Ivan,
I measured the cabinet on the ground-then raised it up so the bottom was 15.5' off the ground and measured again.
So this PROVES (at least to me) that the whole idea of "getting more bass" when the subs are on the ground OR simply that you can "add a magical 6dB" to a whole space measurement is simply false.
Ivan,I just did this in the back parking lot-so not completely free of reflections.
What was the distance of the subs and mic to any buildings?
Art
Behind the sub is the building. It is about 60-70' behind and is basically about 24' high. It has "lossy" walls.Thanks, the distance to boundary clarifications are important to determine whether there was actually a half space to free space change.
The building was about at a 45* angle to the rear of the sub
Behind the mic the nearest "large object" is probably 400' or more. There are some trees and a creek between the mic and the next building
The mic was laying on grass.
Thanks, the distance to boundary clarifications are important to determine whether there was actually a half space to free space change.When I get a chance I can redo the test. I have a 17' tall mic stand that I could use.
For what it's worth, I have done a similar test on a smaller scale, same results. That said, if the test was done at 15.5 feet with the mic on the ground, then the cabinet and mic lifted 15.5 feet high, the second measurement would show a loss in bass response compared to the first.
Interesting... but is 15ft high enough to consider it a move from half to full space? Particularly given the long wavelengths involved?As I said-my forklift only went that high. To go higher would require renting in a larger lift.
I'm thinking that 15ft over 100ft makes for more of a long flat acoustic mirror, than a change of space.
And that floor bounce summation at all freqs will range from minimally destructive at the longest wavelengths to greater destruction at shortest...like what seems to be going on with the plots..
As you can see, the biggest thing that changed was the HIGHER FREQ-NOT the low freq. This is because the mic is not on axis with the sub and the higher freq are starting to become a little directional.
And yet, anyone can hear the difference in bass response between a monitor on a stand versus one on the floor. Why do you think that is?Part of the reason is the position of the ears-the floor (or other reflective surface) and the speaker.
What do you think your results would be if the mic was at "ear" level instead of on the ground?I am going to set up the test again soon. Not this week.
Part of the reason is the position of the ears-the floor (or other reflective surface) and the speaker.
When measuring subs, the mic is on the ground (to avoid reflections).
Reflections can/will cause all sorts of deviations in the response that are totally position dependent.
So when measuring any loudspeaker, you need to reduce/eliminate the reflections as much as possible in order to get any sort of useful measurement of the device.
The idea of measurement is to measure the device itself-without any outside influences.
Thx Ivan, All that makes good easy sense...What brought this on was a discussion in which the "other argument" was "When you put the subs on the floor (vs flown) you get more output".
But still, I'm thinking with a mic on the floor, when you raise the sub off the pavement, you will still measure reflections...
Not the kind of straight line floor bounce reflections we normally think about ala comb filtering, but a kind of integrated weaving of the primary line of sight waves with their reflections off the pavement. An outdoor reverberate field...or better yet, interference field ......solely off the pavement.......that decays oh so quickly...
Guessing that your measurements may be showing this with the relative widening magnitude difference as freq rises.
Also think it only makes sense that a 6db gain from going to half space is a theoretical maximum from perfect coupling, and that less perfect couplings at some distances will also gain some portion of the +6db (albeit more freq dependent). IOW, just like combining subs..
So, I'm really left wondering if anything can be said for the test, in terms of whole vs half space....
What brought this on was a discussion in which the "other argument" was "When you put the subs on the floor (vs flown) you get more output".
Which this test proves is basically wrong. There is no more output-at least that is easily noticeable.
Now if you are close to the sub- you will get more output-but that is due to inverse square-NOT whole vs half space.
I really can't wait to do more tests and get some more data.
I just did this real quick.
I am going to set up the test again soon. Not this week.
One of the measurements will be going ground, 5.5' (ear level) and 16' of the mic on both the ground and flown sub positions. Of course there will be reflective cancellations, but what about broadband?
ALSO, there is another test Tom D. wants me to do that might REALLY throw some interesting discussion/fuel on the fire.
I was one of those that believed ground stacked subs had more output due to coupling with the ground, so thank you for busting that myth.They were sitting on the forks strapped down-so not moving.
