What? Are you now defending above R9 290X reference sound levels on an overclocked reference 980Ti? I am pretty sure someone who pays $1300 USD for dual 980Ti cards finds this unacceptable. I know I would. For a $650 card, this cooler is worthless for overclocking, sorry. It's just as bad as the reference 290X that everyone ripped apart.
I want you to go to this video on this page, turn up your speakers and tell me that's
acceptable for a $650 videocard:
http://www.computerbase.de/2015-06/geforce-gtx-980-ti-test-nvidia-titan/8/
That's only at 52% fan speed, not 70% or even 100%. That's a disaster for a flagship videocard for anyone who doesn't game with closed headphones or noise cancelling IEMs.
Dont put words in my mouth.
I am not defending anything. I just know a little more about sound and noise, apparently. You may not know that noise dissipates just like heat. The pitch and frequencies can carry sound further but really all that is pointless to get into because you know that sitting your ear next to the motherboard is nothing like having your GPU in a case and measuring the sound 2-3 ft away from you. See airflow its self makes a lot of noise but that is completely independent of how far that noise will travel. You measure right beside the card, if it is moving a lot of air, it will be loud. There is no way to prevent that. All you can do is try to control pitches and spread, so it dissipates. When you are that close, most of the sound you measure is directly related to airflow, fast moving air. CFM = noise!!!
You cannot get around that fact but a lot can be done in how you manage it.
I am not making any accuses nor am i saying that the 980ti isnt loud. I do not own one, cant speak about it.
But it absolutely makes no sense to record the dBa directly beside the card. This is not representative of anything but the air flow. This sound can only be altered by changing the size or speed of the fan. This is the noise that must be dissipated and the engineering challenge. The way you deal with it, what you do with it, how that air flow is managed, that is what determines how loud it will be to the end user.
Simply put, not all coolers are the same but they all have to move air. CFM. You should know that larger fans can move the same amount of air while moving slower. The same concept when you use multiple fans. The slower rotation from using a larger fan is quieter because.......why? Think about it.
Smaller fans move air faster to get the desired CFM. The sound is more concentrated at the source but it spreads out and dissipates depending on the design. Right beside the card, you only hear the noise without any dissipation.
Unless that is how you plan on running your PC, with your ear at the edge of the motherboard, then that is a totally unreliable method which is absolutely meaningless to the average user. Honestly, i never really looked at the titanX is so loud claims and examples but if you guys are sighting reviews using these kind of next to card methods, I am just blown away
Surely, if you really dont understand what i am saying, then this will totally blow your mind:
Realistically the reference blower isn't too bad. I could use it without issue. I've owned reference 580's, 680's, 780's, 5870's, 4870's, 7970's, and by all accounts the Ti is quieter than those cards. It's barely louder than 295x2..
OC'd it produces less noise than a stock 290 in "quiet" mode.. "q-u-i-e-t" mode. Or stock 7970, stock 680, or stock 580. They were all not bad. In fact, it's almost the quietest flagship card in a number of generations.. Perhaps the last decade!
Storm in a teacup, move along.
Wha wa
UH?
but that review recording dBa right beside the card says its loud.....
yeah, but apparently there is something about sound you dont know. Or you are just trying your best to paint nvidia cards as bad as you can. Like how you listed two reviews to prove how the 980 reference is so bad. You may have missed the 980 was in the TPU charts along with the MSI Gaming card. Why did you purposefully post a separate review that used totally different methods (with drastically different 980 dBa)?
The only answer is to skew the results . Great job!!!!!