"Black Backgrounds" in Music Reproduction?

[microstrip]

I'm not an EE. I am a line level electrician. I don't know much at all about using meters.
I read threads on people building speakers. And a good friend builds them. My understanding is energy is absorbed in a crossover. It would be best to have a line level device. I also have been told as I examined drivers to make a swarm of subs that dri ers also have different levels of power requirements. Some are very sensitive. Other require a lot of power, but can exert more force.

I make an inference then that a very quiet passage in a musical piece might have the amp only pushing a fraction of a watt. That fraction could be wasted as heat in the crossover and driver. If that happens, that is lost information emanating from the driver, leading to a perception of Black, as there is no sound.
Maybe you can put what I am saying into technical jargon. If it makes sense what I'm trying to convey. Or maybe explain how I'm wrong and power loss is not an issue with modern dynamic driver speakers.

Power loss is not an issue as long as the system is linear - unfortunately what you say makes no sense technically.

The more detailed speaker I have listened and own is by far the Quad ESL63 - 86dB/W. Much more detailed than the more efficient SoundLab A1 PX.

Many aspects of speaker performance, such as distortion, frequency response and dispersion affect our perception of detail.

I am not able to correlate efficiency with detail, although I surely remember some old designs that were very low efficiency and low detail. Low dispersion, as seen in horns, can help the perception of detail. IMO it is how myths develop ...

 

[Kingrex]

Power loss is not an issue as long as the system is linear - unfortunately what you say makes no sense technically.

The more detailed speaker I have listened and own is by far the Quad ESL63 - 86dB/W. Much more detailed than the more efficient SoundLab A1 PX.

Many aspects of speaker performance, such as distortion, frequency response and dispersion affect our perception of detail.

I am not able to correlate efficiency with detail, although I surely remember some old designs that were very low efficiency and low detail. Low dispersion, as seen in horns, can help the perception of detail. IMO it is how myths develop ...

Ok. I get what your saying. But a Quad is a totally different design than a dynamic driver. I have never heard one. I have heard Magnepan and Genesis. They have a sound too. And they require a lot of power too. And neither had a sense of Black.

 

[Holmz]

Ok, it is your feeling. And how do you analyse it technically?

One would plot the linearity.
If it was a motorcycle fork with “stickion” then that would make the fork not move until the force overcame it.
And any resultant square wave movement of the cone will result in harmonics appearing.
It is all pretty friction free, but maybe the surround/suspension can have some hysterysis.

If a voice coil rubs, that would do it, but that scenario usually sounds pretty bad.

If you take a lot of the arguments about cones being light, coils, and cone area… to the limit… then one could end up with planar/electrostatic/magnetic speakers.

I commented earlier on that. I feel a speaker with a massive crossover and very damped driver is not as able to transmit low level signal to sound as well as a speaker with a much more minimal crossover and a tiny diaphragm. When the speaker is playing at high volumes, both present fine details. It's the quiet passages that are heard differently. Thats just my opinion. Sort of how I hear it.

The more complex the XO and the box, usually the worse the transient response.
And it can take a few cycles before the signal is output to match the input, so it take time to ring up.
When the signals all goes away in the program, it also takes a few cycles for everything to ring down.

Louder sections of the program usually have a lot going on.
A quiet section, by definition, does not have much going on… so everything that is happening then happens out of nothing and it is all transient response up, and transient response down. Maybe some flute solo would be an exception.

I agree on the complex XO presenting as different.
Just that reason for the “not as able to transmit low level signal” has a basis in some engineering.

===

The idea of light weight cones and voice coils being faster appears to be somewhat had to imagine.
VC have been light for decades, and cones have also been light. Cone may be stiffer now, and break up further out of band, but an old paper cone is/was light..
If a driver can play at whatever Hz it is commanded, then by definition it is as fast as another driver playing the same frequency.
Any speed thing often ends up being a way to describe the transient response or the group delay.
Which are more often describing the box and XO than the drivers.

If a signal is extending out (delayed) by taking time to ring up , and then taking more time to ring down, then they are always getting wider and wider and doing “accordion out” filling in time where nothing was actually in the program material... so coming in later and lasting longer.