When you did the test, was the sub sitting on the forks of the forklift, or did you hang it off the forks? I ask because most likely the cabinet did not move much sitting on the forks, no matter how loud you cranked (friction.)
In actual use, the subs would be flown (at bottom of a hang or behind the main hang) so there's possibility for the cabinets to move or swing in the air -- they're not stationary. Given Newton's 3rd law of motion, I would guess there is enough force to move the cabinets, especially at high volume. (Even on the ground, my TH118's move and dance around.) I would expect the movement affects the output ever so slightly.
Would it be possible for you to do the next test with the subs "flown?"
Many thanks,
John R.
I am going to set up the test again soon. Not this week.
One of the measurements will be going ground, 5.5' (ear level) and 16' of the mic on both the ground and flown sub positions. Of course there will be reflective cancellations, but what about broadband?
ALSO, there is another test Tom D. wants me to do that might REALLY throw some interesting discussion/fuel on the fire.
Acoustic loading when reflected back through the driver is (I always thought) counterintuitive.
When the resistive acoustic load on the driver is increased say with a horn, the resistive portion of the impedance goes up, not down and so, the power delivered to the speaker with a fixed Voltage, actually goes down proportionally too.
Surely rather than use a fork lift to raise the loudspeaker and leave the microphone on the floor, it would be easier to leave the loudspeaker on the floor and raise the microphone?
Am I missing something where the principle of reciprocity does not apply in this instance? If I am could you provide links/references to any articles that support your claims.
Thanks.
Am I missing something where the principle of reciprocity does not apply in this instance? If I am could you provide links/references to any articles that support your claims.I think I understood Tom's post and, if so, then the answer is that yes there are differences that would make reciprocity not entirely faithful here.
Thanks.
In this case, Tom's contention that the proximity to the ground impacts the impedance and output at a fixed voltage would seem to indicate that moving the sub could produce differing results in both positions that wouldn't necessarily be matched by moving the microphone instead.
Hi Tom, trying to understand what you are saying...
So as a start, I thought to maybe just test the effect on impedance in going from full space to quarter space..
While DATS is a program I use often and useful for finding the frequency of impedance peaks and impedance minima I don't have any faith in its ability to determine absolute impedance. I have ran 3 or 4 consecutive tests in a row and impedance maxima varies conciderably between concecutive tests in a row. Results are only as good as your tools. Give a man a watch and he will be absolutely cetain of the time. Give a man 2 watches and he will never know what time it is.
Results are only as good as your tools. Give a man a watch and he will be absolutely cetain of the time. Give a man 2 watches and he will never know what time it is.
I use the term "meters" instead of watches.And fill the same box with two different brand 18" drivers and you know it's time to change the spec sheet ;) .
Same result :)
Results are only as good as your tools. Give a man a watch and he will be absolutely cetain of the time. Give a man 2 watches and he will never know what time it is.And fill the same box with two different brand 18" drivers and you know it's time to change the spec sheet ;) .
Looks like 4-5 dB difference at 31.5 Hz to me, what's it look like to you?
Art
Not to spam Ivans thread, but here is an interesting white paper from David Gunness
http://fulcrum-acoustic.com/assets/pdf/whitepapers/comments-on-half-space.pdf (http://fulcrum-acoustic.com/assets/pdf/whitepapers/comments-on-half-space.pdf) exploring constructive and destructive interference from the ground bounce.
We often discuss these different spaces as if the sum from a bounce is flat, but it isn't always and can have audible frequency response effects.
I now return this discussion to the speaker geniuses (not me).
JR
John Rutirasiri,
From your response I am assuming that you do not understand the principle of reciprocity. Basically it states that if you swap the positions of the source and receiver (microphone/ear) the result is the same, it doesn’t matter which way round they are.
Results are only as good as your tools. Give a man a watch and he will be absolutely cetain of the time. Give a man 2 watches and he will never know what time it is.And fill the same box with two different brand 18" drivers and you know it's time to change the spec sheet ;) .When we changed drivers in the TH118 (very early on in its history), the measured freq response change was so minimal (around 0.5dB), it was not considered to be worth changing the spec sheet. Especially when you consider that the driver to driver variance is often more than that.
