Speaker Spikes vs Isolation: An Investigation

[Robert R]

While my wife is busy sewing up masks using layers of cotton, scavenged HEPA material from vacuum cleaner bags, and other filter media, I decided to investigate the spikes vs isolation question. I have no expertise here, just curiosity.

My speakers are BG Radia 520s (planar magnetic), running off of a Marantz SR7012. I'm using REW for measurements. I set up a measurement rig with the miniDSP mic, positioned at the center listening position. The speakers are very precisely positioned, with exact same distance to the mike, and toed the same amount using a laser alignment tool.

I evaluated the following materials
1) Spikes with the floor disc:

3) Spikes-->disc-->Vibrasystems EVA-BFR antivibration pad (labeled "Wagner" in the graphs). (No floor disc is shown here, but it was there for measurements) See https://vibrasystems.com/foam-rubber-pad.html. This is a 1" thick pad, sort of a constrained layer approach with a softer foam center and harder rubber outer layers.

2) Spikes-->disc-->50 Shore, 1/4" thick closed cell foam ("Blue pad"). See the blue pad in the next photo.
3) Spikes-->disc-->70 Shore, 1/4" thick rubber disc. See the next photo. I did measure this configuration but did not include it the graphs. So just imagine each material by itself.

5) Spikes-->disc-->Isonode 1" Sorbothane half dome.

I set the system up for "large" speakers, no subs, and Audyssey turned off. I measured only the left channel for simplicity since I only had enough materials for 4 footers. For each configuration, I took three measurements (sweep 20-15k) and averaged them (wash out the variance in background noise, etc.). Also, after changing the materials I checked and adjusted the speaker position. Here are the results, first 20-300Hz:

and 300 to 15000

Observations:
1) There is more variance in performance at higher frequencies than at lower frequencies, contrary to my expectations. Some materials damped response (relative to the spikes), whereas others boosted it.
2) At low frequencies the variance in performance is concentrated around some peaks (80-100, 210-250), while other regions (120-200) are essentially identical. Overall, below 300, the various materials did tend to damp response, except between 230 and 240 where all of them boosted response.
3) The Sorbothane domes (expensive) did not show significantly better damping at low frequencies than the cheaper materials. Indeed at higher frequencies, the domes appeared to damp the response more than the other materials.
4) The thick Vibrasystems anti-vibration pad performed about the same as the 1/4" "Blue pad". Notice how the blue and purple lines more or less track together.
5) At low frequencies, the hard rubber disc seemed to provide the most "even" damping, but at higher frequencies, tended to accentuate the peaks between 2k-4k, returning more energy to the room.

I'm curious to hear from those of you with real expertise to further interpret and explain the results. The original REW file available here https://drive.google.com/open?id=1_cZaJaRQtTdRFw1cHVG_F4OFmxM0tV3a.

Bob

 

[waynel]

I think the differences might come down to slight variation in speaker position both relative to the room and relative to the mic. The speaker height looks to be slightly different based on the spike.

 

[beefkabob]

Also the sorbothane looks awfully squished. How much weight and how much is it designed for?

 

[DownUnderGazza]

Check out the REW distortion and Step Response graphs, not that I expect any differences.

My understanding with hard vs soft footings is in vibration transmission. Both into the floor and back through the speaker cabinet as baffle, side wall and frame vibration. Measuring vibration is not so easy for us amateurs.

 

[Pio2001]

Hi,
The differences in high frequencies are unlikely to be caused by the spikes. Vibrations going through the spikes should more or less sound like what's the neighbour living in the flat located below would hear : mostly low frequencies.
I suspect the differences in high frequencies to come from the exact location of objects in the room. Variations from 1 to 5 kHz might come from your own location. According to where you are standing in the room, the sound waves bouncing around your body might cause what we see here (granted that nothing else has moved in the room, open/close door etc.)
The difference at 11 kHz is strange. This may occur if the microphone moves a few centimeters away from its initial position.

The best candidate to investigate are the variations at 50 and 210 Hz. The odd man out is the Wagner pad. This may be caused by its larger surface, that may transmit more vibrations into the ground. In this case, since the level is inferior, these vibrations would be out of phase with the direct sound.

Someone made a calculation, some years ago in a french forum, about the theoretical frequency response in transmission of a steel spike resting on a steel plate. The question was to find out if spikes were rather transmitting vibrations, allowing the huge mass of the floor to keep the speaker perfectly still, or if on the contrary the spikes were isolating the speaker from the ground, preventing vibrations to get into the floor.
He found that the calculated resonance frequency of the metallic contact was within the audio band, and that this kind of spikes should have a chaotic behaviour throughout the audio band, blocking some frequencies and transmitting others.

But here, the pads below the spikes are supposed to prevent vibrations to get into the floor, which is a good thing.

 

[Robert R]

First, the variations are not likely to variations in height of the speaker relative to the mic. These are BG Radia 520s.

They are 70" tall, with a 50" PM driver. See https://www.stereophile.com/floorloudspeakers/1204bg/index.html for review and measurements. Other than the Wagner pad, the difference in height between the materials was less than a 1/8", as they all were about 1/4" in height (even the squished sorbothane). Given the driver, I'm sure that would not impact the measurements.

Second, as to distance from mic, again, I carefully remeasured and reset after each material change. The mic never moved. Here is my rig:

You can see the 1" ft markings which I use when making multiple seat measurements, so I get the same location/height every time. This is a closet rod, two quick clamps at the ends, and a C-clamp with the mic stand zipped to it.

