What do I mean by modular headphones?
My goal from the very start of this project was to design headphones that would offer me modding possibilities that most other headphones on the market didn't, and that would give me the ability to change their functional and visual characteristics whenever I felt like it, in a completely reversible and reasonably affordable way, without necessarily having to replace an entire headphone every single time I got bored of it or wanted to try something different.
To simplify my definition of modular headphones:
I knew that if I wanted this idea to be plausible, I had to do everything I could in order to make sure the headphones would be very DIY-friendly and easy to work on, completely modular, and offer a wide range of possibilities. I've tried to apply these principles to every headphone I've designed as much as I could, and will do the same or even more when it comes to future models. The principles that I believe define what a modular headphone is, can be summed up into four points:
To simplify my definition of modular headphones:
- They can accept several drivers, with an ability to adapt the headphone to each one
- They offer an ability to manipulate the sound balance/signature while using a single driver
- They can be customized in terms of aesthetics
- Everything done to the headphones has to be reversible
I knew that if I wanted this idea to be plausible, I had to do everything I could in order to make sure the headphones would be very DIY-friendly and easy to work on, completely modular, and offer a wide range of possibilities. I've tried to apply these principles to every headphone I've designed as much as I could, and will do the same or even more when it comes to future models. The principles that I believe define what a modular headphone is, can be summed up into four points:
- Every single part of each headphone has to be removable, replaceable and upgradeable with basic tools, without any damage done to the parts that are being replaced, or to the rest of the headphone, with each process being fully reversible. You should never be forced to replace an entire headphone just because one part of them is damaged, faulty, or you've simply decided to change it or upgrade it.
- Every part has to be designed in a way that allows its production via multiple manufacturing methods, and therefore multiple materials, from affordable to expensive. This in turn allows for the creation of a variety of headphone configurations at multiple performance and quality levels, all using a single basic design platform for each individual headphone model.
- The entire design concept of each headphone platform needs to be modular to such an extent that it enables several select types of drivers to be used with it, while featuring enough functional tuning options to actually allow the headphone to be adapted to those drivers, and offer high levels of performance with them. In addition, the modular features need to allow me to modify the sound of the headphone while keeping the same driver, by manipulating the default sound signature and balance.
- When switching between drivers, or keeping the same driver and playing with sound tuning options, as many parts of a headphone as possible should stay untouched, in order to make the modding process as affordable and simple as possible. Practically speaking, this means that each headphone platform features two groups of parts. The first group are the parts that represent the mechanical and structural elements of a headphone. They have little to do with the sound (to a certain extent) and should therefore be designed in a way that allows them to be reusable and stay the same regardless of the functional configuration of the headphone. Their job is to be as strong, durable, rigid and inert as possible. These are parts such as the headband pieces, adjustment sliders, hinge yokes, fixed parts of the ear cup, fixed elements of the driver baffle, etc. They are still individual parts, and can be replaced independently, but don't necessarily have to be replaced when upgrading the drivers or manipulating the sound signature. The second group are the functional parts, which have a direct influence on the sound of a headphone, and should therefore be designed in a way that allows them to be replaced or modified as required, depending on the desired results, and depending on the requirements of each specific driver. These are parts such as the removable filter frames, acoustic mesh filters, driver holders, ear cup grills, various acoustic treatment elements, etc. They are smaller, lighter, simpler and far less costly to produce than those in the first group, but have a massive effect on the sound of a headphone. As I've said at the beginning, this makes modifying headphones (once the basic structural parts are assembled) very efficient compared to having to build entire new ear cups or even an entire new headphone in order to achieve a different sound.
Wouldn't modular headphones be inferior by default?
A lot of people think, due to the fact that an idea like this hasn't been fully explored in the past, that making a modular headphone that could work with various drivers and/or ear pads is impossible, or would simply never yield positive results.
Before going into why I believe it is possible, I'd like to address the argument that something modular, designed to work in multiple configurations and with a lot of variables at play, is generally supposed to be inferior to something that's designed specifically for a single purpose and to work in a single configuration, with all the parts finely tuned to work together. That argument is correct. It is without a doubt a fact that developing a headphone (or any product) from ground up, and fine tuning every single aspect of it, should lead to the best possible results, at least in theory. However, it's not always the case in practice, especially when it comes to products like headphones, which are often highly subjective. It's actually virtually never the case, hence the fact that pretty much no headphone ever created, at any price point, has been perfect and exactly matching the personal preferences of everyone out there, which is why there are large differences in performance levels, with a lot of different sound signatures between various headphones, even between those that are sold at the same price points.
