View Full Version : What is a "Low Noise Floor"
I read a lot of "audiophilisms" on these boards and one I often read is that a component has a "low noise floor". What is that is supposed to mean? Are you turning the V/C up all the way with no source and if you don't hear static (something relatively easy to accomplish even on a $300 receiver) calling it a "low noise floor"?
AV Doogie 10-25-07, 08:04 PM That would be a sturdy concrete floor upon which running children can not be heard :)
If tweakers are talking about this subject, I believe it refers to the absolute lack of hash and/or white/pink noise in the output .
When I talk about it, a low noise floor means something with a signal to noise ratio better than say 80db (maybe greater?).
speco2003 10-25-07, 08:13 PM Well it applies to all audio devices. In that the area from noise floor to max output (clipping) is the systems dynamic range. If say you have a device with a -70 floor and max output of +3 then you have 73 db of system. If in this same chain you have a device that has a floor of -70 but max out of +20 you still only have 73 db of gain because you can only go as high as your shortes max output. Otherwise you clip. Now there are work arounds of this by using pads that attenuate. Which is what happens in a mixing desk for instance
The noise floor of a device is really where the lowest usable signal is.
Mark Seaton 10-25-07, 08:19 PM Actually, when you start working with rooms that are designed to be very quiet, and if you have loudspeakers that are of higher efficiency, this can become a real challenge. We tend to take it for granted in the consumer world as the background noise level in most homes is rather high, but I know that Keith Yates has repeatedly mentioned that the loudspeakers and electronics tend to set the noise floor in his room designs, although his are pretty extreme cases. I'm sure this is something that can become noticable in any room that is targeting a very low NC rating and has the projector properly isolated.
Steve Bruzonsky 10-25-07, 08:34 PM Subjectively for home theater - when you play movies and tv, you can listen at a lower level yet not feel your missing anything when it comes to voices, vocals, music, micro and macrodynamics. This causes the system to sound more tranparent and natural like the real thing.
Objectively - whatever Mark Seaton says!!!@@
CINERAMAX 10-25-07, 08:47 PM I was staying at the Eurobuilding hotel doing my system design and watching Von Ryan's Express on the telly. At one point the machine gun sound became super realistic, it sounded that good until a shot went through the window. As it turned out the air force base across the street was under attack by General Chaves forces, and there was a C-130 with side mounted machine guns shooting into the airbase itself. This was Chavez first attempted Coup De Tat. The idiot insurgents took all the TV stations but instead of recording Chavez takeover message on BETACAM they recorded it in BETAMAX, so the tapes could not air and the presidential forces quelled the rebellion.
What does that have anything to do with noise floor? Not much except earlier, from the same room I had called Mark Levinson at his Cello atelier, and when he got on the phone he told me: " I am sitting here listening to 30,000 watts of silence."
Raul GS 10-25-07, 10:16 PM This was Chavez first attempted Coup De Tat.
Either the Grey Goose was helping you time travel, or you lost a decade of your life, because Chavez was the President then.
Actually, when you start working with rooms that are designed to be very quiet, and if you have loudspeakers that are of higher efficiency, this can become a real challenge.
I do not see why higher efficiency loudspeakers would make it more challenging.
CINERAMAX 10-25-07, 10:25 PM Either the Grey Goose was helping you time travel, or you lost a decade of your life, because Chavez was the President then.
Raul During February of 1992 Carlos Andres was the president. Oh I see lost ten years.;)
Back on topic:
LOWER NOISE FLOOR is what I experience when I use an Equitech, better dynamics, more robustness.
kucharsk 10-25-07, 10:34 PM I do not see why higher efficiency loudspeakers would make it more challenging.
A more efficient loudspeaker requires less power to produce sound, and thus any hiss or noise in the electronics chain's noise floor becomes obvious.
So for example a low level hiss or hum that might be inaudible in a pair of speakers with a sensitivity of 88 dB @ 1 watt/1 meter might be offensive in a quiet room when coupled with speakers with a sensitivity of 92 dB or higher.
