Home Theater Forum and Systems banner
1 - 20 of 25 Posts

·
Registered
Joined
·
35 Posts
Discussion Starter · #1 ·
Hi Everyone,

What is the best process to use REW for detecting the modal changes using sub woofer risers.
Can rew measure overtones/undertones, vertical mode issues etc.

I'm just trying to put together an accurate picture measuring the changes between no risers and risers.

I've done SPL ,waterfalls, Spectral decay graphs but I'm not sure i'm digging deep enough.

Any advice would be appreciated. Thanks!

Test,
 

·
Premium Member
Joined
·
1,400 Posts
What is the best process to use REW for detecting the modal changes using sub woofer risers.
Can rew measure overtones/undertones, vertical mode issues etc.
Don't know about best, but check this out ;

Open up the EQ window , then open the "Modal Analysis" drop-down menu . This sub-window, sounds like it addresses your initial query .

You'll need to read the help file about this section / & then play around with the settings to capture the relevant info .
- ( fyi, I haven't used this feature / so I can't give you any worthwhile tips on it's usage ) .





<. EarlK
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #3 ·
Hi EarlK,

Oh man!:clap: Thats perfect! I think that may be just what I need. Especially the T60 data.

I don't have an external eq just yet so I didn't even think to go into that section.

Thanks!

Test,
 

·
Banned
Joined
·
516 Posts
Hi Everyone,

What is the best process to use REW for detecting the modal changes using sub woofer risers.
Can rew measure overtones/undertones, vertical mode issues etc.

I'm just trying to put together an accurate picture measuring the changes between no risers and risers.

I've done SPL ,waterfalls, Spectral decay graphs but I'm not sure i'm digging deep enough.

Any advice would be appreciated. Thanks!

Test,
May I suggest that you are not really interested in "modal" changes.
First, a riser will not constitute a massive boundary significantly modifying the actual boundaries determining modal behavior.
Instead you will most likely have more issues with LF resonance contributed by the diaphragmatic action of the riser.
For this the waterfall will be most useful.
Additionally (alternatively), if you have access to a contact accelerometer (a 'contact mic'), when securely attached to the riser surface, you will have an even better idea of the resonant frequencies of the riser itself.

From this, damping in the form of bracing (and possibly porous fill) can be performed and the effectiveness rather quickly determined.

Also, at the wavelengths at issue, you will not have any meaningful reverberant soundfield in a small acoustical space - so I would suggest that any calculation derived from the RT functions will be spurious at best. Anything derived from these calculations will not represent reality.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #6 ·
May I suggest that you are not really interested in "modal" changes.
First, a riser will not constitute a massive boundary significantly modifying the actual boundaries determining modal behavior.
Instead you will most likely have more issues with LF resonance contributed by the diaphragmatic action of the riser.
For this the waterfall will be most useful.
Additionally (alternatively), if you have access to a contact accelerometer (a 'contact mic'), when securely attached to the riser surface, you will have an even better idea of the resonant frequencies of the riser itself.

From this, damping in the form of bracing (and possibly porous fill) can be performed and the effectiveness rather quickly determined.

Also, at the wavelengths at issue, you will not have any meaningful reverberant soundfield in a small acoustical space - so I would suggest that any calculation derived from the RT functions will be spurious at best. Anything derived from these calculations will not represent reality.
Hi,

Let me clarify, I'm interested in the vertical axial modes and decay times in a single listing position. Be that re-verb, resonance or reflections in ms.

Basically, I'm trying to find a way to capture data that either proves or disproves the benefits of home built sub risers in a single listening position with all things being equal. Meaning , no EQ, no room treatments just 2 subs placed 2.5-3 feet off the ground (7 foot 10 inch ceiling) with no subtrap in the riser.

SPL measurements won't reveal much. Waterfalls are hard to interpret (for me) but RT-60 data gives numbers that can be compared.
I don't have access to a contact accelerometer. Just a basic REW setup using a decent mic.