What I'm trying to say is that developing a traditional headphone with a fixed single configuration, doesn't lead to excellent performance and everyone liking it by default. At the same time, if properly designed, there is no reason why a headphone that offers modular aspects and a possibility to replace or upgrade its individual components would have to be inferior by default. The point being that all of the characteristics that make certain single configuration headphones great are applicable to modular headphones too, but with the modular headphones having an added bonus of certain design solutions that aren't present on fixed configuration headphones. Therefore, if done properly, a modular headphone design can potentially allow for the development of multiple great configurations based around a single platform, with each having all the characteristics of a great headphone, such as the use of high quality materials in its production, well designed passive headphone parts, high quality headphone drivers, a high level of attention to detail and great build quality, all while allowing for personalization of subjective aspects, such as the sound tuning in each of these hypothetical configurations.
Before going into why I believe it is possible, I'd like to address the argument that something modular, designed to work in multiple configurations and with a lot of variables at play, is generally supposed to be inferior to something that's designed specifically for a single purpose and to work in a single configuration, with all the parts finely tuned to work together. That argument is correct. It is without a doubt a fact that developing a headphone (or any product) from ground up, and fine tuning every single aspect of it, should lead to the best possible results, at least in theory. However, it's not always the case in practice, especially when it comes to products like headphones, which are often highly subjective. It's actually virtually never the case, hence the fact that pretty much no headphone ever created, at any price point, has been perfect and exactly matching the personal preferences of everyone out there, which is why there are large differences in performance levels, with a lot of different sound signatures between various headphones, even between those that are sold at the same price points.
What I'm trying to say is that developing a traditional headphone with a fixed single configuration, doesn't lead to excellent performance and everyone liking it by default. At the same time, if properly designed, there is no reason why a headphone that offers modular aspects and a possibility to replace or upgrade its individual components would have to be inferior by default. The point being that all of the characteristics that make certain single configuration headphones great are applicable to modular headphones too, but with the modular headphones having an added bonus of certain design solutions that aren't present on fixed configuration headphones. Therefore, if done properly, a modular headphone design can potentially allow for the development of multiple great configurations based around a single platform, with each having all the characteristics of a great headphone, such as the use of high quality materials in its production, well designed passive headphone parts, high quality headphone drivers, a high level of attention to detail and great build quality, all while allowing for personalization of subjective aspects, such as the sound tuning in each of these hypothetical configurations.
What makes the concept of modular headphones possible?
Making any headphone platform adaptable to various drivers, and allow for multiple sound tuning options with a single driver, required a combination of some non-traditional and novel design choices, combined with some well tested principles that many high quality dynamic headphones feature.
Based on my experience with playing with various designs, I've found that there are a couple of crucial elements of an open-back (or semi-open) headphone design that the "sound tuning" most depends on. This was the first step, and allowed me to then start working on ideas of how to make those elements easily controllable and customizable, while keeping most headphone components the same, in order for that headphone to be considered fully modular. Again, this is based on my experience, so take it with a grain of salt. While the methods I'll describe have led to consistent and predictable results with all the headphone designs I've played around with, that doesn't guarantee that the same would be true with absolutely all other open-back designs. It probably would be, and it makes sense that it would be, but I'm not gonna claim that.
To simplify it as much as possible, I'd like to divide those elements into five categories:
1) The raw performance of the driver:
Obviously, this means that starting off with a driver that has good performance in isolation is very important, and the more linear the response of the driver is, with as few localized peaks or dips in the response as possible, the easier it is to manipulate its base sound character in a predictable way. Not every driver can be used for this purpose or is DIY-friendly, but the amount of them available that are is quite large. Picking a driver that is suitable for a certain type of headphone design is obviously extremely important too. A driver developed for use in a fully sealed closed-back headphone, will never work properly in an open-back headphone, regardless of the amount of tuning done to it. Some drivers can be made to work "equally well" in closed and opened headphones, but they're quite rare. In addition, if the concept of modular headphones becomes more popular, it's going to attract more and more companies that can develop and produce high quality drivers to offer their products on the market. This already happened when Grado-like DIY headphones became popular, and has spawned many manufacturers of aftermarket drivers, some of which outperform even Grado's own top of the line drivers.
