You do realize I started that thread, right?
Wow this forum chat escalated quickly... I just wanted to chime in, I know this forum is about measurements and such but I've been putting aside some money for the last 8 months and went on my search for the holy grail of mixing monitors. Yes, there has been quite some hype about these speakers and I was actually on the other side of the hype, mocking the fan boys.
Yet... I went to test the ONE15s and ONE18s a few weeks back, and of all the speakers I have tested since March, none of them told me as much about the areas that would need fixing in a mix as the ONE15s. That's also the reason why I ordered the Amphion 100 amp a few days back and will be getting the ONE15s in January. I want to mix on them and I'm sure the Genelacs or the Focals might have been flatter (no fan of the Neumanns btw. not even the KH310s) but they sound veiled in comparison to the Amphions and they hide the little faults in the mix instead of revealing them. They may be hyped but honestly if you want speakers to mix on, they should indeed be very high an the list of potential buys, they feel fast and full of micro information, I still don't understand how they do it tbh.
"I still don't understand how they do it tbh"
See, that's the problem. There's no magic in speakers, certainly not in monitors. They can just sound different based on a variety of factors that are present in the measurements. It's not like Amphion is even using any exotic materials or drivers that can provide some secret sauce. They're just engineered to certain goals.
Speakers being revealing in a mix can also be a bit misleading. You can also EQ your speakers to have a 3dB boost at 5kHz and they will definitely sound "revealing" too. That's how a lot of HiFi speakers get away with supposedly sounding detailed when they are just exaggerating parts of the response.
In that context,
of course you will hear things you might not have noticed before, because different parts of the music will be emphasized than before. That's why Barefoot Sound has its "MEME" DSP knob to emulate different types of speakers, such as an "Old School" setting that replicates the sound of the oh-so-popular yamaha NS10Ms, which are awful speakers that managed to catch on because they were also "revealing."
But please don't this as me saying you shouldn't enjoy what you like. If you tested a bunch of speakers and picked the amphions, found that they worked best for what you wanted, that's great! It's just when comes to making recommendations in a given price range, I think there are legitimate concerns about value and excellence in engineering relative to the competition. (Then again, there are studios that mix with worse speakers too)
I see. And yeh, I agree, the PIRs are not particularly reliable as predictors of how a loudspeaker will perform on a desk, in a room in which first reflection points have been heavily damped.
The PIR is a weighted average of:
- on-axis response
- early reflections
- sound power
Although (1) and (3) are not affected by the studio environment, (2) certainly is.
Small correction, PIR uses the listening window, not the on-axis. Specifically 12% LW, 44% ER, and 44% SP. In practice it is almost always nearly identical to the early reflections curve, except tilted by an extra 0.5-1 dB.
Let me rephrase: If i have understood correctly (?) predicted room responce is for ”normal” hifi speakers, for normal livingroom conditions, and with its acoustic anomalities. When you rise that speaker, and listening position, room modes and early reflection hits to different positions in audio spectrum all together, since the distance changes to the nearest boundary. And that console or desk boosts mid freqs, and this and that changes.
So my question is: How reliable those predictions about room responces are, when premises of listening conditions changes so drasticly in studio monitoring setup?
Yeah, it's been discussed a few times.
The PIR is not designed for studio setups. However, this does not mean it's not useful: it still is a summary of the speaker's directivity properties, and these are still very much audible in a treated studio context, just to a lesser degree. It's not that much different than listening in a bigger space or listening at shorter distances. A bit more direct sound, a bit less room.
My anecdotal experience is that PIR is quite accurate in different setups up to about 10kHz so long as the listening distance is 6ft/2m or more, in both large and small spaces. Below that, the tilt becomes lesser and lesser, and the actual in-room curve starts to trend closer to the direct sound. Similarly, broadband room treatment will reduce the proportion of room sound relative to the direct sound.
The PIR very closely resembles the early reflections curve, which is an average of five averages:
Front: 0, ±(10,20,30)
Side: ±(40,50,60,70,80)
Rear: ±(90-170), 180
Ceiling: vertical +(40,50,60)
Floor: vertical -(20,30,40)
As noted by
@andreasmaaan, ER is the only variable that will change in a studio.
Because of the substantial averaging that is happening, I find the PIR to be quite flexible to a variety of listening environments. It covers the entire horizontal response of a speaker, so differences in horizontal positioning don't really matter, even if theoretically the weightings might change slightly.
As you noted, speaker height is probably the biggest differentiating factor in a studio. But even then the differences are usually relatively minor. The desk bounce would, after all, make up for some of the difference in speaker height.
Any differences would be most apparent if the speakers are substantially closer to the ceiling than your floor. We can imagine this by simply inverting the ceiling and floor angles; the effect on the ER will usually be small. Here's the ER of the D&D 8C if I swap the ceiling and floor bounce angles:
Considering the ER is only 44 percent of the PIR, the effect will be even less.