August 26, 2009
Here are my intial measurements taken in my room using the HT shack Room EQ Wizard program (REW). I used a Behringer ECM 8000 measurement mic and a Behginer XENXY 502 mixer as a mic preamp. The following measurements are easy to perform, and the equipment is not costly. If you are serious about getting the best bass in your room, it's the smart thing to do.
The sub I'm measuring is the Rythmik 12" servo sealed sub which I have in a 70L box. I have two of them but measured just one. The idea behind these measurements is to determine the best locations in order the get the smoothest possible in room response.
First, nearfield measurements performed with the mic as close as possible to the cone. This removes the influence of the room.
August 24, 2009
In Part 1 I focused mainly on showing how using multiple moderate subwoofers can achieve higher output than a single monster sub. The real benefit, however, lies in the improvement in sound quality. The multi sub approach isn't new. The material here is largely based on online forum discussions with Dr Earl Geddes.
The problem isn't what you think it is
It's very easy to think that the way to achieve the ultimate bass is to buy the ultimate subwoofer. This isn't the case, although it's a good starting point. The reason is that the room itself introduces big problems. If you compare a cost-no-object sub placed in a room with unresolved acoustics issues, it's more than likely that three subs set up following this approach will be more accurate and cost less.
So what is the problem? In a nutshell - room modes. They are talked about a lot more than they are understood. At very low frequencies, a room is unformly pressurised and the wavelengths of these sound waves are large compared to the room dimenions. Here, we only need to move a lot of air and the output will be fairly even anywhere in the room. In the midrange and above, the wavelenths are small compared to room dimensions. Each sound wave will have peaks and dips, but they are so densely spaced that they are spatially averaged, and we don't need to pay any attention to these peaks and dips. In between these two regions, we have the "modal region" where our problems start. In this region, the modes are spaced far enough apart to become a problem. In one seat you have a peak in the response, but in the next you might have a dip. Typically there are many peaks and dips in all locations.
Typical solutions don't really work
There are a few different ways that this problem is often addressed:
- live in blissful ignorance (this option is diminishing quickly for you)
- room EQ
- try to place the sub in the best position and hope for the best
- opt for dipoles
- treat the room with bass traps
It can often achieve a flat response in one location, but make matters worse at others. In well behaved rooms it is better than doing nothing at all, but it's certainly not the best solution.
Open baffle dipoles do have some merit. They share some advantages with infinite baffle, but their output is seriously limited. Their response is not necessarily smoother than monopoles, and multiple monopole sources achieve the desired goal more effectively.
Typical bass traps can help, but ideally the whole room should be designed to act as a bass trap. This makes it much easier to implement the multisub solution, but does not replace it.
This post is more of an introduction, but I'll soon write a more detailed account of the multi sub approach.
August 22, 2009
One of the keys to creating the sound system that works best for you is actually to use that one thing we've been told is evil. Tone controls are for wimps and graphic equilizers - forget it! No, no, we're serious audiophiles here, and please don't mention that word digital.
I use digital EQ in my system and it seems everyone who hears it agrees with me - it's a clear improvement. At times I hear much more expensive systems, and yet my typical response is that I find my own system more enjoyable. The main reason is that I've been able to create a target response curve that sounds best to my ears. If I had to choose, I'd rather have that than a megabuck system that doesn't allow enough shape the response. If it sounds extreme then you realise how serious I am about making this point - EQ makes a huge difference.
Benefits of EQ
It's a very rare system that won't benefit from EQ. In a nutshell, EQ makes a system sound balanced. Without realising it, you may have adjusted to the strengths and weaknesses of your own system. Let's say you have exaggerated bass in your room - other systems will typically sound weak in the bass. Or let's say the treble is tilted up. You may feel that flatter systems are lacking in detail and sparkle. Experimenting with EQ is a learning process that helps you to learn more about your own preferences. Chances are, they will become more refined and you'll gain much more control of the sound of your system.
