24 replies to this topic

[Fury♫]

The dip at 700hz would most likely be baffle step.

[Pulse-R]

baffle step is a rise fury

above 800Hz, the 3dB bump looks to be baffle step.

the idea is to mount the speaker in a baffle big enough to get at least 1/4 wave at the lowest test frequency.

for 100Hz, that's ~0.9m all around the speaker, so 1.8m diameter/square.

[Fury♫]

Pulse-R, on Apr 24 2009, 02:15 AM, said:

baffle step is a rise fury

Depends if you're going up or down the step...
If you want we can call it a shelf?

[Pulse-R]

mmmmmmm, shelf

[crankin]

~thematt~, on Apr 22 2009, 01:52 PM, said:

It is unfortunate that so many people, when 'trying' drivers inside a vehicle, almost always believe that the problems they hear are properties of the driver, when in actual fact, its the acoustic response of the entire package. This includes Xover points, slopes, boundary interaction, environmental interaction amongst other things. Just because you can hear something wrong, doesnt mean its the driver's fault. Rarely it ever is.


The only concern you should have is how the driver interacts with its environment, and how THAT affects the sound.

Amen to that.

[muzzy66]

~thematt~, on Apr 22 2009, 06:37 AM, said:

There are various levels of evolution that allow people to garner a level of understanding, and apply it with increasingly good results.

You have people who deal with the theoretical (scientists)
You have people who deal with the practical (the wrenchmonkeys)

Then you have the camp of people who investigate the theoretical, like Scorpion, Pulse-R, br85 etc, to understand it, and then apply it to the practical. (Engineers). Simon (Pulse-R) and Danny (Fury) are two people I know who have evolved their understanding to the point where they actually achieve results (on MEA anyway).

It is entirely your choice as to whether or not you wish to understand the basic principles. Just dont sit in a room and sling mud on those who do, because they'll always exceed your results anyway.

And BMWTurbo, go have a look at www.Zaphaudio.com. Wonderful reference for the same sort of tests, on varying types of drivers at varying sizes.

Sorry about my first response, I know it was a little blunt / rude and for that I apologise - I respect your knowledge, but I initially took your response as somewhat of a personal attack, hence why I responded a little out of line.

I'll try to clarify myself a little.

A couple of concepts I'd like to bring forward in regards to the theoretical list of driver 'beaming' frequencies, you posted earlier:

1. My understanding, is that beaming isn't one of those things that is digital in nature (i.e. on and off). I believe a driver's dispersion pattern starts out broad, and begins to gradually narrow (i.e. move towards beaming) as the frequency rises - until eventually it reached the point where the beaming is in full effect.

If this is indeed true, then are the frequencies you listed in the above chart the frequencies at which beaming effects begin (i.e. at which dispersion begins to narrow) or is it the frequency at which beaming is in full effect?

For example, the chart suggests that beaming occurs at at 3,440hz for a 4" midrange. However, every real world measurement i've seen on any 4" midrange has shown that amplitude is singificantly reduced as early as 2,000hz / 2,500hz at 60-deg off axis. However, at 15-deg off axis that same drive may not lose amplitude until 3,000hz or 3,500hz.

Could it perhaps be the case that those charts tell you only the point at which beaming is in complete effect, and that it may indeed first begins to impacct drivers response (to a lesser degree) significantly lower in the frequency range?

This would make sense to me in theory, because as the angle of dispersion very gradually narrows, the critical angle (i.e. how far off axis you need to be to recieve the singal) would also reduce proportionally?

i.e.
To to get the signal from the a 4" speaker audibly above 3,440hz you need to be directly in front of the speaker (any off angle will alter measured response above this), to get it at 2.5khz you need to be within 30 degrees (as the dispersion is broader), to get it at 2khz you need to be within 60 degrees (as the dispersion has become broader again), etc.

2. You suggested that of axis response is dependant purely on the size of the speaker, not the brand/model. However, that doesn't take into account that every 4" driver will vary to some degree in cone area - a 4" Focal K2P mid would likey have slightly different cone area compared to a 4" Peerless HDS mid, which in turn would have a slightly different cone area to a 4" Dynaudio Esotar midrange, etc. This means that different 4" drivers from differnt brands / models would theoretically differ to some degree in how they are affected off axis (even if this difference isn't massive).

3. While the point of beaming may be influenced purely by cone area (or primarilly), standard frequency response of individual drivers can also have a significant impact on how good a driver's response is off axis. For example, the Scanspeak AirCirc has a soft rise in the upper frequency response (up around 17k-18k from memory) on axis, and so when listened to lightly off axis, that rise gets brough down a little and it ends up being 0dB at around 18khz. A Seas 27TBFC/G however, has a near perfectly flat on axis response to around 18khz before lightly rolling off. This tweeter (having a similar dome area to the AirCirc) will lose the same amount of output due to beaming, however because it's got no rise in the normal response it ends up about 3dB down (rather then 0dB down) at 18khz [by the way, these numbers are purely estimates, I don't have the graphs in front of me]. So, even though the Scan doesn't get affected any less by beaming, you could still say it has a better off-axis response because when you get it off axis, the response is indeed flatter (at the slight expense of on axis response).