I have not looked at impulse and transcient response of planer speakers, but I would not be surprised if they were not “sharper” than ported boxes with high order crossovers.

 

[amey01]

I've not yet heard a recording that has a quieter background than the system itself. That is, you can always hear background noise on any recording. Furthermore, as you say - no normal residential environment is noise free anyway.

So while manufacturers love to talk about this - I think it's futile.

Disclosure (or qualifier depending on how you see it) - I have horns, so my system naturally has more noise than most.

 

[tima]

Rex, What do you think causes the effect you hear in systems you describe as "black presenting stereo"? Can one culprit be power delivery, or is it something else?

I have found that certain power distributors and 'noise reduction' power cords that aim at filtering out frequencies their designer finds outside the spectrum of music's frequencies as prime examples of gear that leads to a black(er) background. The typical claim is noise reduction. What I do not understand is how these 'non-musical frequencies' affect our bodies or how reducing those frequencies impacts the frequencies we hear within music's spectrum. But the effect, parsed out as a black background, is real.

 

[microstrip]

I have found that certain power distributors and 'noise reduction' power cords that aim at filtering out frequencies their designer finds outside the spectrum of music's frequencies as prime examples of gear that leads to a black(er) background. The typical claim is noise reduction.

IMO it is a lot more complicated than that. I would risk focusing on changes of distortion components involving changes of extremely low amplitude. But it is my speculation. One thing is sure in my experience - there is no direct correlation between equipment measured signal to noise ratio and "black background" (in the sense Fremer and many known reviewers use it)

What I do not understand is how these 'non-musical frequencies' affect our bodies or how reducing those frequencies impacts the frequencies we hear within music's spectrum. But the effect, parsed out as a black background, is real.

No one understands exactly why and how cables change sound - IMO it is an empirically driven technological business. One thing is sure for me - many of the power devices add noise to systems to create a subjective finding of less noise.

 

[hopkins]

Post #20. I was just randomly listening to this album - The Rosenberg Trio - Suenos Gitanos.

It is pleasant. Borderline elevator music at times. Anyway, it is recorded in such a way that the instruments are all realiy well defined, and form a nice "tapestry" of sound. I would say that the background of the recording is quiet.That's the way it is. You can like it or not. The instruments seem "natural" sounding to me. It's nice to listen to on a "good" system.

Why would a system that performs well with this track and dig deep into the recording not work well with a completely different recording (concert hall type) as well?

Someone please explain this to me like I'm a 6 year old.

 

[Atmasphere]

When I go to Walt Disney Concert Hall I do not hear black backgrounds. I hear a fairly high level of ambient noise, including attendees shuffling around in their chairs, coughing, whispering, sniffling, musicians turning pages, etc.

FWIW if it were a live recording all that would be the part that is projected onto the 'black background'.

What is the difference between black background and low noise floor?

Norman Crowhurst talked about the difference in his writings about 60 -65 years ago although he certainly was not putting it in these terms! A low noise floor is just that: low noise. A black background, in addition to low noise, is a noise floor that is free of intermodulations, inharmonic and harmonic information caused by the electronics beyond just thermionic emission, Johnson noise and the like.

If feedback is improperly applied, it will generate a noise floor like I described. Ground loops can cause this too even if you don't hear an outright buzz, they can cause intermodulations quite easily. Zero feedback circuits can thus have a blacker background if they are also low noise and properly grounded. The exception being that if the feedback is properly applied then the intermodulations and the like tend to be at a much lower level. Avoiding ground loops is a whole 'nuther topic about which tomes have been dedicated. But its one of the reasons I'm an advocate of balanced lines since when they are set up properly you avoid ground loops between equipment. That says nothing about ground loops within the equipment though.

Our ears can penetrate a natural noise source like wind or a waterfall; I've heard by as much as 20dB which I really doubt but 10 dB does not seem a stretch. The hiss of tubes and the like is similar enough our ears can hear detail below that sort of noise floor. But if the noise floor is composed of intermodulations and so on even though it might sound the same, our ears have a lot more trouble delving into it.