Looks like 4-5 dB difference at 31.5 Hz to me, what's it look like to you?
Art
When we changed drivers in the TH118 (very early on in its history), the measured freq response change was so minimal (around 0.5dB), it was not considered to be worth changing the spec sheet.Since the chart was labeled TH118, I thought it was a TH118.
The tapped horn I used in this measurement/experiment was part of a "experimental" project for a new product idea. Not a TH118.
Since the chart was labeled TH118, I thought it was a TH118.Yeah-that would be a "logical" assumption. :)
I agree with Gunness's paper about the duality or reciprocity BUT, our ears aren't at floor height. More likely something like 5 feet or more above it.That is why I always try to fly the subs-depending on the venue.
Flown subwoofers and triangulation dictate that some frequencies are gonna cancel, likely resulting in audible changes throughout the audience.
Stacked subwoofers and triangulation virtually negate the chances of this happening, but now we're killing the people in the front.
Pick your poison.
Been following this thread closely as I'm moving my system to another room at the end of the month. I've had my subs centered under the stage and while I like the coupling, they do tend to be loud in the first 2 rows.Just one thing to consider when flying subs.
I'm now thinking of flying them in the center. It'll be flown from 20" truss so I'm hoping the weight won't be an issue.
I'll discuss it with my install company.
Just one thing to consider when flying subs.
Either get them close or very far away from the ceiling.
You need to be aware of boundaries and the cancellations they can cause.
But doesn't this test just prove that there is no increase in output at floor level? I think many of us have noticed a substantial difference in low freq output at normal ear level when placing a speaker on the floor, right? From a practical point of view, the audience rarely listens while lying on the floor. Am I missing something?(probably) ???
Yeah I'm with this guy. I can think of a number of rooms throughout the years where taking the subs off a stage and putting them on the floor made a substantial difference in their perceived output. There was one large club in particular (over 500 seats) that after years of an "everything must be on the stage" policy finally allowed subs on the floor. During a soundcheck we took them off the stage and set them directly in front of where they were. There was no repositioning of any kind other than that.There are a lot of different factors involved in "tests" like this.
The difference was not subtle. We were so amazed that we moved them back and forth several times... just so we could congratulate ourselves on how much better our subs sounded.
Am I missing something here? Maybe floor cancellation if the subs aren't flown 50 feet in the air?
I wonder how much higher than 15' and closer than 100' to mic would change things.The whole "point" of the test/post was that MANY people believe that you can "simply add 6dB" to a whole space spec.
I'm thinking 12 feet from the ceiling. That doesn't sound very far.
Am I missing something here? Maybe floor cancellation if the subs aren't flown 50 feet in the air?Regarding the no-go-zone 2-8ft from a boundary, your typical stage is going to put the sub right within that region which will cause cancellations at all kinds of frequencies. Most people don't consider it, but I've had improvements in response just by rotating a sub 90-180 degrees which was sitting near a wall. The cabinet depth put the driver (when facing the audience) into the +2ft range and caused cancellations.
Regarding the no-go-zone 2-8ft from a boundary, your typical stage is going to put the sub right within that region which will cause cancellations at all kinds of frequencies. Most people don't consider it, but I've had improvements in response just by rotating a sub 90-180 degrees which was sitting near a wall. The cabinet depth put the driver (when facing the audience) into the +2ft range and caused cancellations.
After turning your sub around changing it's distance by 2 feet, do you then adjust the time delay of your tops to bring them into phase alignment? :-)In this particular case I was trying to polish a turd. The subs were a pair of MRX528S's 15 feet from each side of a stage in a fan shaped room and about 5 feet from the nearest seating running parallel to the front wall. The Mains were some decent quality EV tops flown in several day lines above the stage front and 2 rows further back with atrocious spacing for coverage. There was no aligning anything except the flown mains to each other and the best I could do with little slack on the sub lines hanging out of the wall was even out the coverage on either side by rotating the subs. Next time I went back someone had "corrected" my mistake for me.
So this PROVES (at least to me) that the whole idea of "getting more bass" when the subs are on the ground OR simply that you can "add a magical 6dB" to a whole space measurement is simply false.