This is what I meant when I stated above that formulas are lovely, however they can often be mis-interpreted. While it may indeed by true (as the formula suggests) that beaming occurs at 3,440hz for a 4" woofer, that doesn't necessarilly mean that there is no impact at 2.5khz, 2khz, or even 1.8khz (to gradually lesser degrees).

This is why I got a tad defensive earlier - as it seems you were suggesting that everything I said was essentially a load of bs, without offering me an opportunity to explain myself.

Again, I apologise for lashing out at you though - that was inappropriate.

Edited by muzzy66, 28 April 2009 - 01:42 PM.

[muzzy66]

Pulse-R, on Apr 23 2009, 05:15 PM, said:

baffle step is a rise fury

above 800Hz, the 3dB bump looks to be baffle step.

the idea is to mount the speaker in a baffle big enough to get at least 1/4 wave at the lowest test frequency.

for 100Hz, that's ~0.9m all around the speaker, so 1.8m diameter/square.

Zaph also mentioned another possible method, of building the baffle as narrow as possible. Don't quote me on this (as I'm not physicist/engineer) by from my interpretation he suggested that the two possible alternatives were to either design a baffle as broad as is physically possible (in order to being down the frequency at which baffle step occurs) or to build the baffle as narrow as possible (in order to push up the frequency at which the baffle step occurs) - both options aiming to essentially push the effect outside of the critical midrange regions.

I applied the latter to my own bookshelves when I built them, using 4" mids because they were near identical in diameter to the tweeter (reducing the aount of baffle space either side of both drivers) and also to allow the baffle to be built as narrow as possible and try to push the baffle step effects outside of the critical midrange.

Edited by muzzy66, 28 April 2009 - 01:49 PM.

[~thematt~]

See those are good questions. I can answer the ones I'm aware of with the best of my knowledge, but some I'm not full bottle on. Maybe others can chime in.

My emphasis on some of the points above is sometimes not to overtly enforce the physics behind it, but to eliminate the thoughts in many people's minds about the relative importance and understanding of something. Such as with amplifier damping. When I say 'damping is a pathetic and useless number', its not entirely useless, but for anyone on here that has ever quoted it, its impact to them is useless because they dont understand its limitations. They look for fools gold, expecting to get rich.

The frequencies given are fairly close to where the effect begins. Plus or minus half an octave-ish region. Beaming is indeed a progressive result, and sharpens as the frequency gets higher. Obviously it will affect the 90deg off-axis result before it affects the 15deg. The further off-axis, the lower in the frequency curve it becomes evident. But its still a very very good approximation. The point here is that so long as you're not using widebanders or PA gear, your midrange drivers (usually 3 or 4") wont beam inside your Xover region to any noticeable degree. Especially given the fact that its supplemented by a tweeter. That being stated, 'cross firing' and 'angling' speakers becomes fairly useless, in that regard.

Be very careful of measured FR plots, especially ones that dont come attached with LD/NLD plots, because the measurement technique and setup can introduce a lot of errors that become more evident, the further off-axis you go. When it comes to mounting in a vehicle, you may hear differences. Not saying you wont. Those differences though, are due to the interactions with the environment/boundary/Xover/levels etc, not the driver.

FR plots dont differentiate between LD and original signal either. You may have the 2nd Harmonic at 3k, cone breakup at 4k and 3rd Harmonic at 6k. Rotate the speaker off-axis, and you'll get a drop in the 3rd Harmonic, cone breakup will smooth out and the 2nd harmonic peak will still be there strong. Is this audible? Well, yes it is, but none of the above is from the original signal. Its all noise. Changing the angle of a speaker to maintain high levels of 'noise' isnt exactly in anyone's interest, lest not on an SQ forum.

My statement relating to beaming and cone size is to encourage the elimination of 'I dont get a good off-axis response, so I'll sell brand X and buy brand Y to see if thats any better'. Its a VERY common response, and yet a ludicrous thing to do when you understand beaming. If you want better, buy smaller. Either that, or build a waveguide. Simple. If I wanted great off-axis response, I'm not going to buy an Aircirc. I'm going to buy a Hiq. Why? Because the Hiq is a 0.75" dome, not a 1". No magic, marketing, figure fiddling or 'incredibly scientifically amazing design' is going to make a 1" beam less, to be able to compare to a 0.75". Unless I make a waveguide.

Once a basic understanding of beaming is obtained, then you begin to investigate waveguides. Once you understand waveguides, you begin to understand how some speakers 'measure' so fantastically well. The Dyn Esotar for example. Its a 1" dome speaker. Nice motor. Nothing overly special. The thing that makes it perform so amazingly well is the baffle. You are buying the baffle with the speaker. Sure, some people can build a baffle that gives the same response for $100, but there you go. As does the Hiq. Its not some super-duper inverted dome wonderbra technology (like the Utopia Be claims, for example), but a very simple, well-executed baffle/waveguide. You can do the same thing with the same motor, and build the baffle yourself (if done well enough).