Ask anyone with extremely high efficiency speakers like Klipsch's La Scala or Klipschorn about how important matching electronics can be.
sdurani 10-25-07, 10:41 PM ..."low noise floor". What is that is supposed to mean?Shag carpet + thick padding = "low noise" floor.
When used to describe audio devices, the term is just short-hand for a quiet component (some are more quiet than others when no signal is passing through).
Sanjay
Edit: was responding to kucharsk.
If the system is properly isolated and free of ground loops (and the components properly designed), the "noise floor" should be very low. Are you suggesting that their is noise coming from the speakers even when the volume is turned all the way down? Because that would be a noisy system indeed and yes in that instance the noise would be higher through an efficient speaker. But otherwise, in a system without ground loops or stray noise, since the volume control on the preamp would be adjusted to a compensatory level on a system with less efficient speakers, the noise level should be the "similar".
p.s. I put "similar" in quotes as depending on a number of factors "small" might become "large" and semantics comes into play.
mike lavigne 10-26-07, 01:26 AM a few misc thoughts on low noise.
i do throw around the term 'low noise floor' or 'lower noise floor' alot because many of the approaches i have taken have brought dividends in this area. when comparing live music (sound) to recorded/reproduced music (sound) one big clue to which is more real is that live music (sound) has a totally natural ambience (noise floor) whereas recorded/reproduced music (sound) can only approach this ideal to a degree. the difference between a real ambience and reproduced ambience is the recording/reproduction chain's noise floor. and remember......the only way to reproduce the event is to be able to reproduce the ambience of that event.....which means that there should be as little room ambience of the listening room as possible.
when noise is reduced in a system the biggest benefit seems to be increased dynamics and ease.....the sound is more alive and real.
the way phono cartridges and phono preamps work makes it easy to understand these issues. many MC (Moving coil) cartridges have output levels as low as .1mv......some as high a 1mv. typically, the lower output cartridges have more sensitive coils and will produce more nuanced and complex musical information......but they must be matched with phono preamps that are quite high in gain but very low in noise......so when you amplify that very small signal the resulting sound will have a very low noise component. it is relatively easy to make a very low output cartridge be very detailed.......but to make a high gain, low noise phono stage is very challenging. conversly, a high gain cartridge will not need as much gain from the phono preamp to keep the noise low, but it will be challenged to be as detailed and nuanced as the lower output cartridge.
always, low noise allows for more music and less 'other stuff'.
a couple of years ago when i got my battery powered darTZeel NHB-18NS preamp i experienced the benefit of getting away from the noise floor of the power grid and onto batteries. first; batteries are not right for power amps IMHO.....but for a preamp they are perfect.
the biggest benefit was clear increase in dynamic contrasts and more a sense of venue and space. the same reason that SACD gives you so much more of a soundstage than redbook.....lower noise floor = morer space and greater 'boogie-factor' to the music.
then there is the issues of 'relative noise' and 'absolute noise'. maybe read these links........
http://en.wikipedia.org/wiki/Approximation_error
http://en.wikipedia.org/wiki/Dynamic_range
http://en.wikipedia.org/wiki/Signal-to-noise_ratio
http://www.teamcombooks.com/mp3handbook/11.htm
finally; my room is on 4 acres in a separate building from my home......then it is actually a room built inside a room......so it is truely DEAD QUIET. i have a purpose built HVAC system that is 'almost' completely unaudible.....so i can keep the room closed and yet comfortable.
this very low noise environment allows for the music to reveal it's nuances and textures as completely as the media allows.
Terry Montlick 10-26-07, 09:30 AM Well it applies to all audio devices. In that the area from noise floor to max output (clipping) is the systems dynamic range. If say you have a device with a -70 floor and max output of +3 then you have 73 db of system. If in this same chain you have a device that has a floor of -70 but max out of +20 you still only have 73 db of gain because you can only go as high as your shortes max output. Otherwise you clip. Now there are work arounds of this by using pads that attenuate. Which is what happens in a mixing desk for instance
The noise floor of a device is really where the lowest usable signal is.