So a blacker background will allow for more low level detail including room ambience.

IME this is measurable on an oscilloscope to the degree that if you look at the noise floor and it doesn't look a whole lot like actual hiss you may have a problem. The issue is sorting this out while a signal is playing because the signal has so much more amplitude.

Why would a system that performs well with this track and dig deep into the recording not work well with a completely different recording (concert hall type) as well?

I can't think of why not.

 

[gleeds]

FWIW if it were a live recording all that would be the part that is projected onto the 'black background'.

Norman Crowhurst talked about the difference in his writings about 60 -65 years ago although he certainly was not putting it in these terms! A low noise floor is just that: low noise. A black background, in addition to low noise, is a noise floor that is free of intermodulations, inharmonic and harmonic information caused by the electronics beyond just thermionic emission, Johnson noise and the like.

If feedback is improperly applied, it will generate a noise floor like I described. Ground loops can cause this too even if you don't hear an outright buzz, they can cause intermodulations quite easily. Zero feedback circuits can thus have a blacker background if they are also low noise and properly grounded. The exception being that if the feedback is properly applied then the intermodulations and the like tend to be at a much lower level. Avoiding ground loops is a whole 'nuther topic about which tomes have been dedicated. But its one of the reasons I'm an advocate of balanced lines since when they are set up properly you avoid ground loops between equipment. That says nothing about ground loops within the equipment though.

Our ears can penetrate a natural noise source like wind or a waterfall; I've heard by as much as 20dB which I really doubt but 10 dB does not seem a stretch. The hiss of tubes and the like is similar enough our ears can hear detail below that sort of noise floor. But if the noise floor is composed of intermodulations and so on even though it might sound the same, our ears have a lot more trouble delving into it.

So a blacker background will allow for more low level detail including room ambience.

IME this is measurable on an oscilloscope to the degree that if you look at the noise floor and it doesn't look a whole lot like actual hiss you may have a problem. The issue is sorting this out while a signal is playing because the signal has so much more amplitude.

I can't think of why not.

I agree with Ralph's perspective here. What he described is from a talented electrical engineer and amplification designer with many, many years of real-world listening and design experience. An absence of notice can be measured and heard, a comforting fact. Like other components in the reproduction chain, I also believe loudspeakers have a noise floor resulting from multiple distortions from drivers, crossovers, cabinets, and electrical disturbances.

Ralph, I have had some experience with grounding devices that, as I understand their purpose, equalize ground planes and have heard some surprisingly positive and immediately evident results. Not simply a change in sound but an actual reduction in what I perceive as noise floor - stated as an absence of electronic noise. I understand certain loudspeakers also now come with, e.g., higher-end Audiovector models or are recommended to be used with similar grounding techniques. I am curious to know if you have any direct experience with these devices and, in theory, if you believe this to be an approach to further reduction in noise flow.

Thanks!

 

[Atmasphere]

Ralph, I have had some experience with grounding devices that, as I understand their purpose, equalize ground planes and have heard some surprisingly positive and immediately evident results. Not simply a change in sound but an actual reduction in what I perceive as noise floor - stated as an absence of electronic noise. I understand certain loudspeakers also now come with, e.g., higher-end Audiovector models or are recommended to be used with similar grounding techniques. I am curious to know if you have any direct experience with these devices and, in theory, if you believe this to be an approach to further reduction in noise flow.

I've got no experience with them, but I have a fair amount of experience dealing with grounding issues.

To that end I've found that its essential to get it right, which is something that has a lot to do with how the internal grounding scheme of various bits of equipment is done.

Grounding in high end audio is a bit of the Wild West, especially with equipment that has single-ended inputs. Some manufacturers have sorted out how to do it right so to meet electrical safety standards while preventing ground loops and others have not. When you combine equipment from the 'have' column with something from the 'have not' column you can defeat the grounding scheme of the 'have' side, so all hell can break loose. I've seen audiophiles spend months trying to sort out various buzzes and interference problems, all which could be solved by proper grounding. Unfortunately knowing if a manufacturer has this sort of thing together or not is a bit of a gamble.