2) Controlling the position of the driver:
All of the my headphone designs feature some method of attaching the drivers at various distances to the listeners ear, most of them via driver holder parts, which are easily removable, cheap to produce, easy to design in various versions, and are usually designed to have multiple purposes, on top of just holding the driver in place securely. How far away from the ear the driver is can have pretty noticeable subjective effects, and probably easily measurable with the right measurement gear. In my experience, pushing the driver closer or farther by a range of 2 or 3 mm doesn't create major changes in the frequency response, but it does have certain subjective effects, making the sound "more forward" or "aggressive" the closer the driver is, and vice versa. Having an ability to adjust the position of the driver also helps when dealing with other variables, such as the shape of listeners ears, the type of ear pads used, etc. The angle and the forward offset of the driver also leads to certain effects, but the angles are fixed on current designs, in positions that I found best worked for me through testing. In the future model designs, I plan to make the driver positions and angles adjustable "on the fly", without removing them from the headphones, or having to replace any components. This is technically more difficult to pull off, but I have a pretty good idea of how to do it.
3) Controlling the venting of the enclosed ear pad volume:
One of the biggest aspects of being able to adapt various drivers to a single base headphone design, or manipulate the "voicing" of a headphone while keeping the same driver, is the ability to control the venting, or the amount of leakage out of the enclosed ear pad volume. This is crucial for controlling the low to mid frequency performance of an open-back headphone, parts of the audible range that obviously have to be well tuned in order for any headphone to sound natural, as these areas of the frequency range contain most of the major information of anything we hear, regardless of whether we're listening to it in person, or through headphones. If this area is off, without an even and smooth frequency response, vocals won't sound good, instruments won't sound good, and at that point you might as well throw the headphone into the trash. For that reason, finding a way to control the area up to around 2000 Hz has been extremely important to me. I've managed to accomplish that with removable components that I call filter frames, which are at the core of all full-sized headphone designs I've worked on so far, and are one of the biggest reasons why this idea of modular, DIY-friendly headphones is actually possible in practice, instead of just being a nice sounding idea on paper.
Now, a little theory behind this idea first, based on my experience with various designs and quite a lot of experimenting.
Remember, I'm not claiming that any of this is solid science, it's still work in progress, and these are just my subjective and objective observations, if you can call my primitive measurements objective. As I've already mentioned, I found that the low to mid frequency response of any semi or fully opened headphone depends most of all (but not only) on the effects that the entire system of passive ear cup parts has on the amount of leakage, or venting out of the enclosed ear pad volume (let's shorten that to EEPV). To fully simplify this, we can say that the more leakage is possible, or the more "opened" the EEPV is, the lower the SPL up to ~2000 Hz will be, and vice versa. In addition to that, the actual physical volume, or the amount of air enclosed by the ear pad, as well as the shape of the EEPV, can have multiple effects, but more on that later. Now, controlling the amount of leakage, on any open-backed headphone, is usually done with a combination of two things:
All of this means that controlling the low to mid frequency performance of the headphone is best done by relying as much as possible on a combination of the baffle, the ports and the filters, rather than ear pads alone. This meant that the baffles that I had to design had to be as opened as possible, but also as strong and rigid as possible. Luckily, SLS and MJF additive manufacturing technologies are a perfect solution for that, because the materials they use for production are extremely strong and rigid, and there's no limitations to how complex the designs of the parts can be, so I had full freedom in that regard.
But that only solves half of the problem, because replacing acoustic mesh filters on such headphones, including the ones I mentioned above, is impossible without permanent damage, so using that method for modifying the response of an already built headphone is not very practical. This meant that I had to design a headphone in such a way that replacing filters would be quick and simple, would lead to no damage done to either the filters or the rest of the headphone, and would be entirely reversible, without introducing any downsides to the performance or the rigidity of the entire assembly. The best solution I could come up with for that are multi-layered baffle designs with removable filter frames.