The benefits are quite compelling:
- room modes that destroy bass accuracy can be tamed
- active speakers can have the levels better matched
- sub volume can be better chosen
- detailed control of the tonal balance is possible
- dynamic loudness control compensates for the ear's insensitivity to bass at low levels and can supply the right amount of boost/cut at all levels
- dynamic settings can reduce treble at higher volumes, subjectively creating a more relaxed sound at high output
- limiters can protect the system from being over driven by others
- level dependent limiters can reduce bass at high output to protect drivers from damage (an alternative to rumble filters)
Getting a flat response
In my system I use Behringer Ultracurve DEQ2496. It's a powerful unit and great value for money.
The first step is getting a flat response. You need a measurement mic which has a flat response - definitely not a dynamic mic. A DIY mic using the Panasonic capsule or Behringer ECM8000 is suitable. Place the mic in the listening position and run "Auto EQ." Ultracurve will play pink noise and automatically apply eq until a flat response is achieved. This takes about 5 minutes.
If you listen to music once you've set it up as completely flat at the listening position, you'll be surprised to find how bad it sounds.
Trying to do two things at once
We are actually aiming to do two things at once. We want to EQ the bass at the listening position, but at the same time we actually want to get the speakers flat at 1m on axis. In my particular room, I can effectively do these two things at once. I place the mic in the listening position and select the "room eq" option which applies a 1db/octave slope boosting the bass end. When I place the mic at 1m while pink noise is playing, I can see in the RTA window that the response is flat. When I place the mic in the listening position, the response tilts down towards the top end. This tilt is about the same as the 1db/octave applied by the room eq option. The result is that the eq actually makes the speaker flat at 1m for the midrange and above while correcting bass response.
In typical rooms this should work well. In a very live or dull sounding room a different approach may be needed.
Improving on flat
Once a slightly modified version of flat has been achieved, we can now shape the response a little more. This is where experimentation and personal preferences come into play. In my system I use two filters as a minimum. Firstly, a gentle roll off of the treble, starting at around 1kHz and -2.5 dB @ 20 kHz. Then, for music I add bass boost - 6dB boost at 40 Hz which rolls off either side. The boost often becomes overpowering on movies, so no boost for movies is often better. The original mix usually has exaggerated bass already.
Another reason you need subwoofers
Obviously, 5" vented woofers aren't well suited to this kind of EQ. The power and excursion demands are just too much in many cases. We are now into subwoofer territory, and if you haven't considered this option already, then it's time to think seriously about crossing to subs below 80 Hz. If you have not heard a serious sub system set up for music as well as home theatre, then you owe it to yourself to experience one.
After running AutoEQ with my DIY omni speakers and Rythmik subs, these are the settings used to achieve a flat response:
You can see that most of the eq is correcting for room modes. In virtually any arrangement I get a peak around 40 Hz and a series of peaks and dips around 80 - 200 Hz.
I use parametric EQ filters to shape the response:
You can see 6db boost centred around 40 Hz, which subjectively makes the bass twice as loud as a completely flat system. For music this generally sounds "just right." For movies I often remove this boost, as I find the result is often overpowering.
There is a "psychoacoustic dip" around 3.5k as well as a 2 db reduction of the treble. I find this gives a more relaxed sound that I prefer.
Some prefer to use PEQ to deal with modes, but I find it more intuitive to use these filters to shape a flat response. I like to see what I'm getting.
But wait, there's more!
There are a few more benefits and features I haven't mentioned. Let's say you want to try open baffle and need to compensate for the resultant roll off. No problem - simply run through the EQ process again and store a new preset. 5 minutes and it's all done. Let's say you want to make your vented box sealed. Simple again. If you like you can even recreate the same response and apply the necessary EQ. Or what if you want to re-arrange your room for a party that means you have to sit in a less than ideal spot, or push the speakers back towards the wall so the sound gets boomy. Again, not a problem - you can easily compensate and take the boom out of the bass.
A few more features that are nice to have:
- stereo width control - expand the stereo effect
- RTA - some say it's like watching a fire, I like to see the shape of what I'm hearing and how bass the bass goes
- SPL meter
- digital delay to subs
It started here for me:
Ultracurve Review by Thorsten Loesch
Note: the unit reviewed has been replaced by DEQ2496
Ultracurve on the Behringer website
An article on the "house curve" concept at the Home Theatre Forum
August 18, 2009
The conventional approach to room treatment is often adopted without considering or being aware of the options I'm going to present here. The main key to getting it right is to consider the speakers and the room together as a system. Many of the ideas presented here are based on the work of Dr Earl Geddes, however, the opinions expressed are my own.