For reference, a plain old 16-bit CD can record 96 dB of dynamic range. If the peak volume is say, 106 dB, and your equipment can handle 96 dB of signal to noise (child's play for most consumer gear these days), then the noise floor is theoretically 10 dB. You would need to be in someplace like Galaxy Studios in Belgium (designed by my friend Eric Desart), which has an acoustic noise level of only NC 5 (~= 5 dB), to hear this. So for any modern audio equipment which is set up properly, the room (yours or the recording studio's) will determine the audible noise floor.
Regards,
Terry
"But otherwise, in a system without ground loops or stray noise, since the volume control on the preamp would be adjusted to a compensatory level on a system with less efficient speakers, the noise level should be the "similar"."
Only for noise originating ahead of the pre-amp. Most pre-amps have volume controls at their input, not their output. So the noise level of the pre-amp itself will vary depending upon the system it is installed in. Amps with high input sensitivities/gain will make it worse then when used with lower gain amps or amps that have a lower input sensitivity. Same thing with speakers.... the more efficient they are the more you will hear the noise of the pre-amp and/or amps.
And of course someone with a room that is say in the 30 dB range for noise floor has a totally different reference about 'noise' then someone who has a noise floor in the 50 dB range.
Shawn
chirpie 10-26-07, 09:55 AM "Low noise floor" = the only sound environment I can enjoy orchestral music.
^_^
AndreYew 10-26-07, 11:40 AM When talking about room noise, one has to be careful to specify the spectrum of the noise, too. In most rooms, the highest noise is actually at the low end of the spectrum, say 200 Hz down. This is mostly due to rumble of various appliances, road noise, etc. In the midrange, noise can be significantly lower, so there are some things which are audible that may not be predicted by an averaged reading over 20Hz-20kHz.
--Andre
Mark Seaton 10-26-07, 04:07 PM For reference, a plain old 16-bit CD can record 96 dB of dynamic range. If the peak volume is say, 106 dB, and your equipment can handle 96 dB of signal to noise (child's play for most consumer gear these days), then the noise floor is theoretically 10 dB. You would need to be in someplace like Galaxy Studios in Belgium (designed by my friend Eric Desart), which has an acoustic noise level of only NC 5 (~= 5 dB), to hear this. So for any modern audio equipment which is set up properly, the room (yours or the recording studio's) will determine the audible noise floor.
Regards,
Terry
While I agree this *should* be easy to achieve, reality tends to not be so simple. The other issue is that what you might not easily hear at a listening distance of say 13' (4m) can be close to 12dB louder at 1m. Your example (which I realize you simplified) also assumes that all of the gains in the systems are optimized, and that the entire system maintains that noise level without multiple stages summing to a greater level. You are no doubt aware of the same issues as I mentioned with Keith Yates, as he has been designing (and testing) rooms that are in the NC-5 level... before you turn the speakers on. ;)
Let's not forget that THX is recommending something down around NC-22(?) for a certified theater. A tweeter with a sensitivity ranging from 100-110dB @ 1W/1m over the most sensitive hearing range can certainly make perceived silence a challenge.
I've been starting to poke around more devices and common system configurations to get a better handle on the real situation, but still have a lot too look through. I can tell you it is quite frustrating to know how to deliver greater sensitivity which has many benefits, yet run into this front end/setup complication. Serious question I'm still evaluating: Can the serious HT world work with a true 95-98dB @1W/1m loudspeaker?
Terry Montlick 10-26-07, 05:27 PM All good points, Mark. "... any modern audio equipment which is set up properly" includes a goodly amount of handwaving. ;)
Regards,
Terry
ChrisWiggles 10-26-07, 06:24 PM I am surprised you would be asking something like this Q, and I sense alterior motives, but I don't understand what they might be?