Since the filter frames are easily removable, and very cheap to produce, costing a couple of dollars per unit to make, they can be replaced in order to change the sound of a headphone to a larger or lesser degree, depending on the design of the filter frame, and the acoustic mesh used with it. In short, the more transparent the mesh is, the lower the SPL up to ~2000 Hz is, combined with a very small boost in the 2000-4000 Hz region, and vice versa. Steps in acoustic impedance levels between various types of mesh filters are very fine, so it's easy to find just the right one. On top of it, baffles have multiple ports, so it's possible to use a combination of various mesh types at the same time, leading to even finer degrees of sound tuning. In my experience, the effects are very linear and predictable, the shape of the FR curve in that region doesn't change when replacing one mesh with another, and there are no peaks or dips introduced. The only things that change are the tilt and the level of the FR curve, with the effects becoming gradually more pronounced towards the lower frequencies. It's basically like a physical equalizer. This is the case regardless of whether only the filter frame is being changed, or the driver is being replaced as well. So, the method of replacing filter frames is not only useful for adapting a base headphone design to a different driver, but for customizing the sound of a headphone without changing the driver, enabling a transition to a different sound balance by simply swapping the filter frames, and nothing else. It's like having the HD650's, and then after spending 5 minutes to replace a few cheap parts, ending up with the HD600's, with many more degrees of tuning possible in either direction of the sound signature spectrum, and without having to buy another pair of headphones.
Alternatively, I've tested a method of using a certain amount of smaller ports on the baffle, opened or closed with a sliding mechanism, however, this allows for a far smaller range of adjustment, and takes away space that could be otherwise used, ultimately making the headphone more complicated with limited benefits. However, it is an idea I'm exploring and could possibly combine with removable filter frames in certain future designs.
How additive manufacturing helps in this case?
Now, using a multi-layered baffle is only possible if the fit between the filter frame part and the main baffle part is extremely tight, eliminating any possible rattling, noise or any unwanted structural losses in general. Luckily, the dimensional accuracy of SLS or MJF 3D printed parts is extremely high, enabling very tight assembly tolerances or gaps between parts, which means they fit together very tightly, but can still be separated without any damage. On top of that, all the polyamides used by these AM technologies are very "silent", they don't rattle and they don't squeak like ABS plastic can, for example, because of their slightly textured surface. Once the parts are assembled and all of the screws tightened, filter frames and baffles essentially act as a single solid part, with a very solid feel. In fact, without knowing about it, it would be very hard to tell the parts are layered.
4) Controlling the shape and size of the enclosed ear pad volume:
Since the filter frames are now removable, and since they essentially represent the "upper layer" of the baffle, so they define the size and shape of the EEPV to a certain extent, another modification becomes possible. As I've mentioned previously, changing the volume enclosed by the ear pad has an effect on the low frequency performance. Again, in my experience, the larger the volume is, the higher up the frequency range the bass peak is pushed, and vice versa. This means that changing the designs of the filter frames in order to make them thicker, or thinner, is a simple way of either increasing or decreasing the EEPV, and therefore control the peak frequency of the bass response. In addition, since the shape of the filter frame/baffle has an effect on the sound as well, primirily on the 3000-5000 Hz region in my experience, changing the design of the filter frame part, or modifying how "concave" or pronounced the "bowl shape" of the baffle is, enables a bit of control over the frequency response in the highs. However, the higher in the frequency range you go, the more difficult it becomes to control the sound, and the more you depend on the characteristics of each individual driver and the peaks or dips introduced by specific ear pads, earlobe shapes, etc.
5) Controlling the acoustic chamber behind the driver:
The last major aspect of tuning a headphone is dealing with what happens behind the driver, All three of the headphone designs that I'll present, despite all being fundamentally different in terms of design, offer an ability to control the rear acoustic chamber in four main ways, by allowing me to:
So what does this mean?
It means that it's entirely possible to have a single base headphone design that can work with several different drivers, which require slightly different environments in order to function optimally, because now all of the crucial sound tuning elements can be modified while keeping the majority of the headphone components untouched.
