How to put noise in perspective?

[klettermann]

The subject of noise has been consuming me lately, and coming from several directions. First, I recently built a pretty damn quiet media room. Little noise gets, little gets out. That's my starting point. Very nice!

After getting things set up I started going after second order effects and realized that my DAC had a pretty annoying hum. Not loud enough to hear over music but enough to excite room modes that could be heard walking around. That got improved immensely by a DC blocker and chassis damping.

Then a few visits to another site (less measurement focused than ASR) subjected me to ethernet noise gobbledegook. Of course nobody could actually hear it, much less measure it, but nevertheless took great measures to eliminate it. Ditto other well know snake oil nonsense.

Finally, I started fooling around with my own measurements using REW. I COULD detect the DAT hum. I also discovered that my LED track lighting system was creating very measurable room noise (albeit at a very low level, though a soft buzz can be heard in places).

All this lead me to scratch my head and think WTF? People eliminating noise they can't hear or even detect? Others chasing noise orders of magnitude below room noise? The whole topic seems like a cocktail of weirdness. If this goes anywhere I'll post some of my measurements.

Maybe somebody can suggest a rational, measurement-based way to think about it and approach it? To be clear, I'm not talking about electrical noise getting amplified and played through the speakers. Thanks and cheers.

 

[sergeauckland]

That's something I have been conscious of for a long time. Not just ambient noise, but noise and distortion of all sorts. People sometimes obsess about tiny levels of noise and distortion, yet ignore what's actually audible. I find this almost! as bad as the objectivists who claim to hear the inaudible, like cable, amplifier or even DAC differences.

I can accept that good numbers are indicative of good design, but don't let this cloud reality. Do you really have to disregard a device with -90dB noise over one with -110dB noise when your ambient is 40dB SPL? Just get real!

Any noise below -60dB is going to be subjectively silent when playing music, so choose on the basis of facilities, looks and price, not meaningless specs.

S.

 

[DVDdoug]

Yeah, there are lots of sources of noise and some people are more bothered than others. If you are recording, of course you want to eliminate all noise as much as possible.

"Psychologically", I'm more bothered by noise coming-out of my speakers whereas I can ignore a little normal ambient noise. I can hear my furnace when it cycles-on, and my refrigerator which is near my living room home theater when it cycles-on, plus outdoor noises. And the clothes washer & drier when it cycles-or, or the microwave oven if it's running. I'm near a coupe of freeways and I can hear something like white or pink noise. Oh... A jet is just-now flying over. It's not loud but clearly audible. Years ago when I lived in an apartment closer to a military airport, planes landing in the evening were loud-enough that I had to turn-up the TV. Funny, but I don't really remember it bothering me with music.

I grew-up with vinyl and the clicks & pops always annoyed me. Most people didn't complain unless the record was damaged to the point of it "skipping" or "getting stuck". The DJs on the radio were "pretty careful" and they rarely played records with clicks & pops but of course there was FM hiss and AM just sounded terrible with music. I could "live with" the hiss unless it was a distant station with a lot of it.

I WISH my brain could ignore nose from the audio system like it ignores ambient noise!

 

[AndyBell]

Yeah, there are lots of sources of noise and some people are more bothered than others. If you are recording, of course you want to eliminate all noise as much as possible.

"Psychologically", I'm more bothered by noise coming-out of my speakers whereas I can ignore a little normal ambient noise. I can hear my furnace when it cycles-on, and my refrigerator which is near my living room home theater when it cycles-on, plus outdoor noises. And the clothes washer & drier when it cycles-or, or the microwave oven if it's running. I'm near a coupe of freeways and I can hear something like white or pink noise. Oh... A jet is just-now flying over. It's not loud but clearly audible. Years ago when I lived in an apartment closer to a military airport, planes landing in the evening were loud-enough that I had to turn-up the TV. Funny, but I don't really remember it bothering me with music.

I grew-up with vinyl and the clicks & pops always annoyed me. Most people didn't complain unless the record was damaged to the point of it "skipping" or "getting stuck". The DJs on the radio were "pretty careful" and they rarely played records with clicks & pops but of course there was FM hiss and AM just sounded terrible with music. I could "live with" the hiss unless it was a distant station with a lot of it.

I WISH my brain could ignore nose from the audio system like it ignores ambient noise!

There’s an odd psychology to this, isn’t there?

I’m a light sleeper and get irritated when my neighbours are noisy at night because they shouldn’t be.

But if the birds sing, the foxes bark, the cats fight, or the traffic from the adjacent highway is audible, my mind filters it out.

Same with background noise when I’m listening to music - I generally ignore it. But if I get a pop or click from the speakers, that’s another matter.

 

[Keith_W]

It is very easy to measure your room's noise floor.

1. In REW adjust your microphone with an SPL meter using REW's SPL adjustment procedure.
2. Mount your mic on a tripod and leave it undisturbed.
3. Turn on the RTA and let it keep going until you see that the curve stops changing. Then stop it.