But basically, yes, that when a source calls for silence, the system produces as close to that as possible. The biggest challenge to this is noise that is unrelated to the electronics entering the room, which is where good room isolation can be the most imporatnt part of attaining a quiet (low noise floor) space. But there will always be some amount of noise going on in the electronics, and keeping that as low as possible (high SNR) is the goal. This isn't exceedingly difficult to do at least decently, but having an extremely quiet system can become very difficult the greater the complexity, long cable runs, etc, which you can recognize much more severely with things like concerts and big PA systems and the like.
Of course, this all may be slightly a waste of time depending on the noise floor of your sources, if you're listening to very old recordings that have a moderate noise floor already, then having a super-quiet system may be overkill in those instances, but at least you're not laying over your own system noise.
Terry Montlick 10-26-07, 07:10 PM Just a note specific to movie sound.
There is nearly always an audible noise floor placed right in the sound mix. When dialog is recorded, the production sound mixer records "room tone." This is at least 30 seconds of silence recorded on location or the sound stage, during which cast and crew members are required to be quiet.
A movie sound mix is composed of thousands of little snippets of sound spread across several tracks. There will often be gaps in the dailog track, and room tone is put in to fill these gaps. Otherwise, you would notice the difference between silence in the room, which has a very specific sound, and genuine silence!
- Terry
speco2003 10-26-07, 09:04 PM I am surprised you would be asking something like this Q, and I sense alterior motives, but I don't understand what they might be?
But basically, yes, that when a source calls for silence, the system produces as close to that as possible. The biggest challenge to this is noise that is unrelated to the electronics entering the room, which is where good room isolation can be the most imporatnt part of attaining a quiet (low noise floor) space. But there will always be some amount of noise going on in the electronics, and keeping that as low as possible (high SNR) is the goal. This isn't exceedingly difficult to do at least decently, but having an extremely quiet system can become very difficult the greater the complexity, long cable runs, etc, which you can recognize much more severely with things like concerts and big PA systems and the like.
Of course, this all may be slightly a waste of time depending on the noise floor of your sources, if you're listening to very old recordings that have a moderate noise floor already, then having a super-quiet system may be overkill in those instances, but at least you're not laying over your own system noise.
I can assure the PA in our showroom is deathly quiet. We have over 100 various types of Meyer speakers in it. With all sorts of outboard gear and it does not make any hiss or noise when idle. It has nothing to do with PA vs home or cable runs etc.Much of our pro gear has the same or better noise floor than home gear. Please read and understand what I posted about gear in the earlier post.
Dizzman 10-26-07, 09:37 PM Your gear is setup as well so that it is not even breathing hard in order to make lots of noise. so with good gain staging, one should not be able to hear a bloody sound.
I am surprised you would be asking something like this Q, and I sense alterior motives, but I don't understand what they might be?
Grasshopper, one cannot know what is at the top of the mountain until one has climbed it. But at least you are trying to climb it.
speco2003 10-26-07, 10:07 PM Your gear is setup as well so that it is not even breathing hard in order to make lots of noise. so with good gain staging, one should not be able to hear a bloody sound.
Ahh but I have run into some ****** PA whee the gain stages were all screwed up then you hear that noise floor.
Morbius 10-26-07, 10:28 PM If the system is properly isolated and free of ground loops (and the components properly designed), the "noise floor" should be very low.
QQQ,
Even if you have everything isolated and free of ground loops - EVERYTHING
you can possibly think of - you still have noise.
The transistors and / or vacuum tubes themselves add noise - just because they are
at a temperature above absolute zero. This is known as "thermal noise".
Unfortunately, real world devices are not perfect. "Thermal noise" is but one example.
I don't believe there was a comment in the post you quoted that suggested otherwise.
Mark Seaton 10-27-07, 11:40 AM I can assure the PA in our showroom is deathly quiet. We have over 100 various types of Meyer speakers in it. With all sorts of outboard gear and it does not make any hiss or noise when idle. It has nothing to do with PA vs home or cable runs etc.Much of our pro gear has the same or better noise floor than home gear. Please read and understand what I posted about gear in the earlier post.