Test,
 

·
Premium Member
Joined
·
1,400 Posts
This may be a silly question but what is the difference between RT-30 and RT-60?
Are they interchangeable and measure the same thing?
- Here's an online FAQ page that will help answer your question about their definitions. ( Click the pic to see more text. )




Which to Apply ?

- As you can see, the question of when to use RT60 or RT30 ( or even RT20 ) is partly driven by the nominal level of the noise floor for the frequency area of interest ( within the room being tested ) / as well as ( IMO ) applying a reasonable excitation level ( or the average listening level ) for audio activities in that room .

- Look at the graph for one of my speakers located within the "smallish" living room of my flat / "apartment" .

- That excitation level isn't far off from my listening reality of living with neighbors above me / while the obviously ( fairly high ) noise floor from living on a busy street ( seen starting below @ 1500 hz ) somewhat dictates my time-filter choice ( ie ; I have barely enough "acoustical" dynamic range for RT30 to tell me anything meaningful for measurements below 1500 hz ) / above 2000 hz I have a lot more leeway to apply RT60 filtering ( this is due to the fact that I do have some acoustic foam panels in place and use a directional speaker source ( horn-loaded HF drivers , acting above 1K ).




- As previously mentioned , I don't really use this feature very much .

<> EarlK

PS : Also look at using the "Spectrogram" feature ( in conjunction with the window housing EQ/WaterFall/RT-30, for agreements between the 2 ) to look for those long duration resonances . It offers a good "Look-Down" view of the "sea-floor" ( so-to-speak ) and can help pin-point frequency specific problems ( vs ; pseudo-random noise sources ) . If your sub-risers are changing the LF content going out into the room / you should be able to ( eventually ) identify if those changes are good or bad ( with enough parsing of the collected info ) .



- Hopefully these explanations garner more than a, "Clear as Mud" tag :eek: .
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #8 ·
- Hopefully these explanations garner more than a, "Clear as Mud" tag

Read more: Home Theater Forum and Systems - HomeTheaterShack.com - Reply to Topic
Earl, fantastic post!
This really helps immensely!

I can't tell you how long I've been looking for exactly what you posted.

Thank You!
 

·
Banned
Joined
·
516 Posts
Sorry, I misinterpreted your question as wanting to know the impact of a riser (its behavior) on the room response, not simply on the effect changing the position of a speaker has on the modal distribution.
So while the particular methodology to determine the effect was not appropriate, the larger issue of what and how to interpret the results stand.

Modes have nothing to do with reverberant sound fields. They are technically a resonance based upon reinforced superposition determined by the energy wavelength and the room dimensions.
Without going too deeply in to definitions, a reverberant soundfield is a statistically diffuse field with no net transfer of energy - the origin of energy at any point in the space is equally probable from any direction. In other words, it is not resolvable into a vectored source. It arrives 'equally' from all directions simultaneously.

As extensively discussed by folks like Ted Schultz (of Bolt, Beranek & Newman) and others ~30 years ago in various AES and ASA papers, there is no appreciable reverberant sound field in a small acoustical space. In fact, this characteristic is one of the fundamental aspects that distinguishes a small acoustical space from a large acoustical space.

In a small acoustical space the primary behaviors that characterize the space are LF standing waves (modal behavior) and specular behavior at the point where the energy wavelengths become equal to or smaller than the room dimensions. (And more to the point, the only reverberant sound field that will exist is at frequencies far too high to be of concern.)

The larger significance being that RT calculations, based upon behavioral assumptions not in evidence in a small acoustical space, are of little use. Oh, they will give you an output, but only based upon faulty input!

Thus the most practical tools that will actually represent the behavior you are curious about are the cumulative spectral decay and waterfall (or the same information viewed as a spectrograph, as mentioned). This will display the energy distribution at a SINGLE spot in the room.

This test must be repeated for each additional spot in the room that one wishes to evaluate.

