Tom's H. noise charts - reliable?

[DigDog]

what the header says;

i'v seen at times Tom's Hardware noise and heat charts differ wildly from other websites, and i have been wondering for some time how reliable they are. Tom's had a period of bad vibes when it was putting out some dodgy reviews, now i'm curious if the noise charts went the same way. What's weird is that nowhere near the charts is mentioned that the dB scale is logarithmic ...

 

[exar333]

I think it really depends on the case used and their methodology. Short answer is really the reliability is article by article, and you should look closely at what they used and how they measured their results to judge if the conclusions are indeed valid.

 

[Jaydip]

what the header says;

i'v seen at times Tom's Hardware noise and heat charts differ wildly from other websites, and i have been wondering for some time how reliable they are. Tom's had a period of bad vibes when it was putting out some dodgy reviews, now i'm curious if the noise charts went the same way. What's weird is that nowhere near the charts is mentioned that the dB scale is logarithmic ...

Why they need to mention that?Its pretty well known

 

[Ben90]

I converted to the linear values for 680 vs 680 SLI for you so you can better judge the noise difference.

 

[magomago]

Why they need to mention that?Its pretty well known

actually i would bet money that most people reading benchmarks don't really think about what the scales truly mean.

and in the case of audio being measured in deciBels, I'd bet you that even more have *no* idea how to interpret it.

I've never seen the charts, but this is what I'd think they would have to control and maintain for their results to really be valid:

1) measure using the same case
2) measure using the same system configuration
3) measure using the same mic
4) measure from the same distance
5) measure with ambient conditions as consistently quiet as possible

 

[Ben90]

To be completely honest, SPL doesn't tell the entire story. Technically speaking, a person could hear a 150dB sound their entire life and it wouldn't matter if it was at like 0.01Hz. However a 150dB sound at 50000000000000Hz would pretty much ruin your ears instantly.

Deep sea divers already subject themselves to noises over the theoretical limit for undistorted sound in air.

 

[DigDog]

sorry i didn't want to go into technical as i can get nerdy on audio, it's just that if you rate a 37dBa card "quiet" then a 43dBa card can't be "reasonably quiet" ... it's twice the f* noise! a spectrum analysis would be even better but hey, they could just write the spl @ 300hz and that would also be ok.

thing is, i'v got a (see sig) and it's really, really annoying, i can hear it wooshing away even when i'm playing shooty-bomby things and my next card *has* to be proper quiet.

 

[Imouto]

Tom's is a joke. Not even a third rate review site anymore.

 

[BrightCandle]

I would say Toms noise charts are indicative. All the sound testing is only really valid on one site however as they use different setups that produce wildly different results. Its also hard to say what the floor of their test setup is unless they publish it (and as far as I know Toms never has).

I really wish they would publish some frequency data as well or have more sites record the noise.

 

[DigDog]

is there anywhere else one can get reliable noise testing .. pls?

 

[ViRGE]

is there anywhere else one can get reliable noise testing .. pls?

I've always found AT's noise testing to be plenty reliable.

 

[Leyawiin]

Techpowerup's reviews tend to have pretty good subjective and objective noise measures of cards. W1zzard is really fussy about noisy cards.