Toward a Theory To Increase Mutual Understanding and Predictability

[morricab]

Ron,

Your goal is to assemble the best sound reproduction money can buy. I would assume that system would have a low noise floor and would use power/grounding components to achieve a low noise floor. I beg to differ about tube equipment as there are some with excellent grounding schemes. Also your Master Built cables are engineered to produce a very low noise floor. I would think to do otherwise is not getting value for your monetary outlay. You already have decided on the importance of a low noise floor,you might not realise that.

In fact a very low noise floor enhances musicality....period.

And I might add I use a Monarchy NM24 tube DAC with different coupling caps....but I have yet find the bottom noise floor of that piece of gear.

I use a Monarchy M24 DAC (original version) and I too find this to be an excellent DAC. I got a lower noise floor by using the Monarchy power regenerator and DIP. The DAC really loves a low jitter signal and clean power.

 

[morricab]

I agree with you as far as these points go!

I am very conscious of wanting to achieve a lower noise floor; hence my focus on grounding and balanced interconnects for a 45 foot run and the like.

I just never thought of a low noise floor as my primary objective of the hobby.

Hi Ron,
It is not but once you clean up the noise the space and silence between notes and performers really opens up more. I use a power regenerator on my sources (DAC and Phonostage) and this helps immensely with soundstage and image separation. I would also run with it on my preamps if it had the capacity. I need another one! Power amps though I would not do...I have heard this damage the sound much more than help it.

 

[morricab]

Ron, the ambient noise from my listening position was just measured at 36 dbz average. With my entire system powered on and at normal listening volume ( Kondo on phono, Atma MA2 MK !!! monos, VTL 8.2. monos and 4 bridged M2.2 monos ) I measured 44dbz average. The best I could do so far and I am gloriously happy for that. As for the rest of your statement I must admit that I cannot follow you. Swings from pppp to ffff occur more often than not in classical music and having some of the p,s drowned is just as horrible as having the the last of the fs distorted and I have always tried to have it all within the medium of what audiophilia and my possibilities can offer. So respectfully but also forcefully as a classical music lover I have to disagree. But I do know what you mean and I think that my example proves, that you can have the best of both worlds. I would never give up tubes for central parts of my system. By the way, I went through a myriad of tubes to get my Kondo quiet, whereas the Lamm frontend incl. phono stage was exemplary.

44db with everything on is good. I was getting something similar in my previous room but with my horns I get a bit of hum and buzz so I am likely higher than that (I haven't measured though to be sure yet).

 

[RogerD]

I use a Monarchy M24 DAC (original version) and I too find this to be an excellent DAC. I got a lower noise floor by using the Monarchy power regenerator and DIP. The DAC really loves a low jitter signal and clean power.

Mr. Poon wrote about my coupling cap conversion using Mundorf....I think they were M-Supreme 2.7 kind of tight space but doable....great bass

 

[morricab]

I think we are talking past each other, rather than disagreeing.

Obviously, if reducing the noise floor brings you closer to a realistic musical experience then I am all for it!

All I said was that focusing on noise to the detriment of the realism of the musical experience does not make sense to me: "A lower noise floor which takes my system further away from what I consider to be realistic reproduction of music does not make any sense to me."

I agree Ron, I would not choose a piece of gear JUST because it was quieter but in other ways audibly inferior.

 

[morricab]

Thanks, RogerD. Maybe we all finally have something in common to put our arms around.

View attachment 31207

Upon viewing this wikipedia graph once more I noticed the caption at the bottom "Signals below the noise floor cannot be measured."

Reading this immediately reminded me of a rather heated discussion I engaged in 4 years ago with Mark Levinson and primarily John Curl.

There, they both expressed their concerns that their professional-grade and professionally calibrated sensitive measuring instruments routinely failed them, yet their ears and their colleagues' ears could easily discern.

More importantly, was John Curl's eventual admission that every last one of his designs (and everybody else's designs) included at least one serious but unknown flaw for which their measuring instruments could not discern.