Because of this, it's possible to not only fine tune the sound of a headphone, and achieve different sound signatures when using a single driver, but even after building one configuration of a headphone, it's possible to keep most of the "expensive headphone parts" the same, put an entirely different driver in it, and still have the ability to adapt that headphone to function properly with that new driver, by simply changing the "cheap headphone parts". This is great because it's not only fully reversible, but is also far more efficient and affordable than changing the entire main baffle, or the entire headphone. The best part is, the amount of headphone drivers available is enormous. Even if we just limit our choices to the "Grado mod drivers", the 46 mm diameter ones, there's multiple high quality choices on the market. Expanding to the entire aftermarket driver industry, there are literally hundreds of choices at various price points. And at the end, there are the OEM drivers, since pretty much every major headphone company sells their own proprietary drivers as spare parts, so it's possible to build custom headphones using them as well.
It also leaves us with some interesting implications regarding the potential life-times of such headphones, the possibilities to gradually change or upgrade them in the future, replace parts with new updated and improved versions, replace them with parts produced out of higher quality materials, maybe made with some newly developed additive manufacturing technology that's yet to become available. The possibilities are literally endless. Of course, theoretically, it still takes a lot of work to develop a modular design that actually works in practice, and it's a far more difficult task than developing a single-configuration headphone. But, as I've said at the start, my idea has been to design fully modular, DIY-friendly and future-proof headphone platforms that could serve me well for years, and I think this comes quite close to what I've had in mind. Does it require more work than simply going into a store and buying a headphone? Sure. But it's very fun to play around with, and that's what headphone DIY should be all about. Especially when it's possible to end up with great results, while potentially saving money.
Based on my experience with playing with various designs, I've found that there are a couple of crucial elements of an open-back (or semi-open) headphone design that the "sound tuning" most depends on. This was the first step, and allowed me to then start working on ideas of how to make those elements easily controllable and customizable, while keeping most headphone components the same, in order for that headphone to be considered fully modular. Again, this is based on my experience, so take it with a grain of salt. While the methods I'll describe have led to consistent and predictable results with all the headphone designs I've played around with, that doesn't guarantee that the same would be true with absolutely all other open-back designs. It probably would be, and it makes sense that it would be, but I'm not gonna claim that.
To simplify it as much as possible, I'd like to divide those elements into five categories:
1) The raw performance of the driver:
Obviously, this means that starting off with a driver that has good performance in isolation is very important, and the more linear the response of the driver is, with as few localized peaks or dips in the response as possible, the easier it is to manipulate its base sound character in a predictable way. Not every driver can be used for this purpose or is DIY-friendly, but the amount of them available that are is quite large. Picking a driver that is suitable for a certain type of headphone design is obviously extremely important too. A driver developed for use in a fully sealed closed-back headphone, will never work properly in an open-back headphone, regardless of the amount of tuning done to it. Some drivers can be made to work "equally well" in closed and opened headphones, but they're quite rare. In addition, if the concept of modular headphones becomes more popular, it's going to attract more and more companies that can develop and produce high quality drivers to offer their products on the market. This already happened when Grado-like DIY headphones became popular, and has spawned many manufacturers of aftermarket drivers, some of which outperform even Grado's own top of the line drivers.
2) Controlling the position of the driver:
All of the my headphone designs feature some method of attaching the drivers at various distances to the listeners ear, most of them via driver holder parts, which are easily removable, cheap to produce, easy to design in various versions, and are usually designed to have multiple purposes, on top of just holding the driver in place securely. How far away from the ear the driver is can have pretty noticeable subjective effects, and probably easily measurable with the right measurement gear. In my experience, pushing the driver closer or farther by a range of 2 or 3 mm doesn't create major changes in the frequency response, but it does have certain subjective effects, making the sound "more forward" or "aggressive" the closer the driver is, and vice versa. Having an ability to adjust the position of the driver also helps when dealing with other variables, such as the shape of listeners ears, the type of ear pads used, etc. The angle and the forward offset of the driver also leads to certain effects, but the angles are fixed on current designs, in positions that I found best worked for me through testing. In the future model designs, I plan to make the driver positions and angles adjustable "on the fly", without removing them from the headphones, or having to replace any components. This is technically more difficult to pull off, but I have a pretty good idea of how to do it.