As to test conditions. I always sat in the same spot, on the right side of the room away from the left speaker, and no changes were made in other stuff. That said, I could set up the delay function for measurements and leave the room before each measurement to ensure precisely no variance.

Yes, the sorbothane was pretty squished, so I'm likely exceeding its load rating. They were the hardest to get installed because they tend to roll over. The version you see was the most stable. I have some sorbothane sheet material somewhere--I'll dig that out and test again. Have plenty of time on my hands....

I think the big picture is that while we're trying to decouple bass frequencies, we're also changing mid range and upper range behavior as well in a rather complex way.

 

[Wes]

Interesting.

I don't suppose that you happen to have a laser interferometer handy?

 

[Robert R]

No, but I'm expecting a laser particle counter from Amazon in the next couple of days to test the masks my wife is making. After all, we're objectivists through and through.

 

[NTK]

But you will also need an electrostatic classifier to generate the correctly sized particles to test your wife's masks.
https://www.tsi.com/products/partic...fiers-and-dmas/electrostatic-classifier-3082/

 

[Robert R]

https://www.amazon.com/gp/product/B071RF6B2Y/ref=ppx_yo_dt_b_asin_title_o00_s00?ie=UTF8&psc=1

I think this will do the job...

 

[NTK]

I was just joking (and I thought you were joking about the laser particle counter). Those electrostatic classifiers are for highly specialized research applications.

What you have ordered should be sufficient, in my estimation.

 

[Wes]

So what accounts for the mid range and upper range differences found?

 

[DownUnderGazza]

So what accounts for the mid range and upper range differences found?

Now this is the question!

 

[tw99]

Might have been an idea to try to repro one or two of the setups to see if they are self consistent?

 

[tuga]

Jim Lesurf has a piece about this in his website:

Spikes and Cones – What’s the point?

In recent years it has become common for items of audio equipment to be mounted using “spikes” or “cones”. These come in all sorts of shapes, sizes, and materials, at all kinds of prices. The Hi-Fi magazines sometimes ‘review’ these accessories, and recommend their use. However, are they worth buying and using? On this page I’ll consider their use with loudspeakers, and discuss some alternatives.

http://www.audiomisc.co.uk/cones/speak.html

 

[A800]

No, but I'm expecting a laser particle counter from Amazon in the next couple of days to test the masks my wife is making. After all, we're objectivists through and through.

I can't wait for the measurements.

 

[tecnogadget]

I did the same research as you did with Spikes vs my very own Constrained Layer Damping fórmula feets. With the same results as you in the Frequency Response department...no discernible change.

Having son little effect in FR, I think we could only find effects (if they ever exists) in Impulse Response. I’ll try to review my own measurements and see what I can spot there.

 

[tecnogadget]

A few years ago I measured my floorstander with Umik 1 pointing 0º at tweeter level height and 1-meter distance, the listening room had porcelain flooring.
The comparison involved bare speaker over a 5mm square sheet of MDF (to avoid any speaker rocking since being porcelain tiles flooring it was not 100% flat), DIY Constrained Layer Damping feet, and Spikes.

DIY CDL feet: Sandwich of 3mm thick sheets of rubber straps and 3mm thick sheets of mdf. Rubber-Mdf-Rubber-Mdf-Rubber-Mdf-Rubber
(sorry for the bad photo)


Frequency Response: Minor changes, only noticeable below 400hz, small 1dB changes that possibly won't be heard.
Do not look for pretty FR since this measurement was only meant to find "different kind of feet" differences. Don't even remember where the speaker was placed for the measurements, and of course just RAW response, no PEQ or processing at all.

Impulse Response: I'm not able to find or interpret any meaningful difference. Any contribution will be appreciated.
FLOOR

DIY CDL

SPIKES

Decay: Looks like Spikes improve decay especially below 40hz at 160ms, but still not sure about its audibility.
FLOOR

CDL

SPIKES

Here is the measurement set if anybody wants to play with it or finds andy meaningful plots.
Floor vs CDL vs Spikes REW Measurement Set

 

[DSJR]

I think you people with large bare rooms and plain un-carpeted floors forget one small thing?

Unless I'm mistaken, the use of spikes was a UK centered trend to start with? The reason it worked so well here in the UK I think was the fact most of us had carpeted floors with underlay, which made many smaller stand mounted speakers wobble alarmingly. Speakers from Linn, which majored on the precise perceived 'start-stop-snap' of percussive notes above most else (forget the hideous tonal colouration of most of these early models) really did respond to floor spiking and in the case of my then Isobariks, just shoving a cassette box between the rear of the cabs and the back wall to stop movement just didn't work as well over all frequencies. Using cross head screws into the floor-boards and then spiking into the heads of these, seemed to aid the bass reproduction and my experience over the years is that getting the bass better, usually has benefits further up the frequency range.

These days, my current boxes (Harbeth SHL5s) don't seem to give a hoot how they're mounted (the bass they have is the bass they have) and in their case, they sit on plain oak custom size 'tables' via small bump-on feet which allow some stand movement perched on top of our carpet and in this case it's distance from the back wall which determines how much they boom (I'm not allowed to change them right now).

None of this matters on non-percussive music in general, but for better reproduction of drum kits and so on, it used to make a difference on many speakers, especially passive crossover domestic models which can slur the dynamics a little if under-driven as many domestic stereo's used to be. Just my vibe and obviously UK-centric.

 

[tecnogadget]

DSJR
I think you are right with your point as to why the advent of the Spikes trend .
That must have taken place some decades ago, and so much have happened since then with so many Snake-Oil companies/reviewers promising Jesus juice with their magical speaker feet solution, that these kinds of investigations are interesting.