The conventional approach
In a typical dedicated or semi dedicated room, the conventional approach to room treatment usually involves the following:
- bass traps (typically removable)
- absorption panels for first reflections
While this approach can work quite well, an alternative is presented here which many will find more satisfying.
Considering speakers and the room together
Speakers and the room together form a system. The relationship between the two is critical. Conventional speakers are often designed for their on axis response, while neglecting the power response. As a result, the off axis output which is radiated to the room has a different response. The ambient sound is therefore coloured relative to the direct sound. For this reason, room treatment is often used to reduce the damage.
An alternative approach is to first start with speakers that are designed to interact with the room in a way which does not require damage control. Some options include:
- controlled directivity speakers with waveguides
- open baffle speakers with behaviour approaching constant directivity
- omnidirectional speakers which radiate evenly in all directions
- AI Audio loudspeakers or their Gedlee kit versions
- Linkwitz Orion or the Nao
- Linkwitz Pluto or B&O Beolab
If we start with any of these speakers, we no longer need to be concerned as much about reducing reflected sound in the same way. The sound radiated by the room now matches the character of the direct sound. So what are we trying to achieve now?
I suggest the following guidelines:
- high level of bass damping
- modest amounts of diffusion
- little if any absorbing panels
- careful speaker placement
Firstly, the room itself should provide a high level of damping in the bass range. Idealy this should include the entire envelope acting as a large bass trap. Installing multiple layers of plasterboard (drywall) joined with a flexible adhesive such as Liquid Nails will achieve this goal. If the existing structure is very solid and undamped (masonry/concrete), then false walls and ceiling is a good place to start.
When starting with a room with significant damping, the problem of room modes is much easier to deal with. In addition to starting with a well damped room, it's also advisable to place multiple bass sources in locations determined by measurement. Three bass sources will typically provide a good balance of price and performance.
Read more about the multi sub approach >
Broad band treatment
Most domestic rooms can be considered acoustically small. They require different treatment to large spaces such as a commercial theatre of concert venue. Unlike these venues, it's preferable in a home environment to retain as much reverberant energy as possible. As a result, we should start with diffusion. It's important to avoid placing diffusers too close to the listening positions. Nearfield placement results in poor performance. Ideal locations include the wall behind the speakers, ceiling and side walls.
Experimentation should be used with absorbing panels. When forced to sit with a wall close behind, it may be worthwhile placing absorbing panels on the wall. Idealy such panels should be as thick as possible and relatively dense to be effective over a broad range of frequencies. Panels that are too thin will only work at high frequencies.
In nearly all cases, speakers should be given room to breathe. Firstly, the tweeter should be as close as possible to seated ear level. They should have at least one metre clearance behind and to side walls, but preferably more. In the case of omni speakers, Linkwitz recommends that they be placed wider and closer to the listener than open baffle speakers.
Read more about omni placement at Linkwitz lab >
Geddes loudspeakers have unique recommendations for toe in. Geddes recommends 45 degree toe in relative to the side walls. This means the axis of each speaker will cross in front of the listening position. As a result, the listening position will be off axis. This is discussed in an online discussion at DIY audio. View thread >
This recommendation may or may not apply to other speakers.
Ideally the decision to purchase speakers should be made while also considering room issues. If your room isn't a dedicated room and treament isn't an option, then it becomes even more important to choose speakers with a well behaved polar response. In this case especially, omni, open baffle and controlled directivity designs should be seriously considered.
More information is avaible in the white papers and scientific papers on the Harman International website >
August 17, 2009
My interest in omni speakers was sparked recently at an audio gathering I hosted. We featured three sets of speakers, my own open baffle speakers and two pairs from Gainphile, one of the other attendees. He presented his own open baffle speaker as well as a Linkwitz Pluto clone.
Listening to a familiar track on Gainphile's omni, I was struck by the imaging! On the one hand, it had the spacious sound stage I enjoyed from my open baffle speakers, yet the imaging was sharper! It gained my immediate interest as a result.
