. 

 

Here's a Prokofiev symphony getting to virtually 0dB.

W09OEQQ.png

 

And here's some JPop:

YEFtzdJ.png

  • Like 1
Link to comment
Share on other sites

Yep, so for the intents and purposes of this thread that means we must expect, and be prepared for, the CD player producing close to its maximum output level occasionally (classical) or a lot of the time (highly compressed music).

Link to comment
Share on other sites

1 hour ago, Steffen said:

I'm not sure what you're saying, that there is some secret sauce hidden in the electronic signal, besides the physical reality?

I'm a great supporter of live music, especially of the acoustic kind (i.e. no loudspeakers), but this cannot replace component testing, let alone level matching between components (as per the OP's question). You can't listen to an individual component, only to an entire system from source to speaker/room. To get insight into how individual components affect the signal along the way you need to use test signals and compare inputs and outputs.

A electronic test signal, whether it be sine wave triangle wave, or square wave or other variants is just that, a  test signal. If we are to get closer for audio equipment to be better at reproducing actual music, a music signal needs to be what is compared to.

 

 This raises an interesting observation that certain electronic music can indeed get close to test signals.  Lou Reed's Metal Machine music as example is the most audacious attempt I know of at presenting sound that has little if any relationship to conventional music as we know it, therein it has a certain beauty of its own.     

 

 

Audio component testing can certainly use test signals, but should also include comparison to music and indeed live instruments if their budget stretches that far, I am sure we would have better equipment made if the real thing, was  being compared to.

 

Edited by stereo coffee
Link to comment
Share on other sites

33 minutes ago, Steffen said:

Yep, so for the intents and purposes of this thread that means we must expect, and be prepared for, the CD player producing close to its maximum output level occasionally (classical) or a lot of the time (highly compressed music).

The CD as you acknowledge, simply reflects what has been recorded. As I will explain below the maximum attainable level of CD - where undistorted sound is reproduced, is rarely, if ever attained.

 

So to understand what is attainable in undistorted level we need to know about a subject called companding. Its history goes back to a failed patent of Murray G Crosby which described emphasis and de-emphasis. In the 1960's two manufacturers developed equipment based on Crosby's work,  that was widely used in the recording industry, namely DBX and Dolby. Companding is a portmanteau of compression, and expansion.  A recording could be compressed when recording and then played back expanded, thereby capturing the available dynamic range.  

 

Prior to companding the maximum dynamic range a recording could offer was limited by the tape format to 55db, which as expressed as bits approximately 9.8 bits  

 

In the mid to late 1960's DBX companders could attain 75-85db which was very close to the 16 bit standard., so recordings as at say 1969 were equal to the ability of CD players of 1988 with respect to capturing dynamic range. in fact some of the best recordings are from this era, where large reel to reel recorders and either Dolby or DBX companders were used.

 

When digital recording was first made available, the prior 0db levels used with analog tape, had to be reduced to -7db   so a step backward occurred with many recording studios and recordings they could offer. If the recording engineer attempted anything over -7db they were greeted with very harsh overload by the recorder. Indeed some recording engineers worked out that they could in some cases marry the benefits of companding once again with digital equipment. DBX type 1 enables this. But digital had, and still does have very specific needs where the analog to digital conversion takes place in order to regain the 16 bit ability analog previously enjoyed.  https://warehousesound.com/r/dbxTYPEIV.pdf

 

I would need to know more about CD manufacturing equipment before proceeding, but I strongly suspect recordings are at least 15 years ahead in capturing dynamic range, than the commercial CD's we purchase. At a guess  replay equipment we presently use with possibly 105db ability in some cases matches an available CD, because of dynamic range losses when the CD is manufactured. We can partially blame ourselves too, for not adopting companding in the CD era for domestic use. But there is still opportunity to do that if we begin to show interest. 

 

 

  • Love 1
Link to comment
Share on other sites



Well I don't know if I may have stumbled on a the answer to my question.  And if someone can confirm that it makes sense or otherwise it would be great.

 

So I saw this mentioned

There is no fixed connection between digital level in audio files and analog (mechanical) level on record. We always set cutting level according musical content.
If we know the reference level is supposed to be 5 cm/s we set to cutting level to meet this requirement.
5 cm/s is quite low for reference tone. It corresponds to -9 dB (= 16 µm peak to peak amplitude) of Neumanns reference level at 1 kHz (0 dB = 14,14 cm/s = 45 µm PP).