The thing is, you don't always hear an obvious buzz when a ground loop is present!

I get called out a lot for my stance on balanced lines but this is one reason I take a stand for doing balanced lines properly. Since a balanced connection (if done properly) ignores ground, it allows for ground loop immunity. That goes a long ways for allowing also for a good noise floor without intermodulations and harmonic junk mixed in.

 

[hopkins]

I've got no experience with them, but I have a fair amount of experience dealing with grounding issues.

To that end I've found that its essential to get it right, which is something that has a lot to do with how the internal grounding scheme of various bits of equipment is done.

Grounding in high end audio is a bit of the Wild West, especially with equipment that has single-ended inputs. Some manufacturers have sorted out how to do it right so to meet electrical safety standards while preventing ground loops and others have not. When you combine equipment from the 'have' column with something from the 'have not' column you can defeat the grounding scheme of the 'have' side, so all hell can break loose. I've seen audiophiles spend months trying to sort out various buzzes and interference problems, all which could be solved by proper grounding. Unfortunately knowing if a manufacturer has this sort of thing together or not is a bit of a gamble.

The thing is, you don't always hear an obvious buzz when a ground loop is present!

I get called out a lot for my stance on balanced lines but this is one reason I take a stand for doing balanced lines properly. Since a balanced connection (if done properly) ignores ground, it allows for ground loop immunity. That goes a long ways for allowing also for a good noise floor without intermodulations and harmonic junk mixed in.

I have no technical understanding of any of this, but my amp's designer explained at length that one of the key features of his design had to do with exactly that - a balanced design that eliminates noise issues.

Anyway, I was able to compare it for several weeks with his previous model which did not have the same design. The difference was striking and my impression was...that the background was blacker... How is it in absolute terms compared to other systems? I'm not going to dare guess. There are too many variables at play.

I will quote his explanations at the time. All I understand is that those aspects are significant in his design...

I also happen to agree with his conclusion. Ok, he's the designer, but he also tells me when his attempts fail...

I am sure that some other designers deal with these aspects in their own way.

Why people would argue about this being beneficial beats me.

How does the new Fractal matrix converter work:

We have same 16V supply voltage and we have the same converter (well 15 fractal bits instead of 31 on the new version but the supply voltage and output impedance remain the same).
However we are going to use some tricks in order to get much more power out of this same circuit, eliminate the ground reference voltage circuit and minimise power consumption when no signal is being generated.

First step is making the audio signal more compact by "folding" it. Digital rectifier is used to flip the negative signal half into a positive signal half, freeing up the DAC range previously occupied by this negative signal.
Now we can double the peak output voltage to 16V peak by shifting all bits one position upwards by hard wiring (fixed setting).

We now only have positive signals representing the already positive signals but also the negative signals (now flipped around so they also become positive).
The positive peak voltage now equals 16V instead of 8V and the negative peak voltage (now turned around) also equals 16V and both still fit nicely in the existing DAC with 16V supply voltage.

When shifting all bits one position upwards, the MSB is no longer connected to the DAC resistor array. The MSB is now used to identify the -original- polarity, so we know what signals were originally positive and what signals were originally negative.
The LSB (of the converter itself, not the input) is no longer needed as all bits move one position upwards. So we only need 17 bits in order to reproduce a 18 bit signal. This relaxes the converter accuracy requirements a bit as less bits allow for slightly higher component tolerances.

First the MSB is used for digital rectifying (to get positive signals only). This is done by only inverting the incoming data when it is negative and passing the data as-is when (not invert it) when the signal is positive.

So we now managed to generate a positive-only output signal at the existing DAC output. The output voltage equals zero volt (not 8V) when no signal is applied. When feeding this signal to the speaker without reconstruction we would get a heavily distorted signal.

So next we have to restore the DAC output signal to its original form and that is handled by the Power switch matrix. It basically consists of an electronic (super fast) double pole, double throw switch (DPDT) controlled by MSB.


Original signal is positive:

+ speaker terminal is connected to DAC output, -speaker terminal is connected to ground. We now have a +16V peak signal -across- both speaker terminals.