My omni experiment
I decided I had to try this for myself, and have more time to experience omni speakers. I built a simple prototype using my existing drivers and crossover, and using Ultracurve to eq to the same target house curve that I'm accustomed to. The tweeter fires forwards, while one midbass fires up to the ceiling. The crossover point is 3.5k.
How does it sound?
Unfortunately, I didn't manage to achieve the same focused imaging of Gainphile's Pluto clone. I did, however, get a chance to experience the sound and appreciate some other benefits. Firstly, the sound stage is large and spacious like an open baffle speaker. It's clearly a bigger sound than you can achieve with a box speaker. It also has a very large sweet spot, even bigger than open baffle. The entire vocal range is covered with one driver which is radiates in an omnidirectional pattern on the horizontal axis from 80 - 3.5k. As a result, the sound changes much less across the room. When hosting a dinner party I had people sitting in places that I normally wouldn't when listening to the system, yet the sound didn't change much. No other type of speaker could achieve this result as well. Most would fail terribly.
I don't feel the sound is quite as transparent as an open baffle speaker. While the sealed box is lightly stuffed with dacron, I get the impression that there is a small amount of box coloration that isn't there with open baffle.
Omni speaker theory
For a long time I felt omni speakers had the wrong idea. It didn't make much sense to me to radiate sound the same way in all directions, but instead to direct the sound where you want it. And yet the legendary Siegfried Linkwitz recommends omni speakers, so one has to at least give them some serious consideration. The idea is that it has an even polar response in all directions. This means that sounds which are reflected around the room will be closer in character to the direct source. Less coloration of reflected sounds results. The increased reflected sound creates a larger sound stage and creates a spacious sound.
In order for a driver to operate as an omnidirectional source, there is a relationship between the size of the driver, and the wavelength. A 6.5" driver will be omnidirectional below 1k in all directions. The Linkwitz Pluto therefore crosses at this point to the tweeter. Most tweeters aren't suitable for such a low crossover point, so a larger driver with high excursion is used. The 2" Aura driver used in the Pluto operates as an omni source up to 3k.
It becomes clear that my prototype is a compromise. The midbass should ideally be crossed at 1k, but since it is crossed at 3.5k it is increasingly directional above 1k. This means more sound is directed towards the ceiling. For permanent use, this could be addressed with acoustic treatment. From a horizontal point of view it behaves as an omni through it's entire range.
More information, as well as kits are available at Linkwitz Lab
Pros and Cons
+ small and unique form factor possible
+ very good DIY kits available
+ huge sweet spot allowing them to be appreciated over a much larger area
+ spacious sound stage of open baffles yet with potential for sharper imaging
+ even polar response results in less coloured room reflections
- difficult to achieve high efficiency or high output
- very low crossover point required for tweeter
Overall omni speakers are a very attractive option which are rarely considered at all.
The ideal application for omni speakers is as a music system in a typical living room. High output won't be required and their huge sweet spot will mean they are appreciated more by all. Their novel form factor and modest size is also a bonus in this application. This isn't to suggest they aren't suitable for serious use in a dedicated room - indeed they are well suited to this. However, my personal preference is for speakers with higher output capability.
Looking for something more conventional?
TL speakers >
Looking for a speaker out of the box?
Open baffle speakers >
Yes, you heard it right - two subwoofers isn't really up to the task. I know what you're thinking and no this isn't about insane output. The primary reason for using three subs is actually all about sound quality! The increase in headroom and output is also a key benefit, but not the main reason. Of course, not everyone has a dedicated room. The context of this post is a serious dedicated home theatre or music room in which we are aiming for accuracy and fairly serious output.
The problem with one subwoofer
There are two problems with attempting to do the job with just one subwoofer:
- Output - the excursion and thermal power requirements are expensive and detrimental to accuracy
- Room acoustics - room mode issues aren't properly addressed
If the driver is placed in a sealed box, the efficiency will be very low at 20 Hz. In many cases it will be as low as 70 db 1w 1m! Compared to a driver in a vented box, the excursion requirement is doubled. If we place this ultra high excursion driver in a vented box it now becomes very difficult to deal with vent chuffing with air moving at high velocity. At very high excursion it become impractical to design a vented box. The typical solution then becomes using two passive radiators. This removes the vent turbulence problems, but further increases the cost.
