Placement options – Impacts of location, orientation and treatments

JonFo · 12-09-2007, 02:05 PM

We all know that placement has huge effect on the sound of our speakers. Along with the general room layout and treatment, where we put the speaker is vital to the perceived sound and its quality.

During my SL3XC build, I documented some of the effects of placement and treatment of the wall behind the speaker. So today, I decided to expand on that using the left-over SL3 from my center build.

So here goes a series of posts on placement, with supporting measurements.

First, the context. A pic of my SL3, the measurement mic, the VOM and the Variac (variable AC voltage device) used to update results on the panel Voltage vs frequency / SPL thread.

The laptop is used to do Windows Remote Desktop to the Automation PC in the equipment room; it has the microphone amp, and other Audio I/O for measurements (an M-Audio FireWire 410). I use the latest R+D (essentially ETF v6, from acoustisoft.com)

JonFo · 12-09-2007, 02:07 PM

All the following measurements were done energizing the Panel only, since for the Variac tests I did not want to have to deal with the output of the woofer as well. Also, for these placement related metrics and discussion, I wanted to separate the bass, as it is VERY placement dependent, and much has been written and discussed about how to place a speaker with a box woofer. I figure it’s more relevant to discuss placement impacts for dipole ESL’s, since that’s what we have

First up are the Measurements of the SL3 at the usual recommended 3 feet from the wall behind it. It is exactly aligned to the wall behind, with no toe out.

The wall in this shot is untreated, bare drywall.

JonFo · 12-09-2007, 02:08 PM

Here is the impulse response of the 2.5’, untreated real wall corresponding to the Freq. response plot above:

Note the massive spike at 4.2ms and its reversed echo at 16.5ms.

JonFo · 12-09-2007, 02:10 PM

Note the massive spike at 4.2ms and its reversed echo at 16.5ms.

These are really bad, as they cause comb filtering as seen in this Bode response of the same measurement:

JonFo · 12-09-2007, 02:12 PM

Now we move on to a treated surface behind the speaker.

This is the same distance, but with a treated surface behind it:

Note the decrease in artifacts at 4ms.

JonFo · 12-09-2007, 02:13 PM

And the accompanying Bode response. Notice how much less comb-filtering there is in this plot.

JonFo · 12-09-2007, 02:14 PM

Here is the contrast of the treated vs untreated surfaces. Note the much cleaner impulse of the treated wall.

JonFo · 12-09-2007, 02:15 PM

And to conclude the treated (blue) vs untreated (red) sequence, here is the frequency response of the treated wall at 2.5ft.

Note that the untreated wall has more energy, unfortunately, it’s smeared in time. The blue trace of the treated wall shows a smoother response, especially in the critical midrange.

JonFo · 12-09-2007, 02:17 PM

Now that we see how much havoc the bare wall reflecting energy back at the ESL panel can cause, let’s examine a more typical setup, where the speaker is toed in a bit.

Here is the setup

JonFo · 12-09-2007, 02:18 PM

So here is an impulse measurement of the same location (bare wall behind), but now toed in. Notice that the nasty reflection is gone? That’s because, just like a billiard ball bouncing off a side, the sound is ricocheting at an angle away from the speaker. So no nasty spikes around 4ms.

JonFo · 12-09-2007, 02:19 PM

While I saw a good bit of benefit from diffusion behind the panel in the center channel (link to thread here), as it is perpendicular to the wall, in this situation where one has a toed out speaker, and we place diffusion behind it, we see some of the energy that had been reflecting away from the panel, now being redirected back through it by the diffuser, some of which we probably don’t want. My theory is that one wants an abfusor behind the panels. I have some RealTraps diffusers (which also do absorption) on their way. I will update this thread when I get them.

And here it is toed in and with some diffusion behind it: Note how it now has a reappearance of the artifact at 4.9ms (a little longer, as the bounce had a bit longer to go due to the angle to the wall).

JonFo · 12-09-2007, 02:20 PM

And here is the Bode response comparing the two mode, with diffusion and without. Diffusion helps in some areas, like the smoother response between 1 and 2Khz, but up higher it clearly smears things a bit more. Interestingly, this hardly translates to any significant difference in the frequency response plots.

JonFo · 12-09-2007, 02:21 PM

So, what do we learn from all of this?

First, do not put the speakers directly aligned at 90 degree angles to surfaces. Again, think of a billiard ball, you do not want it coming right back at the point of launch. You want that rear wave to travel more than 15 or 20ms before arriving at the listening position. Ideally, it should be diffused and slightly attenuated as well.

Putting the usual toe-in on speakers is a good thing for these dipoles.

Be careful with diffusion, it’s good in many circumstances, but you really want to see if it’s also better with some absorption in the high-frequencies. And some might prefer no diffusion.

Finally, some level of absorption behind the speakers is good, especially if it’s absorptive in the mid-bass @160Hz through 500hz. As this helps mitigate the dipole cancellations in that frequency. And absorption of high-frequencies mitigates bounce-back and time smear in the <20ms arrival window that the human hearing is most sensitive to (anything between 0 and 20ms is deemed to be the ‘same’ sound).

That’s it for now. Next time, I’ll dive into what does distance from walls do. .

Rich · 12-09-2007, 02:53 PM

Your research verifies what I discovered in playing with acoustic treatments behind my Summits. Diffusion sounded bad. The high frequencies were harsh and smeared. Micro traps were slightly better, but still didn't really help imaging much. But the High Frequency Mini Traps from Real Traps did an excellent job at clearing up the imaging and the sound at all frequencies. These traps absorb fairly evenly from about 250 hz. all the way up (basically the entire frequency output of the Summit panels). In my experience this provided the best sound from the Summits. I look forward to hearing about your experience with the diffusers. I am tempted to try the diffusers in front of my HF Mini Traps to see what the combined effect would be.

edwinr · 12-09-2007, 03:17 PM

Thanks for posting that info. Very interesting. I am lucky to have a brick wall behind my Summits. I think this wall offers a nice balance between diffsion and reflection of the rear sound wave.

Speaker setup is tricky. I recently moved my Summits a few inches to accomodate a new pull down HT screen. I lost the sparkle I was getting before. I moved my speakers another few inches in another direction and changed the toe in slightly. I had the sparkle back...