Hi speco',
I expect your system is extremely quiet by most standards. I have been bit a few times by noise floors that were only obvious if I literally stuck my ear within inches of a tweeter, yet were obvious in a well constructed, and well isolated room. The context of the observation of noise floor is the real issue. PA actually has some significant benefits that allow you to achieve a rather low noise floor. Most importantly, the electrical signals further down the chain are both much stronger in normal use, and have greater peak capability than consumer gear. With less "room to work" one or two components can quickly make life very difficult in a very quiet HT.
Steve Bruzonsky 10-27-07, 11:45 AM When I use my laptop in my isolated dedicated quiet home theater, and I read one of Speco's AVS posts, I hear a loud whissing sound. Yet when I do this in my den I don't hear anything!!!
Mark Seaton 10-27-07, 11:46 AM Just a note specific to movie sound.
There is nearly always an audible noise floor placed right in the sound mix. When dialog is recorded, the production sound mixer records "room tone." This is at least 30 seconds of silence recorded on location or the sound stage, during which cast and crew members are required to be quiet.
A movie sound mix is composed of thousands of little snippets of sound spread across several tracks. There will often be gaps in the dailog track, and room tone is put in to fill these gaps. Otherwise, you would notice the difference between silence in the room, which has a very specific sound, and genuine silence!
- Terry
Thanks for the additional insight, and that of course makes perfect sense. I think the point enthusiasts should gather from is that we would like HT systems to have a low enough noise floor that you can actually hear/feel/experience the "room tone" that is recorded. While this subtle detail isn't at the top of the list of sonic critiques, it is exactly this sort of recorded information that serves to further draw us into the movie.
The further down in level the "room tone" is of the listening space, the less it sounds like one room tone on top of another, and more like the intended space.
ChrisWiggles 10-27-07, 12:09 PM I can assure the PA in our showroom is deathly quiet. We have over 100 various types of Meyer speakers in it. With all sorts of outboard gear and it does not make any hiss or noise when idle. It has nothing to do with PA vs home or cable runs etc.Much of our pro gear has the same or better noise floor than home gear. Please read and understand what I posted about gear in the earlier post.
Yes, I didn't mean to insinuate that there was anything inherent to PA systems per se, but rather the fact that the kinds of stereo/HT systems you encounter in fixed indoor settings, especially just regular surround systems for HT, are a lot less complex than say a full concert where you have dozens of mics and channels, effects, and hundreds of cables running back and forth across hundreds of feet. And where enormous volume is a primary goal. There are countless places noise can creep into systems like that, which is why usually you have balanced cables running everywhere, something that may often be overkill in a much more controlled environment like a rather straightforward home surround or stereo system. That was really all I was saying, not that the equipment is to blame, just rather the messy nature of such a large system usually assembled under enormous time pressure, and as anyone can attest, usually it's just logistically impossible to keep such a system as dead quiet as you can fairly easily with a simple HT. Of course, it's also not as crucial, as the ringing in your ears if you're not wearing hearing protection is probably a much bigger problem than the hiss in that type of system anyway...
speco2003 10-27-07, 02:58 PM Yes, I didn't mean to insinuate that there was anything inherent to PA systems per se, but rather the fact that the kinds of stereo/HT systems you encounter in fixed indoor settings, especially just regular surround systems for HT, are a lot less complex than say a full concert where you have dozens of mics and channels, effects, and hundreds of cables running back and forth across hundreds of feet. And where enormous volume is a primary goal. There are countless places noise can creep into systems like that, which is why usually you have balanced cables running everywhere, something that may often be overkill in a much more controlled environment like a rather straightforward home surround or stereo system. That was really all I was saying, not that the equipment is to blame, just rather the messy nature of such a large system usually assembled under enormous time pressure, and as anyone can attest, usually it's just logistically impossible to keep such a system as dead quiet as you can fairly easily with a simple HT. Of course, it's also not as crucial, as the ringing in your ears if you're not wearing hearing protection is probably a much bigger problem than the hiss in that type of system anyway...