Of course, all of these challenges can be navigated and balanced - this is a part of good design. However, the purpose of this post is to present a sensible alternative.
Let's consider an example to illustrate. Simulations have been performed in Win ISD pro alpha.
1. 3 modest subs with Peerless XLS (12") in an 80L vented box tuned at 22 Hz with a modest 300 w input each.
2. One sealed sub with Exodus Audio Maelstrom-x (18") in a 200L sealed box with 1.5kw input power.
Max SPL chart - Red: 3 x 80L Peerless XLS 12" vented Green: Exodus Maelstrom-x 18" sealed
On paper it doesn't look like much difference, but in fact in an actual room the difference in output could in fact be easily increased. In a room with substantial gain, different filtering could be used to get more output in the top end, with room gain brining up the bottom end to match. The efficiency of each Peerless driver is about the same as the larger Maelstrom, and the total radiating area is greater.
The Maelstrom uses 33mm one way excursion to achieve this output, but the Peerless drops down to 5mm at tuning with a maximum of 12mm at 30 Hz. As a result, the Peerless can be used in a vented box and air velocity remains at an acceptable level with a practical vent.
Please note: the point is not to attempt to show that either driver is superior, but to illustrate a different approach to achieving a target. Both are excellent drivers which are also very different. Many would in fact use the Maelstrom in a very large vented box tuned below 20 Hz. Used in this way with a very powerful amplifier it could achieve things which the three Peerless subs couldn't.
So it can be seen that the three modest subs can actually achieve more output than the single large sub. They do it with less power and excursion, and will often fit more easily into the room.
The real advantage, however, is the way in which three subs can integrate into the room. This will be covered in Why two subs isn't enough - Part 1
Producing an impressive pair of DIY speakers that perform well beyond their price is easier than you think. Of course, hardcore DIY enthusiasts will do it the hard way - it's their hobby and it's all about putting in the most amount of effort for small improvements! But for the rest of us, there are ways to make it easier.
First, lets talk about the hard part - coming up with a crossover and driver combination. Hardcore enthusiasts will buy their own measurement equipment, invest a lot of time learning to use it all, and then go through many different crossover iterations and even go as far as measuring many different drivers before deciding which ones to use. You won't find instructions here on how to become the speaker design guru others will aspire to emulate. Instead you'll find a starting point in getting there the easy way. It's not easy to the point of being completely idiot proof, but it's easy compared to designing everything from scratch and reinventing the wheel just for fun.
What's the big deal about crossover design?
Surely it's just soldering a few bits together - a few resistors, capacitors and inductors, right? Not quite. There are many different factors to juggle and get right. Crossover design is a big part of what makes two loudspeakers using identical enclosures and drivers sound different. Some describe it as an art, others as science but you might call it a combination of both. It's certainly not something you would expect to perfect on your first project. Considering that it's more challenging than it sounds, and that the experts take many years in perfecting the art and science of crossover design, it makes a lot of sense to follow in the shoes of those who know what they are doing. At least, for a first diy speaker project.
I suggest you are best to start with a project that has a very good chance of working very well. It will be rewarding, and quite likely a good introduction to a long term satisfying hobby.
Where to find existing designs
First, I'll suggest some good quality websites which feature a range of competent designs that many have built:
Humble Homemade Hifi
This is the website of Tony Gee featuring a large range of speakers. Many of them would be sheer madness to try as a first time project.
Another popular suspect is John "Zaph" Krutke of Zaph audio. John has measured a very large collection of drivers. There's a good chance that he's already measured the any popular DIY driver you are interested in.
Troels Gravesen DIY projects
A great range of projects, many of which will appeal to those looking for less conventional designs including hybrid open baffle and efficient designs. He has even gone as far as developing his own driver!
You can also find a good selection of projects at Parts Express in their DIY project showcase. There is considerable variety in the projects, including some novel and innovating designs.
Share and learn
One of the best places to share your DIY experiences and learn is DIY audio. It features a HUGE gallery of projects in the System Pictures and Description thread. There are some inspirational projects in there. Some of them are so good they look like they were built in a factory. Some of them are that good because they were built in a factory.