This is the noise floor of my living room taken on a weekend afternoon. My listening room has a large sliding door that opens to my backyard, the green/yellow measurements show the difference between open/closed.

The typical noise floor of a listening room rises at low freqs for a few reasons. Rooms act as low-pass filters which you can see easily in this measurement. The difference between open/closed doors is not that great below 100Hz, about a 5dB difference. But there is up to a 15dB difference with the doors open at 1kHz. Even though I live in a glorified tent (typical rubbish cheap Australian capital city home construction) my noise floor is commendably low with the doors closed. That's because I some simple measures to seal all the air gaps - foam gaskets in all windows and doors. It not only cuts the noise down by quite a lot, it also reduces heating bills.

 

[klettermann]

It is very easy to measure your room's noise floor.

1. In REW adjust your microphone with an SPL meter using REW's SPL adjustment procedure.
2. Mount your mic on a tripod and leave it undisturbed.
3. Turn on the RTA and let it keep going until you see that the curve stops changing. Then stop it.

View attachment 501989

This is the noise floor of my living room taken on a weekend afternoon. My listening room has a large sliding door that opens to my backyard, the green/yellow measurements show the difference between open/closed.

The typical noise floor of a listening room rises at low freqs for a few reasons. Rooms act as low-pass filters which you can see easily in this measurement. The difference between open/closed doors is not that great below 100Hz, about a 5dB difference. But there is up to a 15dB difference with the doors open at 1kHz. Even though I live in a glorified tent (typical rubbish cheap Australian capital city home construction) my noise floor is commendably low with the doors closed. That's because I some simple measures to seal all the air gaps - foam gaskets in all windows and doors. It not only cuts the noise down by quite a lot, it also reduces heating bills.

Thanks @Keith_W . That's the kind of stuff I'm starting to explore. Below are a couple of my first attempts using REW for this purpose. I didn't do the spl adjustment, I'll do that next time. Anyway, the LED lights are pretty obvious though I can't hear them when music is playing. But the room is pretty quiet so if nothing is playing you can hear such odd buzzez, hums etc If I can track then and kill them it would be nice just on principle. FWIW, my room is kind of the opposite of yours - sealed, mass-damped, double sheetrock on hat-channel/RISC clips, sealed electrical outlets etc etc. I can hear the blood in my ears. All that said, any comment on these traces? Do I seem to be doing it right? Cheers,.


1. LED lights off

2. LED lights on

 

[Keith_W]

That seems to be what a typical noise floor looks like, yes. Although I would have let it run a lot longer than 20 averages. But as long as the curve stops changing shape it's fine.

A SPL meter is mandatory for this procedure. Otherwise your measurement might mislead you.

 

[klettermann]

Great, thanks much. Eventually I'll be back with s lot more to discuss.

 

[klettermann]

That seems to be what a typical noise floor looks like, yes. Although I would have let it run a lot longer than 20 averages. But as long as the curve stops changing shape it's fine.

A SPL meter is mandatory for this procedure. Otherwise your measurement might mislead you.

I'm back already. I'm using a calibrated mike. REW days that in this case further calibration isn't need, it's already done (by the cal file I guess). Make sense? I know you use ,Accurate which I suppose is different. Cheers,

 

[Keith_W]

When you say "microphone calibration", be precise about what you are talking about. There is calibration that only corrects deviation from a non-flat response but does not measure absolute SPL. These will need to be used with an SPL meter. And then there are USB mics which are pre-calibrated for absolute SPL and do not need to be used with an SPL meter.

I use both Acourate and REW. The programs do different things. Doing sims and designing filters is a breeze in Acourate, because that's what it was designed to do. But it's not as good as REW for "room" measurements. Also, since most people on ASR use REW, I also use REW. Otherwise I can't talk to you

 

[klettermann]

When you say "microphone calibration", be precise about what you are talking about. There is calibration that only corrects deviation from a non-flat response but does not measure absolute SPL. These will need to be used with an SPL meter. And then there are USB mics which are pre-calibrated for absolute SPL and do not need to be used with an SPL meter.

I use both Acourate and REW. The programs do different things. Doing sims and designing filters is a breeze in Acourate, because that's what it was designed to do. But it's not as good as REW for "room" measurements. Also, since most people on ASR use REW, I also use REW. Otherwise I can't talk to you

LOL, certainly part of the problem is that I DON'T know what I'm talking about - which is why I'm here. Anyway, I looked at the REW help for measuring room spl. It explained that with a calibrated mike and calibration curve, such as umik1, that it was good to go. Make sense?

 

[DVDdoug]

Note that our ears are most-sensitive to mid-frequencies. Fortunately, most "natural noise" has more energy in the low frequencies where it's harder to hear. 50 or 60Hz power-line hum is harder to hear than hiss or "whine" (although power line hum often contains more-audible harmonics and then it sounds more like a "buzz").

Most SPL measurements are A-Weighted to (partially) compensate for that and more-accurately represent what we hear. Almost all SPL meters are A-weighted or they have a setting. A-Weighting always gives you a better (lower) number.

I can see that REW is including dB-A and dB-C.