Additionally, another very practical tool for the mapping of the space, is the use of tones. Being careful not to overheat your speakers, stepping through the frequency spectrum, one can easily identify and note the various resonant frequencies and the nulls for a particular spot. This can be rather quickly repeated for other locations and noted on a grid map system drawn on graph paper with the additional provisions for vertical notations, noting both the horizontal and vertical distribution of the nodes and anti-nodes in the room. And be aware, there will be both horizontal and vertical modes. If you are only concerned with one listening height, this restriction in the degrees of freedom with make the process proceed that much more quickly. In that case, a wheeled chair adjusted to the proper listening height can make things rather easy - especially for checking multiple positions in the room for each particular frequency.

BTW, http://www.hunecke.de/en/calculators/room-eigenmodes.htmlthis modal calculator is useful in visualizing some of the 'general' modal distributions in 3 space - both horizontally and vertically - as long as you realize the idealized nature of modal calculators and their assumptions and understand how a real world room will vary from such predictions...

Oh, and depending upon the location of the speakers, you might also benefit from examining the potential additional effects of SBIR as well as various first order reflections off surfaces such as the floor. There are several good SBIR and floor bounce calculators available online for this including Thomas Barefoot's bounce calculator, that may also prove useful in taking into account non-modal localized reflections that might effect your response as well...

Thus you map the physical distribution for all of the listening positions of concern.

The persistence of the resonance is displayed in the waterfall or spectrograph. But as this is not a reverberant sound field, RT calculations (they are not measurements per se) are not particularly useful and do not present the an accurate representation that so many generally assume they do. In other words, you use the right tool for the problem, and RT calculations are not appropriate for modal behavior.

RT calculations are very useful in large acoustical spaces where the reverberant noise floor becomes of sufficient gain to mask low level direct signals. But such is not a concern in a small acoustical space.

Stick with the CSD/waterfall plots or the spectrograph display (if you are comfortable with this variation in displaying the same results). And try using the tones (also available in REW) to more quickly map the distribution of the various nodes(nulls) and anti-nodes(peaks) in the region of interest.

Good luck and good hunting.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #10 ·
Thank You Sac,

Thats quite the post!:gulp:

I will digest it and put it into practice.

BTW, I AM only concerned about One listening position.

I have to admit, the benefits of a sub riser vs. no sub riser are taking on a whole can of worms I didn't expect to open.:unbelievable:

Its not as clear cut or easy as some would have lead me to believe.

I'll keep learning and pluggin away though.

Thanks for your very thought out post.

Test,
 

·
Banned
Joined
·
516 Posts
If you are simply concerned with the modal response at one listening position, things are actually very simple. It is over a distributed listening area that things become complex.

In a single position given a single height, things are pretty much 'go-no go'. Ideally you are hoping that the modal distribution at the listening position is neither a node nor an anti-node.

(Ignoring that factors such as SBIR can also potentially contribute to response anomalies), raising and lowering the speaker with a stand can effect the modal distribution as well as the spatial loading of the sub - with a greater effective LF output resulting from the placement of the sub against the floor-wall intersection than simply against the wall on a stand. Thus you also weigh the difference of greater effective LF energy with the sub on the floor against the wall, or slightly less effective output on the stand - combined with the modal distribution at that point.

Thus, while waterfalls and spectrographs can indeed indicate relative comparisons for the sub on the floor against the wall or on a stand against the wall, if you are simply interested in which sounds better, simply playing tones while listening at the listening position will give you a good inventory of the performance. Otherwise generate a waterfall for each of the 2 vertical speaker positions and select the one that looks best in terms of overall gain and persistence for that location.

Again, it need not be complex to simply determine which speaker position is better (or, at least, less bad...).

If both are problematic, I would then try repeating the process, moving the sub to either the left or right (at each height) and again evaluating the performance, opting for the position featuring the least pronounced anomalies

Since we are primarily concerned only with the effects of speaker location on performance, if you want to try yet another even simpler method, place the sub (securely) in the listening position, and then move about the sub mounting locations, evaluating both horizontal and vertical positioning corresponding to the height on the stand and determine which spot has the best response. Place the sub there and you are done.

...Don't make this more difficult than it is!