If you do follow these links and build one of the projects, make sure you come back and tell me about it!
August 16, 2009
Experience has taught me that the best bass doesn't always come from a subwoofer. There are still many subs that don't match the accuracy of a good woofer. My Rythmik subs restored my faith in subwoofers.
Admitting to bias
I should admit up front that I can't be completely unbiased when talking about Rythmik audio. Back in 2006 I suspected that they had a great product backed by very sound theory, and proposed a website redesign. I had to hear them for myself, and I also wanted to know if they really were as good at the website asserts. It didn't take me long to verify that they really are exceptional.
It didn't take long to realise I was listening to accurate subs. The absence of distortion was instantly recognizable and the sound was dry from the start. The sound reminded me of the sound of some Focal woofers I'd heard at Aranmah acoustics. The cost of those woofers is about the same as buying the Rythmik driver and amp. Not bad considering the Rythmik has greater extension and output.
Set up and calibration
Set up correctly, the Rythmik subs are magic. But like any subwoofer, it's possible to set them up so that they sound terrible. In my system I use Behringer Ultracurve with a Behringer ECM mic for EQ. I use it to shape the response and this includes the bass response. It also helps determining suitable sub locations. In my room placement isn't critical - the room has a lot of bass damping, which means most practical locations for subs work well.
I use an 80 Hz crossover point and a 4th order electrical high pass supplied by the plate amp.
An earlier version of a Rythmik sub has been independently measured at the Home Theatre Shack:
HT shack Rythmik measurements
One aspect clearly shown in the measurements is that the distortion is low at all drive levels within the range required for music. The distortion below 40 Hz climbs fairly high, but this is mostly related to more artificial home theatre LFE use, for which the ear often has no real life reference. Subjectively the Rythmik subs are very clean and dry.
My previous subs sounded better in a H frame dipole arrangement, but I was surprised to find that the Rythmik subs sound just as accurate in a sealed box. I believe the reason is related to the higher level of damping control provided by the servo which eliminates box coloration in a similar way to infinite baffle and dipole subs. I've written more about this in my post about my open baffle speakers.
I have no hesitation in recommending these subs to anyone for serious home theatre and music listening. They don't use the highest excursion drivers or the most powerful amps, but I believe they operate in a sweet spot. Where more output is desired, I suggest multiple subwoofers, which is actually better anyway in terms of room acoustics.
Rythmik Audio Website
It all started with a saturday afternoon experiment. A very quick and dirty chipboard baffle and H frame. Three years later I'm still hooked on the sound. The difference became most obvious when I tried to go back to box speakers. After listening for a few weeks then changing back, it became clear that open baffle speakers have a certain magic to them I find hard to live without.
What is an open baffle speaker?
An open baffle speaker is variant of a dipole where no box is used. Electrostatic loudspeakers are also an open baffle dipole, but I use the term here to refer to speakers based on conventional dynamic drivers.
The key attributes to an open baffle speaker are:
- No box - the output from the rear of the driver is used constructively, rather than attempt to absorb it's energy inside a box; this results in reduced box coloration and an improved interaction with the room.
- Dipole radiation - increased rear radiation and reduced side, floor and ceiling reflections
- Velocity source bass interacts differently with room modes.
In changing between a box speaker and open baffle, I attempted to remove anything that might sway me either way. I used the same drivers and crossover, and used EQ to get the same frequency response.
How do they sound?
The biggest difference is in the sound stage and imaging. The sound stage is bigger - wider, deeper and more immersive. In switching from open baffle to dipole, the sound stage seems immediately small. I find the sound stage of open baffles to be more realistic and natural. I also find that the imaging is more stable. It's less likely that you will need a centre speaker (and more difficult to include if you do want one). The sweet spot is bigger and as you move sideways across the room, the sound stage is less inclined to collapse to one side.
The downside is that the imaging is not as tight and focused. When sitting in the sweet spot and where everything is just right, and you are listening to a recording where the imaging isn't confused, box speakers have the potential to create a tighter image where it's easier to pin down where the sound is coming from. Open baffle speakers seem to stretch the sound stage. It's like comparing a projector to an LCD TV. You can't expect the projected image to be equally sharp. It's a matter of personal preference. Which do you prefer? It's a decision you can't make without experiencing both options.