Original signal is negative:

+ speaker terminal is connected to ground, - speaker output is connected to the DAC output. We now have -16V peak -across- both speaker terminals.

Note that both speaker terminals work in tandem, each carrying only half the signal.

We now have +16V and -16V peak signal instead of the original -8V and +8V peak signal, we doubled the peak to peak output voltage on the speaker terminals from 16V peak to peak, to 32V peak to peak.

Since we start from zero volts and slowly increase the DAC output from zero volts in tiny steps (instead of flipping many bits around).
The zero crossing is much cleaner and the converter is now much more tolerant to bit errors as these now become fully masked by louder signals instead of sticking out when the signal should have been zero.


Advantages:

- Output voltage doubled to 32Vpp using the same converter and the same supply voltage (output power increases by more than factor 4).
- Because output equals true zero volts (ground potential) all bits are in the same state (off) so idle current is almost zero (minimum power dissipation).
- No ground reference circuit needed as the DAC output already sits at zero volts when there is no signal. The slight degrading of the ground reference circuit is eliminated.
- More tolerant to minor bit errors (few LSBs).
- Fractal converter with fewer fractal bits can be used.

Disadvantages:

- More complex circuit.
- For single-ended output (driving an active subwoofer) we need a balanced to single-ended transformer to convert the signal -between- the speaker terminals into one single ended signal referenced to active subwoofer ground


First sound impressions (limited because I only modified one channel):

- Better dynamics, tight and well controlled bass.
- Airy and clear trebles.
- Improved separation of both, instruments and voices.
- Improved realism.
- Can play -very- loud (calculated 120dB peak for one channel, 126dB peak for stereo), so it should be able to drive less sensitive speakers

 

[Holmz]

I agree with Ralph's perspective here. What he described is from a talented electrical engineer and amplification designer with many, many years of real-world listening and design experience. An absence of notice can be measured and heard, a comforting fact. Like other components in the reproduction chain, I also believe loudspeakers have a noise floor resulting from multiple distortions from drivers, crossovers, cabinets, and electrical disturbances.
…

The speakers do not play sounds when there is no power to them.
But yeah - any sounds being played could have “noise floor raisers” like IMD, cabinet and basket ringing, chuffing, etc.

In terms of distortions, I suspect that if there are room reflections that are causing “musical wraiths”, and the system has time distortions and smearing of the signal with poor transient response, then there gets be a more perfect storm possible.
If the room is super dead, and the transient response super sharp then it is probably easier to have it be dark/black.
But a livelier room with more diffusion can still likely not have obvious “musical wraiths”.

There is a lot going on, and I suspect that it involves more than the electronics .

I have no technical understanding of any of this, but my amp's designer explained at length that one of the key features of his design had to do with exactly that - a balanced design that eliminates noise issues.

Anyway, I was able to compare it for several weeks with his previous model which did not have the same design. The difference was striking and my impression was...that the background was blacker... How is it in absolute terms compared to other systems? I'm not going to dare guess. There are too many variables at play.
..:

It is not a stretch to assume to assume that one could use measurements.
Yeah subjective listening is right next to just listening and enjoying the system. But one could break out a scope or some data capture device when there is a problem.
Or one can use those tools in order to set up a system.

Other than one problem with house wiring and a TT… It was only when I got a DAC and had horrible ground loop issues when I needed to figure things out in a structured way.
But since then we have had a couple of moves, and it certainly seems like it can be a help to run cables and ensure that the noise floor is as low as possible.
It still needs to sound good, but it usually doesn’t sound great when there is obvious “extras” on the lines to start with.

I suspect that the amp designer uses measurements.

 

[matthias]

First sound impressions (limited because I only modified one channel):

- Better dynamics, tight and well controlled bass.
- Airy and clear trebles.
- Improved separation of both, instruments and voices.
- Improved realism.
- Can play -very- loud (calculated 120dB peak for one channel, 126dB peak for stereo), so it should be able to drive less sensitive speakers

Are you talking about the prototype of an upcoming new PowerDAC?

 

[hopkins]

Are you talking about the prototype of an upcoming new PowerDAC?

No the current model (SX) versus the previous one