Complexity arises if you wish to try to treat the space and to mitigate modal behavior at a particular spot < and even more so over a larger distributed listening area, rather than having the flexibility to adjust the listening position!
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #12 ·
...Don't make this more difficult than it is!
Hi SAC,

Believe me, I had no intention of making it difficult in the first place.

But...Here's a story...:whistling:

One fine day, I came across a thread in other forum talking about the benefits of risers.
Not just for decoupling purposes either. Rather, to address vertical axial modes and RT-30 reverberant times.

My subs (I have 2) were always on the floor. Have been for a couple of years.
Below grade, concrete floor with thick burber carpet over top.
So decided to build a couple risers. The first 2 were at 7 inches. The 2nd set of risers were at 2.5 feet adjustable to 3.
After I put them up I noticed an audible difference for the better. To me anyway.

Well, i posted my experience in another forum and got asked for proof. I had no idea how to use REW at the time.
So, I didn't have any scientific data to back up my perceived "audible" improvement.

After learning how to use REW (from some great guys who post here as well) I made the basic SPL, Waterfalls, Spectral decay measurements and guess what?
They really didn't show much difference. So, I was really in a tough spot.
Did I just fall victim to the placebo effect?
Was the improvement all in my head?

It was then that I came here asking for help. I'm so glad I did.
I'm still working on which set of test data shows the difference.
So far the results have been pretty negligible.

In my room at least, yes I do have slightly shorter decay times across most frequencies from 25Hz to 120Hz with subs on the risers. The question is, Is Thats audibly detectable?

Could that be the reason why my ears are perceiving an improvement.

I'm working on it. lol


Test,
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #13 ·
Here are my results trying to pinpoint either "placebo" or "real audible" differences.









Now I realize that the RT-60 test is NOT really applicable here. I just posted because a few reviews of the ACS subtrap riser used RT-30 data to prove their case.
I will post asap the RT-30 results.

Test,
 

·
Banned
Joined
·
516 Posts
Actually, the evidence is already there!

But before we examine that, let’s very quickly peruse a few of the factors at play.(Note: I don't expect most will be interested in trying to break down the complex factors that contribute to this specific case, especially as we are dealing with both incomplete behavioral information and also a reality that is complex in that it exists in a gray area somewhere between oft cited 'ideal' model behavior, and are thus a bit more complex and less clear cut than one might hope. And for those folks, there are plenty of other posts to spend their time reading. But, if anyone wants to have a bit of fun balancing the various contributory issues...)

There are three primary factors (of which we will focus primarily on two).

One is spatial loading. Very simply, one can think of this as a given amount of energy radiating into a volume of space. To use a very simple example, let’s compare a given amount of energy radiating into a given space versus the same energy radiating into a volume only ½ as great. The effective energy density will be twice as great in the volume that is ½ as large; and the energy density will be ½ as great for the energy radiated in to a volume twice as large. There is an inverse relationship.

When a speaker is placed within the space adjacent to a reflective boundary, part of the energy is radiated directly into the space, and part of the energy is reflected by the boundary back into the space.

For the purposes of this discussion (and in the absence of more detailed information) we will assume a point source that radiates energy equally in all directions (but realize a real world device is anything but this!!! And recognize the practical deviation from theory that results).

So, for our purposes, a speaker positioned at the juncture of the floor and wall will radiate energy into ¼ of a sphere – with a Q of 4 (or pi steradians – in spherical measure). This would correspond to the original subwoofer position.

Alternately, when the speaker is mounted off the floor in the ‘middle’ of the reflective boundary, the degrees of freedom are increased, and the energy is radiated in more directions (1/2 a sphere) with a Q=2, or 2pi steradians. This would approximate the placement of the subwoofer on a riser against the wall.

The second directly related aspect (an important factor but one of which we will not dwell too much as we lack any such objective information) is the loudspeaker directivity factor – the actual spatial dispersion of the energy from the speaker itself. This is determined by a variety of speaker design factors and the specific distribution if beyond the scope of this post. But suffice it to say that to the degree that the radiated energy is not omnidirectional, this will further affect the distribution of the propagated energy in the available space and modify the above mentioned statistical energy distribution predictions.