One of the more subtle qualities of an open baffle speaker is the absence of box coloration. High end speaker manufacturers work very hard to try to remove the impact of the rear wave, but it is never completely removed. As a result, budget drivers on an open baffle can compare very well to much more expensive drivers in a box.
Bass - monopole vs dipole
Open baffle purists insist on dipole bass as well. This is where one has to balance cost, size and output. To match the output of a conventional sub or woofer an open baffle needs to move much more air. Open baffle bass has a number of advantages:
- No box coloration
- Different room interaction
There are two schools of thought regarding the room interaction of dipoles with the room. One is in favour of dipoles as a means of minimising the excitation of room modes. This view has been made popular by Siegfried Linkwitz. A different view is promoted by Dr Earl Geddes, in which it is considered more desirable to deliberately excite as many room modes as possible with monopole subwoofers in order to achieve a smooth in room bass response. My own preference is for the Geddes approach.
My initial listening tests showed open baffle bass can be more accurate with conventional subs. However, once I obtained some Rythmik Servo subs, I discovered this isn't always the case. I set up two Rythmik subs - one a monopole and the other in a H frame. Both were calibrated so that their output and response were identical. Then listening to one of my familiar reference tracks with acoustic double bass, I compared, switching from one to the other until. Which one was more accurate? They were so close, it's hard to say any differences were real or imaginary.
I believe there is a reason the Rythmik servo subs didn't reveal a difference. The servo provides extra damping which vastly reduces the impact of the rear sound wave. Box coloration is reduced to the point that you can't tell if you are listening to a monopole or open baffle. This is why I now run my subs as monopoles. There is no reason in this case to give up output.
Navigating complexity and performance
Open baffle speakers are normally considered expensive, complex and large yet with serious output limitations. This is not necessarily true, depending on certain choices.
My open baffle speakers are relatively simple and have high output without being large. The main panels have dipole roll off below about 300 Hz. This means eq is needed between this point and 80 Hz where the main panels cross to the subs. Midbass drivers can easily handle this eq and in real use I find the system sounds stressed before running out of excursion. The efficiency in the midrange is actually quite high at around 94 db 1w1m. Where high output is more important, an efficient woofer could be used as a monopole crossing where dipole roll off occurs.
There is a lot more that can be said about open baffle loudspeakers. If you'd like to read more, then I recommend the following websites:
Linkwitz Lab - the best online source of information on open baffle loudspeakers
Music and Design
August 15, 2009
If you are looking for a high end transmission line kit, I recommend LSK TL6. (Not currently available, but soon to be redesigned from the ground up.)
Why choose this design over JV60 + TL box?
JV60 is a good value kit that with some mods can sound very good indeed - it is a smooth laid back sounding speaker that sounds even better with a good quality TL (transmission line) box. The crossover parts are cheapies and that is addressed with my assembled crossover upgrades. However, there are better drivers available and better design choices. The stock JV60 was designed to avoid failure when driven hard. One can remove the polyswitch protectors but the crossover point is a little high in order to optimise power handling over off axis response. Where very high output is desired, there are far better choices.
TL6 is about to be redesigned from the ground up. It will soon be a completely new design. New drivers. New crossover. New box.
After selling my modified JV60 kit in a TL box, this is what the new owner had to say:
The Etude TL is a 2.5 way MTM floorstander with extension below 25 Hz in typical listening rooms. The bass is articulate and the overall sound is neutral with a slight mellow bias in the voicing. This speaker has been built by DIY enthusiasts all around the world.
If you have built the TL, please share some photos.
In choosing a speaker to build, it's worthwhile to understand a little about your preferences. Some speakers are ruthlessly accurate and highly resolving. They will expose a poor recording without holding anything back and for many this means not being able to enjoy many albums. I have found the Linkwitz Orion to be in that category as well as Vaf Signature speakers. The Etude TL is a different type of speaker. It does not expose detail to the same extent and tends to be a little more forgiving with less than perfect albums. The sound is neutral and natural.