When describing this unknown flaw from a different perspective, Robert Harley used the term "catastrophic" whereas John Curl used the term "serious".

However, signals below the noise floor are potentially audible. Example, tape hiss on old recordings is noise; however, it is possible to hear sounds that are softer than the noise floor of the tape. Our brains can extract this information because it is correlated with signals above the noise floor. A highly resolving system will be able to dig below the noise on the program material...if the system itself is not imposing an even higher threshold on the sound.

Now, where this is NOT true is when the noise floor itself is correlated with the signal (not the case with true noise like tape hiss). How does this happen? With negative feedback. Crowhurst demonstrated that negative feedback creates a signal correlated "noise" floor that will obscure low level information because the "noise" itself is correlated with the signal amplitude.

Curl's flaw was likely reliance on the use of negative feedback to "linearize" the signal.

 

[microstrip]

Hi Ron,
It is not but once you clean up the noise the space and silence between notes and performers really opens up more. I use a power regenerator on my sources (DAC and Phonostage) and this helps immensely with soundstage and image separation. I would also run with it on my preamps if it had the capacity. I need another one! Power amps though I would not do...I have heard this damage the sound much more than help it.

You can get an used cheap pro 500W power amplifier, a similar power 2x40 power transformer and a 50Hz sine oscillator - I have used such system to power sources with success! The performance is quite similar to that of the PS Audio P10 when used just for sources (the P10 is however much more powerful).

 

[morricab]

No. What is measured is not what is heard. See this article I wrote for Widescreen Review Magazine: http://audiosciencereview.com/forum/index.php?threads/dynamic-range-how-quiet-is-quiet.14/

Your photography example is all wrong I am afraid. The first one is showing what happens when your shutter speed is too slow for the subject movement. And the second the effect of depth of field. Neither has anything to do with noise.

Indeed, noise is sometimes added to overly soft photographs as to give them the illusion of higher sharpness. Here is a crude attempt at that where I used the out of focus shot you posted (on left) and added some noise to it (right):

Notice how much sharper the right image now looks and has more "3-D pop."

The reason for that is that noise has high-frequency spectrum and that is what is missing in a soft picture. By adding it in, even though it is wrong and artificial, it changes that perception visually.

Not telling you to go and add noise to your system but be careful about using analogies that are not correct.

A better example would be the artifacts one gets with cheap digital cameras for low light photography. They use inexpensive CMOS detector elements that are far noisier than the high quality CCD elements (that work well down to single photon counting when cooled properly). The cheap camera has a very grainy image due to noise artifacts interfering with the picture. The CCD much less so.

 

[DaveC]

I agree that all else being equal, a lower noise floor has the potential to make for substantial improvements. The details a lower noise floor uncovers improves soundstage and imaging as well as allowing you to hear subtle details in vocals and acoustic instruments.

With cables, the key is simply a heavier ground connection to reduce resistance between component grounds. Shielding reduces noise but also makes the sound dull as capacitance is increased, this is not usually a good trade-off.

Grounding is very important and often overlooked especially if multiple AC lines are used, plus the new Furutech NCF parts seem to work very well wrt reducing noise. AC power and grounding are VERY important towards reducing noise:

I had DaveC replaced the IECs and OUTLETs on my Weizhi PRS-6s with Furutech NCFs. As expected, the sound is more organic, lower noise floor ... very happy with the results. Still can't believe outlets can make such a difference.

DaveC also added a ground post to each Weizhi. The most remarkable improvement is tying the Weizhis together with his ground cable. It removed layers and layers of veil. Passages that were once congested and muffled are now crystal clear. Sound is more 3D, airy, tighter imaging, resolving, vivid ... I can crank up the volume several notches without any harshness.

http://www.whatsbestforum.com/showt...s-this-morning&p=436956&viewfull=1#post436956

 

[NorthStar]

Very interesting example Amir. That's probably why LPs and tapes sound more 3D immersing too...?

 

[RogerD]

I agree that all else being equal, a lower noise floor has the potential to make for substantial improvements. The details a lower noise floor uncovers improves soundstage and imaging as well as allowing you to hear subtle details in vocals and acoustic instruments.