Uhhh I cant hear you?!?!!?? No very good I understand. I am very lucky in that being in one of the best designed shows on the strip I mix at a very comfortable level at FOH and never have ringing in my ears. And I have also been very lucky in that when I did tour I had a great company behind me. My current employer is very good about paying for my hearing check up every 6 months. And so far after 4 years here I have not lost any yet from my baseline which was very good as well.
I would think that most of the gear we would use in this forum and even alot of the lesser gear, is quiet. I see from the standpoint of just the room noise floor as well. The idle showroom can sometimes be 55db just from air handlers.
tbrunet 10-28-07, 10:53 AM The noise floor is a real world ingredient or variable that prevents a given design from achieving a theoretically perfect dynamic range potential i.e.
16 bits offers a maximum dynamic range of 96dB
A 24-bit sample resolution offers a dynamic range of 144dB
A "Low Noise Floor" is anything downward of ~70dB
kucharsk 10-28-07, 06:10 PM Edit: was responding to kucharsk.
If the system is properly isolated and free of ground loops (and the components properly designed), the "noise floor" should be very low. Are you suggesting that their is noise coming from the speakers even when the volume is turned all the way down? Because that would be a noisy system indeed and yes in that instance the noise would be higher through an efficient speaker. But otherwise, in a system without ground loops or stray noise, since the volume control on the preamp would be adjusted to a compensatory level on a system with less efficient speakers, the noise level should be the "similar".
That's precisely what I'm saying.
Electronics all have an associated amount of amplifier noise even with the volume turned all the way down.
The noise can be minimized through careful design, but more efficient loudspeakers will make it more apparent.
For example, many Carver amplifiers were known for having a hiss to them with most any speaker above 90 dB efficiency - even with the volume turned all the way down.
tek-noid 11-08-07, 03:01 PM Is there a specification that can be used to best determine if a component has a low noise floor? S/N Ratio?
Is there a rule-of-thumb for what is acceptable, desired, [practically] ideal?
Any particular brands better at this than others?
Thanks,
James
Terry Montlick 11-09-07, 08:25 AM Is there a specification that can be used to best determine if a component has a low noise floor? S/N Ratio?
Is there a rule-of-thumb for what is acceptable, desired, [practically] ideal?
Any particular brands better at this than others?
Thanks,
James
Yes, S/N ratio gives you the noise floor -- relative to the max possible signal.
I'd say anything with around 100 dB of S/N for a direct signal (as from a CD player) is just great. And my guess is that all manufacturers have models which achieve this.
Regards,
Terry
zamboniman 11-09-07, 12:56 PM Don't most pro rigs.... studio/mix, etc.. use noise gates? Which would give the "appearance" of lower noise floor when a the system is idle.
BAMAVADER 12-14-07, 09:52 AM I have a Onkyo 390 CD changer that has a very soft hum when off that can only be heard with ear next to the player. Is this an example of floor noise and is it normal?
thanks...
Peter Nielsen 12-14-07, 10:34 AM QQQ,
Even if you have everything isolated and free of ground loops - EVERYTHING
you can possibly think of - you still have noise.
The transistors and / or vacuum tubes themselves add noise - just because they are
at a temperature above absolute zero. This is known as "thermal noise".
Unfortunately, real world devices are not perfect. "Thermal noise" is but one example.
Very true. The culprit is usually the power amplifier. It is difficult to make big amps silent. With inefficient speakers, it may be a non issue, but with highly efficient speakers it may be intolerable.
TacT audio has addressed the problem in a different and ingenous way in their digital amps: When you lower the volume, the voltage in the output stage decreases and consequently the noise floor is lowered too. (In traditional power amps the noise floor is the same regardless of volume).
Peter
|
|