3) Controlling the venting of the enclosed ear pad volume:
One of the biggest aspects of being able to adapt various drivers to a single base headphone design, or manipulate the "voicing" of a headphone while keeping the same driver, is the ability to control the venting, or the amount of leakage out of the enclosed ear pad volume. This is crucial for controlling the low to mid frequency performance of an open-back headphone, parts of the audible range that obviously have to be well tuned in order for any headphone to sound natural, as these areas of the frequency range contain most of the major information of anything we hear, regardless of whether we're listening to it in person, or through headphones. If this area is off, without an even and smooth frequency response, vocals won't sound good, instruments won't sound good, and at that point you might as well throw the headphone into the trash. For that reason, finding a way to control the area up to around 2000 Hz has been extremely important to me. I've managed to accomplish that with removable components that I call filter frames, which are at the core of all full-sized headphone designs I've worked on so far, and are one of the biggest reasons why this idea of modular, DIY-friendly headphones is actually possible in practice, instead of just being a nice sounding idea on paper.
Now, a little theory behind this idea first, based on my experience with various designs and quite a lot of experimenting.
Remember, I'm not claiming that any of this is solid science, it's still work in progress, and these are just my subjective and objective observations, if you can call my primitive measurements objective. As I've already mentioned, I found that the low to mid frequency response of any semi or fully opened headphone depends most of all (but not only) on the effects that the entire system of passive ear cup parts has on the amount of leakage, or venting out of the enclosed ear pad volume (let's shorten that to EEPV). To fully simplify this, we can say that the more leakage is possible, or the more "opened" the EEPV is, the lower the SPL up to ~2000 Hz will be, and vice versa. In addition to that, the actual physical volume, or the amount of air enclosed by the ear pad, as well as the shape of the EEPV, can have multiple effects, but more on that later. Now, controlling the amount of leakage, on any open-backed headphone, is usually done with a combination of two things:
- Ear pads – To simplify it as much as possible, the less transparent the ear pads are (eg. Leather), the less leakage they allow out of the EEPV, and therefore generally lead to higher SPL in the low to mid frequency region. More transparent ear pads (eg. Velour) will allow for more leakage, and will lead to a lower SPL in this region, if all other variables remain the same. However, ear pads usually have various other effects on the sound as well, throughout the audible range, and these effects usually aren't linear, leading to various dips or peaks in the reponse, sometimes quite nasty ones too. Which material is used for the ear pad, how much it absorbs or reflects certain sound freqencies, how perforated it is, how large the perforations are, how dense the foam inside of it is, how deep the ear pad is, and how large the ear pad opening is, are some of the variables that all add their own little contribution to the sound. For that reason, relying on ear pads alone to tailor the sound of a modular headphone is more or less a game of trial and error, and will rarely lead to desired results. When making a single configuration headphone that's not meant to be modified in the future and that will only use a single type of driver, designing an ear pad to tailor the sound in a specific way is possible, but we're talking about modular headphones here. So relying on ear pads alone is impossible.
- Baffle ports – The other way of controlling the leakage out of the EEPV are the ports on the baffle surrounding the driver, which vary in shapes and sizes from headphone to headphone, and are covered with filters that have a certain acoustic impedance rating. Paper filters are usually used on most headphones, because they're cheap and can do the job in most cases, especially when the ports are fairly small relative to the overall area of the baffle. The better way of doing it, and this is how most dynamic high-end headphones do it (HD800, HD700, T1, K812, all Focals, etc.) is to make the baffle ports as large as possible, and then cover them with more advanced, but more expensive, metal or polyester acoustic mesh filters. Acoustic mesh filters are specifically designed for use with headphones, have very linear effects on the frequency response, and are available in a wide range of acoustic impedance ratings, from very transparent, to very dense, making it very easy to find the right filter for any application. I've been using such filters, provided by SAATI as test samples. Now, making the openings on the baffle very large, also means that the design of the baffle has to be quite complex, with a lot of structurally beneficial design elements and bracing, combined with high quality materials, in order to ensure that the baffle is rigid and inert enough, keeping the impact it has on the sound of a headphone to a minimum, and allowing the driver itself to shine. At the same time, the baffle itself can't be too heavy. It would be easy to make a perfectly inert baffle by just laser cutting a 1 cm thick piece of steel and drill some venting holes in it, but the headphones would be unbearably heavy.