The third aspect that affects the response within the room is the modal response. This constructive (additive) reinforcement results in regions of superposed (combined) resonance manifest as a standing wave – meaning that the low frequency energy distribution wave is distributed in space in a regionally identifiable manner. While the placement of the energy source can modify the spatial distribution of the energy, the massive boundary room dimensions determine the frequencies affected.

So, what does all of this mean?

By moving the speaker from the intersection of two reflective boundaries to that of simply one boundary, we partially change the spatial loading of the speaker. The reason that it is only partial change is that the additional boundary is not so far away as to be rendered inconsequential and remains partially active in the scenario.

Energy that is within ¼ wavelength from the boundary will be reflected and sum coherently with the source, thus supporting those frequencies. Energy that is greater than ¼ wavelength away, will combine destructively. Thus for a source spaced further away from the additional boundary, the effective volume will be reduced to the degree that energy greater than ¼ wavelength away will not be reinforced.

And that is exactly what we see in the results.

Using 2.5 feet as a rough measure (as this does not accurately account for the spacing from boundary to acoustical origin of the source), this measure corresponds to a quarter wavelength of a 10 foot wavelength with a frequency of ~112 Hz. Thus energy at frequencies below 112 Hz that are incident with the floor boundary will still be reinforced (as it is at the floor-wall intersection), while those frequencies higher than ~112Hz with wavelengths greater than 1/4 wavelength will not be reinforced - just as we see. (But note, all of this energy will not cease to be reinforced, as the speaker is still against one of the boundaries. So the difference will not be as dramatic as one might expect! For you see, even if the effect were ideal, it would not be much more than ~3 dB SPL! And on the scale selected for display, that is a very small incremental difference!!!

And if one will note the waterfall of the raised source, that is exactly what we see – both a slight increase followed by a slight decrease in gain of the energy above ~112Hz. We have a source that is slightly closer with a bit more direct energy compared to the floor mount in the reinforced frequencies, while having slightly less spatial loading gain a bit higher. Not a HUGE difference, but it is there. And as these frequencies are closer to the more critical ranges of hearing, their presence, or lack thereof, will be noticeable (to some degree).

Now I am not sure exactly what one expected to see in terms of a change in response due to a difference of simply 2.5 feet. One is certainly not going to experience some difference of 15-20 dB SPL! Nor is the change able to be examines in such a discussion as a ‘pure’ change – meaning that each of the principles stated is not implemented in a technically ideal manner! Thus the effect of each variable is partial and relative.

As the difference in position is relatively slight, one would expect to see a slight change in the vertical distribution of the vertical modes. Likewise, we see a slight modification in the reinforcement of the mid bass frequencies above ~112 hz.

I suspect that much of the changes are minimized as the listener is still within the ‘nearfield’** response of the speaker. By this I mean that the direct energy is dominant (>~15 dB) over the reinforced energy reflected by the various surfaces.

Thus we do not expect a dramatic change. But we reasonably do expect a small change, which is indeed manifest.

But as such issues as modal behavior are still determined by the massive room boundaries which have not changed, and the energy being distributed has not experienced a change other than by way of position, we are not going to see a radical change in modal behavior. At most we might see a slight shifting of the modal distribution within the room which may or may not specifically effect the particular listening position at issue.

Thus to recap, we have a number of variables contributing to various degrees to the response of the low frequencies. As each exists in an ‘incomplete’ sense, the effects will be real, but they may not be dramatic. And this is exactly what is seen. In the gain response in the region above ~112 Hz. And to a lesser degree, a slight modification of the spatial distribution of the modal response.

Measurements are a great resource, but they are not a substitution for understanding the underlying behavior. Instead while they can sometimes make gross differences readily apparent, they can nevertheless be used to help to understand more subtle changes - especially those that are a result of multiple interacting variables. In some cases. more tests to isolate various variables might be worthwhile. In this case I think we are quickly reaching a point where we can spend allot more effort to determine what we an already reasonable predict based upon basic physics behavioral models. And as has been stated, it is simply the attempt to correlate perceived behavior with general tests. And in this case we are working with quite a few contributory parts of the specific behavioral puzzle missing (such as the directivity of the source and the acoustical impedance of the various front wall, side wall and floor wall surfaces that will each impact how much energy is reflected and the frequency content of said reflected energy from each surface. So we simply assume they are equal - which could constitute a significant source of error..

