Is it possible to estimate a computers dBA level?

[Silent but deadly]

I'm currently in the process of planning a HTPC build and one of my goals is to make it as quiet as my gaming rig when it is idling. The problem is though is that I don't know exactly how loud my gaming rig is at idle, so I don't know whether the 26 dBA noise level of the system I wish to base my HTPC on (the Lenovo TS140) is more or less than my rig. Fortunately all but one of the components that produce noise has been reviewed by SPCR. With that in mind, would it be possible to estimate how loud it is?

The components that make noise at idle are as follows.

Scythe Kotetsu

I run mine at just over 700rpm to avoid a clicking noise at lower RPMs. According to SPCRs data, the Kotetsu produces 12 dBA @ 670rpm and 14 dBA @ 790 RPM.

Asus GTX 670 DirectCU II TOP

SPCR have reviewed the non-TOP version of the card so their data may not be 100% applicable to the TOP version (which exactly the same aside from a higher clock speed) but I think the data is at least a good estimate. According to said data, the card produces between 13 and 14 dBA at idle.

Noiseblocker NB-eLoop B12-2

According to SPCRs data, the B12-2 produces 11 dBA @ 700rpm and 12 dBA @ 900rpm. Unfortunately the RPM that SpeedFan/my motherboard reports the fan as running at varies (due to some odd behaviour getting the fan to ramp up) but I'd be willing to bet that it's somewhere in the 700 to 900rpm range.

Scythe Gentle Typhoon 1850

SPCR hasn't review the exact model of fan that I have but to my ears, the GT 1850 sounds better and is quieter than both the B12-2 and the Scythe Glidestream on my Kotetsu.


So just to recap.

Heatsink - 12 to 14 dBA
GPU - 13 to 14 dBA
B12-2 - 11 to 12 dBA
GT 1850 - an estimate of 11 to 13 dBA

I know that two or more identical components, producing the same amount of noise will increase the overall dBA by around 3 dBA (please correct me if I'm wrong). Does that also apply to other components that produce roughly the same level of noise but with a different sound profile?

Would it be safe to estimate my computer as idling at less that the TS140's 26 dBA?

[quest_for_silence]

Does that also apply to other components that produce roughly the same level of noise but with a different sound profile?

I guess it doesn't work that way: each of those sources has its own radiation pattern.

By the way, stop all the fans but the Glidestream, crank it up to about 1200-1300rpm, and you'll have an estimate of how much loud about 26dB will sound.

[Reachable]

Go to this site:

https://www.noisemeters.com/apps/db-calculator.asp

and enter the decibel levels one by one in the special calculator (or do the math out with all the horrible Greek letter thingys.)

[quest_for_silence]

Go to this site:

https://www.noisemeters.com/apps/db-calculator.asp

and enter the decibel levels one by one in the special calculator (or do the math out with all the horrible Greek letter thingys.)

It's the same thing as above: it applies to identical sources with the same radiation pattern.

[Reachable]

Go to this site:

https://www.noisemeters.com/apps/db-calculator.asp

and enter the decibel levels one by one in the special calculator (or do the math out with all the horrible Greek letter thingys.)

It's the same thing as above: it applies to identical sources with the same radiation pattern.

True, but I was wondering how much the overall noise level would be reasonably expected to vary from one vantage point to the next. The components the OP is considering calculate out to far less than the limit he set. I'd be inclined to advise going ahead with it. Would that be unwise, or not?

[quest_for_silence]

True, but I was wondering how much the overall noise level would be reasonably expected to vary from one vantage point to the next.

I guess it may vary wildly: if you consider the SPCR GTX 960 review, Lawrence Lee managed to run the whole system at 14dB under load (slightly above the GTX 670 idle level) with a comparable system (actually he used a semi-fanless Kingwin PSU, while we don't know which PSU the OP got), I mean same CPU cooler, somehow similar fans and so on.
On the other hand the calculator says just under 19dB on average.
Which guesswork may be more representative, 14dB or 19dB? Subjectively an almost 5dB difference is well noticeable.
I'd add that we also don't know how Lenovo measured those 26dB, which (per SPCR standards) is a fairly loud sound level.

The components the OP is considering calculate out to far less than the limit he set. I'd be inclined to advise going ahead with it. Would that be unwise, or not?

I'm sorry, I didn't get you.

[Reachable]

True, but I was wondering how much the overall noise level would be reasonably expected to vary from one vantage point to the next.

I guess it may vary wildly: if you consider the SPCR GTX 960 review, Lawrence Lee managed to run the whole system at 14dB under load (slightly above the GTX 670 idle level) with a comparable system (actually he used a semi-fanless Kingwin PSU, while we don't know which PSU the OP got), I mean same CPU cooler, somehow similar fans and so on.
On the other hand the calculator says just under 19dB on average.
Which guesswork may be more representative, 14dB or 19dB? Subjectively an almost 5dB difference is well noticeable.
I'd add that we also don't know how Lenovo measured those 26dB, which (per SPCR standards) is a fairly loud sound level.

The components the OP is considering calculate out to far less than the limit he set. I'd be inclined to advise going ahead with it. Would that be unwise, or not?

I'm sorry, I didn't get you.

Sorry, my fault. I misunderstood the opening post. My advice above would be invalid.

[Silent but deadly]

Does that also apply to other components that produce roughly the same level of noise but with a different sound profile?

I guess it doesn't work that way: each of those sources has its own radiation pattern.

By the way, stop all the fans but the Glidestream, crank it up to about 1200-1300rpm, and you'll have an estimate of how much loud about 26dB will sound.

That's a good idea, I'll have to try that... I might also try seeing how much I can turn up all the fans before they exceed the ~28 dBA noise floor of the noise meter app on my phone. I know it's not accurate but it will at least give me a rough idea of noise levels.

if you consider the SPCR GTX 960 review, Lawrence Lee managed to run the whole system at 14dB under load (slightly above the GTX 670 idle level) with a comparable system (actually he used a semi-fanless Kingwin PSU, while we don't know which PSU the OP got), I mean same CPU cooler, somehow similar fans and so on.
On the other hand the calculator says just under 19dB on average.
Which guesswork may be more representative, 14dB or 19dB? Subjectively an almost 5dB difference is well noticeable.
I'd add that we also don't know how Lenovo measured those 26dB, which (per SPCR standards) is a fairly loud sound level.

Another good point regarding the results from the 960 review. It pretty much answers my question regarding similar noise levels from different components... at least, I think it does.

In regards to the measurements from the calculator, even using the highest values for my parts, it still under 20 dB... now I just need to reread some of the stuff on A-weighting.

As for the accuracy of Lenovo's data. I'm always sceptical of noise measurements when there's a very small difference between idle and load figures (I seem to remember reading the the Lenovo TS140 is 27 dBA under load), with the exception of SPCR of course. I usually just assume that the lower figure represents the noise floor of the testing environment or the measuring equipment. Case in point, pretty much every case review over at anandtech or similar websites. They do knowledge that there testing procedure for noise measurement isn't accurate but when you quite clearly can measure anything under 30 dB @ 1', I do wonder what the value of such data is.

Getting back to the TS140. If I do end up getting one, I'll be sure to document my opinions on it and any attempts to make it quieter. With the growth of NAS systems (which help to make our main rigs quieter by moving mechanical storage to a separate device), I'm sure there'd be interest from the like minded individuals around these parts.

[CA_Steve]

Something else to consider is the tonal quality. Certain frequencies or patterns (like a hum that comes and goes) can be more annoying than a broadband louder noise.