Your point three though, is well executed. Yes, there will be a difference between different brands, but for what reason? Why does the Aircirc rise, and the Seas 27c does not? Is it because the Aircirc solved beaming? Does it actually have a better off-axis response? Thats a no, and a no, on both points. The Aircirc rises because of its motor. What goes up, must come down. If the rise is the action, what is the effect? What is the impact of that rise? Have a look at the distortion curves, and tell me that the rise is all good news.

It short, in response to the OP's question, my statement was boldly stating that aiming doesnt have an effect in the way he intended it too. It doesnt produce a 'good' response when pointing at you, and a 'bad' response when pointing away. Installation is key. Build a waveguide and you can have a 'good' response all the time. Cross a midrange below its beaming frequency, and install it well with respect to reflections, and you dont need to aim at all.

You can spend months, rotating a speaker to get it to face the 'right way' to give you the desired result. Good for you. You can also spend a day to make a half-decent waveguide that provides the same response. Much better. Hence my statement

Quote

The only concern you should have is how the driver interacts with its environment, and how THAT affects the sound.

[muzzy66]

~thematt~, on Apr 28 2009, 07:42 AM, said:

The frequencies given are fairly close to where the effect begins.

Are you absolutely certain that they are the frequencies they begin at, as opposed to the frequency at which the dispersition becomes completely "beamed" (for want of a better term)?

Only reason I'm being so persistent about this, it seems to make perfect sense. If the full effect of beaming occurs at 3,440hz (approximately) for a 4" driver, but also had impacts at lower freqnencies (the extent of which reduces as the frequency does) then that would completely explain why every 4" mid I've seen data on loses something as early as 2khz/2.5khz, but not a single 4" I've seen goes without loss at 3.5khz off axis (even if its only by 15 degrees).

At that 3.5khz point, every single 4" seems to be down significantly no matter how small the angle, so that would make complete sense if 3.5k is the point at will the full impact of beaming is felt.

If beaming only begins to impact at 3.5khz for a 4" driver, then you would think that the effects of beaming (and hence the loss of amplitude) would only be minor at that frequency, however.

I'm sure I also read a quote on Zaph Audio where he stated that he crosses over mids in his designs as low as possible so as to escape off axis drop-of (and broaden critical angle for listening). Don't quote me on it though, because wherever the quote was I can no longer find it!

~thematt~, on Apr 28 2009, 07:42 AM, said:

What goes up, must come down. If the rise is the action, what is the effect? What is the impact of that rise? Have a look at the distortion curves, and tell me that the rise is all good news.

The AirCirc's 'rise' exists between 15khz - 20khz on axis, and isn't present on the Seas (it actually begins to roll off from around 15khz).

Sadly the test's I've seen of both drivers sho response up to 20khz, but only plot distortion to 10khz. From 2.5khz up to this point, the AirCirc has significantly lower distortion when compared to the Seas but I sadly can't compare the two at the point of the rise (15khz) as I don't have data available.

I understand that you prefer to suggest people use wave guides and techniques to avoid off of output off axis, but for many people (especially those who lack knowledge of these, and the install skills to create an effective one) it can be much easier to simple rotate a tweeter (or mid) a few degrees to the left and improve their midrange response this way. It may not be the most complicated method, but it generaly does achieve the desired result

One way or the other, I really do need to do some more research into waveguides etc, because to be completely honest I don't have a huge knowledge of them at all - and I really don't want to make any arguments for or against them unless I have the knowledge (either physical or conceptual) to back it up. I've been wanting to read that waveguide thread in this forum, but I only pop on here maybe once or twice a day for a short time and reading that thread seems it may take an hour!

Edited by muzzy66, 28 April 2009 - 05:30 PM.

[~thematt~]

muzzy66, on Apr 28 2009, 03:24 PM, said:

Are you absolutely certain that they are the frequencies they begin at, as opposed to the frequency at which the dispersition becomes completely "beamed" (for want of a better term)?

.....

I'm sure I also read a quote on Zaph Audio where he stated that he crosses over mids in his designs as low as possible so as to escape off axis drop-of (and broaden critical angle for listening).

I'll tell you what, I'll take on an action to review more in-depth, the actual impacts of beaming and whether the calcs refer to full-effectiveness or it only 'beginning'. I'll come back when I have an answer.

Zaph designs his speakers based on his knowledge and understanding, and is very good at making the required changes to a 'system' (as a whole) knowing the positives and negatives of each change. He knows all about beaming, but given he operates in a home audio environment, those changes are to allow proper crossover points between drivers, not to compensate for poor location due to environment.

The biggest problem I have with 'rotating' a driver is, you're not just 'rotating' a driver. You're impacting heavily on the entire system. Its the interaction of the environment that you're effectively changing, the way the driver interacts with reflective surfaces, its baffle, diffraction, introducing (or eliminating) comb filtering, diffusion etc. Acoustic Xover points and slopes will change. Before you can get this right, you need to understand why you're doing it. Otherwise you'll be there for months.

Thats the great thing about waveguides. Besides simply the benefits that a waveguide produces, you KNOW you're modifying the environmental interaction, not just 'the direction the driver points'. You do it because you're understanding more about the interaction