But I fear the disappointment in such investigation may be that the expectations are that of a dramatic change in modal behavior. I would expect little to no change in the degree of dominant axial modal behavior (e.g.: time decay aspect of the modal resonances) as neither the determinate variables of energy nor room dimensions have been altered. And to the degree that we have NO reverberant sound field at such long wavelengths in a small acoustical space, any reference to RT calculations are spurious. Neither modes not specular reflections, both of which dominate and define a small acoustical space, have anything to do with reverberant soundfields – in fact, they are the antithesis of such reverberant behavior! So references to RT calcs are a red herring in such a space. Save the attention to RT calcs for large acoustical spaces!

-----------------------------------------------------------------------------------

I apologize if any of the terms or their use are confusing. Some of the terms refer to specific acoustical definitions (of which some may not be intimate), and in other cases some of the topics are very quickly mentioned in passing, assuming some pre-requisite understanding – as this was not meant to be a complete acoustics text! (despite its length!) So if anyone wishes any clarification or explanation in simpler terms, PM me with a US phone number or a Skype/Yahoo Messenger username and we can talk and address any specific concerns or confusion that may exist. Also, there are graphics that an make many of the mentioned concepts much clearer as well. We can dig into this as casually or as deeply as anyone might desire. And hopefully anyone interested can walk away with a better functional understanding of just what is happening and how to quantify it in terms of measurements.

** (Note: I am not using the term (s) 'nearfield/farfield' in the usual strict sense addressing the relative intensities of energies from multiple drivers with respect to each other relative to distance and orientation, beyond which the soundfield becomes ‘far field’ and the relative interdriver balance ceases to change with angle and distance; and the response field becomes uniform with angular and distance issues. Instead we are here focusing on the ratio of direct to reflected energy. Expressing this more clearly and simply, in the traditional use of the term “nearfield” the response is specified in terms of the interaction of individual drivers and the perceived character changes with relative angle and distance to each individual driver; while in the far field the response is coherent and the response ceases to vary with angle and distance within its uniform region of the power spectrum. To the degree that the direct energy can be expected exhibit a level greater than ~15 dB relative to the reflected energy, the direct energy dominates and tends to mask the reflected energy.) I hope that the two different uses of this term do not confuse folks here.
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #16 ·
Oh jeez test
WHAT?:yikes::rofl2::D

I tried Mpray. I thinks its time I drop the issue.
I doubt any amount of evidence would ever convince anyone either way (for risers or against).
I'm almost certain that if I had 6 people over too listen too the subs on and off the risers, 3 would say OH yea big difference and the other 3 wouldn't notice anything at all.

Blah.

Time to use my new found knowledge to improve my room using REW a BFD, some bass traps and other room treatments.

It was fun though!

Test,
 

·
Registered
Joined
·
35 Posts
Discussion Starter · #17 ·
Thanks SAC,

Another great post! I had to read it 3 times but I understand it all now.:scratch::D

Pretty much your post is what I needed to put this bit of testing to bed.
It pretty apparent that any benefits will be in the Upper mid bass range (which is good) but in my case not really enough to write home about.


Thanks for your awesome articles!

Test,
 

·
Registered
Joined
·
112 Posts
test4echo101 said:
WHAT?:yikes::rofl2::D

I tried Mpray. I thinks its time I drop the issue.
I doubt any amount of evidence would ever convince anyone either way (for risers or against).
I'm almost certain that if I had 6 people over too listen too the subs on and off the risers, 3 would say OH yea big difference and the other 3 wouldn't notice anything at all.

Blah.

Time to use my new found knowledge to improve my room using REW a BFD, some bass traps and other room treatments.

It was fun though!

Test,
At least you learned how to use REW out of it.
 
1 - 20 of 25 Posts
Top