With cables, the key is simply a heavier ground connection to reduce resistance between component grounds. Shielding reduces noise but also makes the sound dull as capacitance is increased, this is not usually a good trade-off.

Grounding is very important and often overlooked especially if multiple AC lines are used, plus the new Furutech NCF parts seem to work very well wrt reducing noise. AC power and grounding are VERY important towards reducing noise:



http://www.whatsbestforum.com/showt...s-this-morning&p=436956&viewfull=1#post436956

Single point grounding is by far the best solution but not used in home audio. It is a best solution because it creates a separate pathway back to ground,allowing IC's to carry the audio signal. I hear it everyday....much more information without the downside you mention.

 

[DaveC]

Single point grounding is by far the best solution but not used in home audio. It is a best solution because it creates a separate pathway back to ground,allowing IC's to carry the audio signal. I hear it everyday....much more information without the downside you mention.

What do you consider "single point grounding"? I think this is often done in home systems, especially when you plug your entire system into a single power distributor, or have the grounds of your distributors connected with a low impedance ground cable.

Downside? The only downside I mentioned was shielding and unless you have identical cables with and without shielding, measure them and listen to them, it's not going to be apparent. Issues are often identified only by their absence...

 

[Rodney Gold]

I built my room to be a silent one.. round 20db with all equipment on..it makes a big difference tho I didnt make it my primary objective in the build.. but if you are doing a dedicated room its definitely a big plus and is not that difficult to include. An even bigger plus is that if its silent , it has to be soundproof... meaning you can play louder without hassling others

 

[DaveC]

I built my room to be a silent one.. round 20db with all equipment on..it makes a big difference tho I didnt make it my primary objective in the build.. but if you are doing a dedicated room its definitely a big plus and is not that difficult to include. An even bigger plus is that if its silent , it has to be soundproof... meaning you can play louder without hassling others

20 dB is very impressive, congrats!

My room is decently quiet, 30 dB or so and doesn't change with the system being on or off... would love to have a dedicated room and cut it down 10 dB though!

 

[RogerD]

What do you consider "single point grounding"? I think this is often done in home systems, especially when you plug your entire system into a single power distributor, or have the grounds of your distributors connected with a low impedance ground cable.

Downside? The only downside I mentioned was shielding and unless you have identical cables with and without shielding, measure them and listen to them, it's not going to be apparent. Issues are often identified only by their absence...

high frequency rolloff.....anyway,look @ the single point section. Single point is standard in recording studios...not home audio

http://audiosystemsgroup.com/InfoComm-Grounding2012.pdf

http://www.interfacebus.com/How_to_Specify_an_Equipment_Chassis-Rack-Grounding.html

 

[microstrip]

I built my room to be a silent one.. round 20db with all equipment on..it makes a big difference tho I didnt make it my primary objective in the build.. but if you are doing a dedicated room its definitely a big plus and is not that difficult to include. An even bigger plus is that if its silent , it has to be soundproof... meaning you can play louder without hassling others

20dB with what kind of weighting? It makes quite a difference, usually people refer to dBA.

We can not separate our subjective perception of noise in a room from its RT60 versus frequency - when we enter the room we are immediately analyzing our own noise versus room noise.

 

[stehno]

However, signals below the noise floor are potentially audible. Example, tape hiss on old recordings is noise; however, it is possible to hear sounds that are softer than the noise floor of the tape. Our brains can extract this information because it is correlated with signals above the noise floor. A highly resolving system will be able to dig below the noise on the program material...if the system itself is not imposing an even higher threshold on the sound.

Now, where this is NOT true is when the noise floor itself is correlated with the signal (not the case with true noise like tape hiss). How does this happen? With negative feedback. Crowhurst demonstrated that negative feedback creates a signal correlated "noise" floor that will obscure low level information because the "noise" itself is correlated with the signal amplitude.

Curl's flaw was likely reliance on the use of negative feedback to "linearize" the signal.

