Heat, Noise and Heat Vs. Noise
We test cases in stock configurations for a few reasons. Chief among these is that we don’t want to take responsibility for ventilation or noise problems, where temperature and noise have an inverse relationship that depends primarily on fan speed. We would need to experiment with dozens of fan model and placement options to find the ideal cooling-to-noise ratio, and chassis manufacturers are in the best position to do that. They do, after all, have the strongest motivation to improve case performance.
And then there’s the problem of price. Aside from specials and temporary discounts, cases like the fan-packed Antec Three Hundred Illusion cost more than single-fan enclosures like the Antec Three Hundred. Value might not be a big part of our search for the most premium case, but we aren't about to throw away the price/performance chart entirely.
CaseLabs presented a unique testing challenge in that its Merlin doesn’t include fans, though that might not have been a problem if fans were among the options on its configuration sheet. They aren't, though. Still, anyone prepared to drop $500 on a high-end enclosure has the money for their own fans. Testing the Merlin without any just to reflect a shipping configuration wouldn't be fair or realistic.
I don’t have more than two of any particular fan model in my lab, so I picked a pair with a fairly good cooling to noise ratio. I placed one behind the CPU cooler to pull heat away from it, and another at the bottom of the case to feed the graphics card. We can see the positive effect on thermal performance, and would confidently predict that a third fan would have likely yielded even lower CPU temperatures if it was added to the second front-panel mount.
Fans also add noise, but in a peculiar way. Two fans with opposite noise patterns could cancel each other out acoustically, though that never really happens in practice (they don’t stay synchronized). And then there’s the nearly unpredictable problem of beat tones, where overall SPL measurements aren’t detailed enough to do those calculations.
We do see that the Merlin SM08 has the highest operational noise levels, even without any added fans. That makes sense from a design standpoint because nothing stands between the graphics card and the vented front panel that might otherwise have reflected fan noise back into the case.
CaseLabs' submission approached ideal thermal conditions with a couple of quiet fans added, however, making it an average overall performer in the comparison of cooling-to-noise. Corsair’s Graphite 760T leads this chart.