Since the mic (or person listening) is already in half space, you get the loading. But no additional loading when both are in half space.
I think you need to consider that listening ear are at least 1.6 meters above the ground level. Then comb filter interaction will cancel some frequencies.So have you EVER seen a spec sheet that said the mic was at ear height? And what is "ear height? Sitting or standing?
As you can see there are two deep holes. The first one is at 80 Hz and the second one is about at 240 Hz, (of course the last one is well above frequency xover). This is obvious depends on the height of both subs and listener ear. Another think to consider is that as you move away from the flying source, the comb tends to tilt up. This means that only further away people can be considered fully in half place.
(https://s24.postimg.org/ei3xugs1h/Immagine.jpg) (https://postimage.org/)caricare immagini (https://postimage.org/index.php?lang=italian)
Models are nice-but they DO NOT represent everything that is going on in real life.
There are a lot of different factors involved in "tests" like this.
For example. Room modes. The room modes can often have a greater effect on sub performance than anything.
And the reflections (which are part of room modes) can greatly affect the output at different locations.
I will agree that in some cases, putting them on the floor results in greater output. But it could be because of something other than in the air vs ground.
With my test, there was only 1 variable-since I was outside and pretty far from reflections.
What I was addressing was the "general statement" about half space vs whole space.
And unless you were in a very LARGE room, when the subs were on the stage-there were nowhere near whole space.
So-as usual-it depends.
Ok so I'm trying to get my head around this. Is what you're saying is that there's no such thing as an advantage to half space loading versus full space? That the reduction in low end that we all experience when we raise tops on stands or put subs on a stage is caused by room modes? And that the low end would return if we could just get the cabs high enough - and far away enough - from boundaries?Room modes and reflections are related and different.
In a way, that actually all makes sense. It's just looking at an observed phenomenon from another perspective.
And I know you're going to say "it all depends" ;) In general... :D
I can get my hands on a forklift, some subs and smaart. What should I do to make a valid measurement?Get away from reflections
Like this?Yep.
(http://i.imgur.com/SCbLFqN.jpg)
Distance from sub to mic will be ~25M and the forklifts will be colormatched.
Yep.
Just remember that the colors MUST match-or the whole thing doesn't make sense ;) ;)
Great work, thank you Jelmer!
At 120ms distance,
it looks like building reflections are coming in, and the effect of subs flown vs ground is getting triangulated out to minimal...
At 30ms distance,
first thing I looked for was acoustic reciprocity as was discussed earlier in thread...
Looks like the sub flown/mic on ground, and the sub on ground/mic at 14 ft,..... are about the same trace....so looks like pretty good reciprocity...
What is a little surprising, is how close the sub on ground/mic at ear level trace , is to those two traces, without some bounce cancellations.
Also look like there is a good case to made for ground stacked subs up close...
Do these observations seem correct to you?
Great work, thank you Jelmer!
At 120ms distance,
it looks like building reflections are coming in, and the effect of subs flown vs ground is getting triangulated out to minimal...
At 30ms distance,
first thing I looked for was acoustic reciprocity as was discussed earlier in thread...
Looks like the sub flown/mic on ground, and the sub on ground/mic at 14 ft,..... are about the same trace....so looks like pretty good reciprocity...
What is a little surprising, is how close the sub on ground/mic at ear level trace , is to those two traces, without some bounce cancellations.
Also look like there is a good case to made for ground stacked subs up close...
Do these observations seem correct to you?
I made all the pictures clickable so you can see a higher resolution image.I was looking at your graphs embedded in the thread last night, and only discovered this because you mentioned it this morning - so I'm mentioning it again because it really helps.
+1 Good point. This is what i tryed to explain in my previous post with the prediction software. They are going to loose efficiency in the upper range because they behave like a low pass filter including notches when the reflected wave is completly out of phase. IMO flown subs are only good to achieve high directivity control (if the array is long enough) and for deep bass slam. Ground stacking is more efficient for mid bass punch and requires less subs.[/list]
- Typically our ears aren't at ground level. With flown subs the cancellation caused by the floor bounce is a moving target over frequency response and space. This is apparent from your measurements and is proven in other replies to this thread. In the time domain, time smearing is introduced until we become uncoupled (1/3 WL).