This project is based on the JV60 kit which previously was sold in Australia through Jaycar. The kit included a pre-assembled crossover, drivers and damping material. The crossover shown here is a modified version in which the tweeter level has been adjusted and unnecessary parts have been removed. If you have this kit then some crossover upgrades are suggested.
What is a transmission line speaker?
A transmission line (TL) speaker loads the woofer into a tapered tube which is filled with acoustic damping material. The line tunes the woofers to extend the bass lower. It also absorbs the midrange output from the rear of the driver and thereby minimises box coloration.
In effect, a TL combines the best of sealed and vented designs. A sealed enclosure allows more acoustic damping to reduce box coloration but allows no bass reinforcement. A vented enclosure doesn't allow sufficient damping to remove box coloration to the same degree. The result in either case is a compromise that isn't required in a TL speaker.
More info on wikipedia:
Wikipedia - TL
What's the catch?
The most obvious downside is enclosure size. A TL will normally be larger than a sealed or vented enclosure. However, if the line is folded, the size is still acceptable. Another issue is coming up with a design. It isn't quite as simple as a simple vented box.
The drivers are well regarded Vifa classic units - Vifa D25AG a 1" aluminium dome and the Vifa P17 6.5" midbass. They are now discontinued, but currently still available in Australia.
How does it sound?
With the tweeter pad, the sound is smooth, open and relaxed. The midrange is natural and uncoloured and the bass is articulate and deep. Acoustic double bass sounds very natural. The bass tends to be natural rather than punchy, although this also depends on the room. It is perhaps slightly under-stated.
The Vifa D25 aluminium dome is one of my favourite tweeters. It has a nice smooth extended response and no hint of the harshness that some associate with metal domes. The Vifa P17 is a very well behaved driver and the box helps to keep coloration low.
Different music styles
I've always felt this speaker shines the most with vocal jazz music. Acoustic double bass is very natural on these. Norah Jones, Harry Connick Jr, Dianna Krall, and Natlie Cole among others stand out. Of course, any decent speaker will handle all types of music, but if your music preferences are more towards rock then this might not be the best choice. Where bass authority is important I suggest adding subs, but for dynamics and raw energy and high output you probably shouldn't be looking at a conventional floorstander such as these. You might instead consider an Ewave.
Placement is an important to get the best result. Like every floorstander, this one requires breathing space. Bring them out from the front and side walls and toe them in so that they face the main listening chair. Ideally their offset from the front and side walls should be different to avoid boundary effects. At least 1m clearance is recommended. If you must place them closer then acoustic treatment should be used.
When placed correctly, the sound image should float out in the middle of the room, extend behind the speakers and sometimes project out the sides as well. With your eyes shut there should be no sense that the sound is coming from two speakers.
Use with subwoofers
Where more bass force and impact is desired, a sub is recommended. Be careful with your choice, to ensure that your sub is up to the task. You will also need to put in some effort at integration. It is much harder to get good bass from a sub because there are many more ways it can go wrong.
More about bass integration
I have used Rythmik servo subs with my TLs and recommend them highly. When integrated well they extend the bass deeper and the bass improves further.
Home Theatre use
Ideally for home theatre I would suggest a speaker with greater efficiency such as the Ewave. The higher efficiency is likely to mean spending less on amplification. Hence instead of adding power amps to your receiver, you may instead use more efficient speakers. However, the TLs are as good as any floorstander for home theatre and the crossover is ideally suited to building matching surrounds.
New to DIY?
Few people realise how much value you can add to your system by DIY. The most critical part of your system is the speakers, and this is where you can add the most value with DIY. What makes DIY so exciting? It's simply this - if you make wise choices, you can create a system that will beat one you couldn't otherwise afford.
Above: Current home theatre setup with diy omni speakers and Rythmik servo subs
Rythmik Servo subwoofers
Articles - coming soonDIY - creating the system you think you can't afford
Yes, it's possible to create that system you thought you couldn't afford. It's easier than you think. No excuses!Majoring on the big issues
It makes sense to put the most effort where you get the most return, but common sense isn't so common in audio! Here are my suggestions about what is most important.
Open baffle speakers
Uncovering the magic of one of the most under-rated speakers in mainstream audio.
Getting the bass right
An approach that works well for music and movies.