 

Also found this;

 

"The Ultimate Analogue Test LP is a milestone and very powerful reference for those truly concerned about getting the very best performance from an analog (vinyl) front end and also for those professionals who manufacture the related analog front end products. To start with, Track One very accurately provides a 1kHz tone at 7 cm/s mono (in-phase, lateral) NAB OVU standard reference level, which has been the industry standard since 1964. This supercedes the long-outdated RCA 5 cm/s (3.54 cm/s stereo) test level. Track One can also be used as an outside track wow and flutter reference as will the inner Track 10...two ways to approach this measurement. A close look shows that each track on this disc can be used in multiple ways far beyond the list provided in the limited space on the record jacket. Hats off to all of those involved in making this outstanding test LP." – Keith Herron, Herron Audio

 

And this from the NAB standard;

 

Reference Recorded Program Level-Monophonic

 

3.10 It shall be standard that the reference recorded program level shall produce the same reference deflection on a standard volume indicator (ASA Standard C16.5-1961) 2s that produced by a 1000 cps tone recorded at a peak velocity of 7 centimeters per second.

 

Reference Recorded Program Level-Stereophonic

 

3.15 'It shall be standard that the reference recorded program level for each channel in its plane of modulation shall produce the same reference deflection on a standard volume indicator (ASA Standard C16.5-1961) as that produced by a 1000 cps tone recorded at a peak velocity of 5.0 centimeters per second. (Approximately 3 db below the monophonic Reference Recorded Program Level.)

 

I read; to calibrate dBVU one uses the test record and sets the dBVU level to 0.  So the 7cm/s sets the dBVU (volume scale) to 0.

 

So using my figures, if I increase the phono level by 9dB i.e. to the 7cm/s reference I have a voltage of just over 2V.  This is just over 1dB difference between cd and phono of what is going to the line stage.

 

I don't consider 1dB would be a audible difference and I'm not hearing one.

 

Can anyone shed more light on this?  Where are all the analogue experts now :emot-bang: 

Link to comment
Share on other sites

  • 1 month later...
On 24/02/2020 at 5:20 PM, tesla13BMW said:

Well I don't know if I may have stumbled on a the answer to my question.  And if someone can confirm that it makes sense or otherwise it would be great.

Yes you have found your answer!

 

The meaning of the CD player output figure is quite specific. You asked: " For CD what does "0dBFS 2.3VRMS Signal Output" really mean?".  It simply means that when playing a test sine wave recorded on the CD, at the maximum possible recording level the 16-bit red book standard can accommodate, 0dB,  the analogue output will be 2.3V RMS (plus or minus small manufacturing tolerances). This will be the case at 100Hz, 1kHz or 10kHz and frequencies in between.  There is no RIAA equalisation curve to complicate things!

 

Modern CDs are typically mastered so that a piece of music reaches, or almost reaches, 0dB, for at least some of the time.

 

 

The meaning of the magnetic cartridge sensitivity is also specific but it is a much more complicated story.  A 0.4mV cartridge output at 1kHz when tracking a sine wave cut with a tip velocity of 5cm/s is one thing.  How the signal is processed before being sent to the cutting head for a particular disc is another.  The frequency response and the levels sent to the cutting head are manipulated so as to get the best overall result.  It is a compromise between on the one hand attaining a high tip velocity so as to improve the signal to noise ratio, and on the other hand reining in the tip velocity so as to guarantee trackability, lessen THD and IMD, avoid sacrificing an extended low frequency response, and perhaps even to fit in extra material on that side of the disc. 

In practice the recorded level achieved on a stereo vinyl disc can be greater than what the standards might suggest.

 

The following graph shows measured recorded levels for a selection of vinyl discs. (The measurements were reported in a Shure Technical Seminar in 1978.) It can be seen that some discs were found to have a recorded level at 1kHz well in excess of 5cm/s. (This graph appears on the webpage http://pspatialaudio.com/max_velo.htm )

 

shure%20velocities_enhanced.png

 

 

Edited by MLXXX
  • Like 1
  • Thanks 1
Link to comment
Share on other sites

  • Recently Browsing   0 members

    • No registered users viewing this page.




×
×
  • Create New...
To Top