Morricab, I'm way outta my league here as with most things, but with all due respect:

1. I always thought tape hiss was not a noise floor threshold but rather an audible noise generated by the friction induced by the physical tape lightly brushing against the recording heads???? If there's any truth to that, then I can't imagine tape hiss in and of itself being a noise floor threshold but rather just another audible noise in a certain frequency range for which indeed we could and should hear other info above and below and maybe even right smack dab in the middle of it.

2. I know you're an intelligent gent so I wanna be careful as not to offend or pretend I am, but I along with some intelligent gents like yourself have held the belief that information below the noise floor threshold cannot be heard. And I hate to use this because it can have the appearance of circular logic, but it would seem this supposed inability to hear anything below the noise floor belief might be somewhat substantiated at least to some degree by the Wiki graph caption stating "Signals below the noise floor cannot be measured." and I would suggest substituting measured for heard.

3. As you offer a potential reason for Curl's serious but unknown flaw, please bear in mind Curl made it abundantly clear that he was speaking not only of all his own designs but also ALL of his colleagues' designs including Mark Levinson who was present yet made no attempt to refute Curl's statement. IMO, and being front and center in that discussion, I interpreted Curl's admission of serious but unknown flaws as being a perplexing universal matter for every designer he communicated with and all who may address negative feedback in various and differing ways. As I recall everybody engaged in that discussion seemed to have much the same interpretation / understanding I had.

 

[Rodney Gold]

I think it was a weighted on my cell app, tho there is no reason not to believe it , it is a room within a room
The room is also treated with bass traps , diffusion , absorption etc
I have no fans or anything that creates any other noise other than audio in the room
Even my hard drives were chosen for their low noise levels.

The most frequent comment of any audiophile entering the room is of its silence and the fact their voice sounds so natural.

 

[RogerD]

Morricab, I'm way outta my league here as with most things, but with all due respect:

1. I always thought tape hiss was not a noise floor threshold but rather an audible noise generated by the friction induced by the physical tape lightly brushing against the recording heads???? If there's any truth to that, then I can't imagine tape hiss in and of itself being a noise floor threshold but rather just another audible noise in a certain frequency range for which indeed we could and should hear other info above and below and maybe even right smack dab in the middle of it.

2. I know you're an intelligent gent so I wanna be careful as not to offend or pretend I am, but I along with some intelligent gents like yourself have held the belief that information below the noise floor threshold cannot be heard. And I hate to use this because it can have the appearance of circular logic, but it would seem this supposed inability to hear anything below the noise floor belief might be somewhat substantiated at least to some degree by the Wiki graph caption stating "Signals below the noise floor cannot be measured." and I would suggest substituting measured for heard.

3. As you offer a potential reason for Curl's serious but unknown flaw, please bear in mind Curl made it abundantly clear that he was speaking not only of all his own designs but also ALL of his colleagues' designs including Mark Levinson who was present yet made no attempt to refute Curl's statement. IMO, and being front and center in that discussion, I interpreted Curl's admission of serious but unknown flaws as being a perplexing universal matter for every designer he communicated with and all who may address negative feedback in various and differing ways. As I recall everybody engaged in that discussion seemed to have much the same interpretation / understanding I had.

If the recording system's noise level is lower then the playback system more information will be inaudible. I believe that it is imperative that a playback system's noise level needs to be equal to the recording. That can be problematic because there are many variables on both sides. I always engineer the largest path to ground that way the lowest noise floor reveals the most fidelity.

 

[DaveC]

If the recording system's noise level is lower then the playback system more information will be inaudible. I believe that it is imperative that a playback system's noise level needs to be equal to the recording. That can be problematic because there are many variables on both sides. I always engineer the largest path to ground that way the lowest noise floor reveals the most fidelity.

I don't see how having all components plugged into a single power distributor isn't single point grounding?

Also, pathway to ground doesn't matter at all, it's on the first couple pages of the pdf you linked to. What is important is reducing resistance between component grounds.

And on shielding, measurable high frequency roll off isn't the issue, but dull sound still is. I'm not totally sure why, but I'm certainly not alone in noticing the issue.