In addition, this design principle has certain other benefits as well. Logically, the larger the ports on the baffle are, the less solid plastic/metal areas there are to reflect the sound back. Keeping the baffle as opened as possible also leads to a subjective sensation of a more opened sound because it's possible to hear the surroundings, and there's less heat build-up as well, leading to more comfortable listening sessions.
All of this means that controlling the low to mid frequency performance of the headphone is best done by relying as much as possible on a combination of the baffle, the ports and the filters, rather than ear pads alone. This meant that the baffles that I had to design had to be as opened as possible, but also as strong and rigid as possible. Luckily, SLS and MJF additive manufacturing technologies are a perfect solution for that, because the materials they use for production are extremely strong and rigid, and there's no limitations to how complex the designs of the parts can be, so I had full freedom in that regard.
But that only solves half of the problem, because replacing acoustic mesh filters on such headphones, including the ones I mentioned above, is impossible without permanent damage, so using that method for modifying the response of an already built headphone is not very practical. This meant that I had to design a headphone in such a way that replacing filters would be quick and simple, would lead to no damage done to either the filters or the rest of the headphone, and would be entirely reversible, without introducing any downsides to the performance or the rigidity of the entire assembly. The best solution I could come up with for that are multi-layered baffle designs with removable filter frames.
Since the filter frames are easily removable, and very cheap to produce, costing a couple of dollars per unit to make, they can be replaced in order to change the sound of a headphone to a larger or lesser degree, depending on the design of the filter frame, and the acoustic mesh used with it. In short, the more transparent the mesh is, the lower the SPL up to ~2000 Hz is, combined with a very small boost in the 2000-4000 Hz region, and vice versa. Steps in acoustic impedance levels between various types of mesh filters are very fine, so it's easy to find just the right one. On top of it, baffles have multiple ports, so it's possible to use a combination of various mesh types at the same time, leading to even finer degrees of sound tuning. In my experience, the effects are very linear and predictable, the shape of the FR curve in that region doesn't change when replacing one mesh with another, and there are no peaks or dips introduced. The only things that change are the tilt and the level of the FR curve, with the effects becoming gradually more pronounced towards the lower frequencies. It's basically like a physical equalizer. This is the case regardless of whether only the filter frame is being changed, or the driver is being replaced as well. So, the method of replacing filter frames is not only useful for adapting a base headphone design to a different driver, but for customizing the sound of a headphone without changing the driver, enabling a transition to a different sound balance by simply swapping the filter frames, and nothing else. It's like having the HD650's, and then after spending 5 minutes to replace a few cheap parts, ending up with the HD600's, with many more degrees of tuning possible in either direction of the sound signature spectrum, and without having to buy another pair of headphones.
Alternatively, I've tested a method of using a certain amount of smaller ports on the baffle, opened or closed with a sliding mechanism, however, this allows for a far smaller range of adjustment, and takes away space that could be otherwise used, ultimately making the headphone more complicated with limited benefits. However, it is an idea I'm exploring and could possibly combine with removable filter frames in certain future designs.
How additive manufacturing helps in this case?
Now, using a multi-layered baffle is only possible if the fit between the filter frame part and the main baffle part is extremely tight, eliminating any possible rattling, noise or any unwanted structural losses in general. Luckily, the dimensional accuracy of SLS or MJF 3D printed parts is extremely high, enabling very tight assembly tolerances or gaps between parts, which means they fit together very tightly, but can still be separated without any damage. On top of that, all the polyamides used by these AM technologies are very "silent", they don't rattle and they don't squeak like ABS plastic can, for example, because of their slightly textured surface. Once the parts are assembled and all of the screws tightened, filter frames and baffles essentially act as a single solid part, with a very solid feel. In fact, without knowing about it, it would be very hard to tell the parts are layered.
4) Controlling the shape and size of the enclosed ear pad volume:
Since the filter frames are now removable, and since they essentially represent the "upper layer" of the baffle, so they define the size and shape of the EEPV to a certain extent, another modification becomes possible. As I've mentioned previously, changing the volume enclosed by the ear pad has an effect on the low frequency performance. Again, in my experience, the larger the volume is, the higher up the frequency range the bass peak is pushed, and vice versa. This means that changing the designs of the filter frames in order to make them thicker, or thinner, is a simple way of either increasing or decreasing the EEPV, and therefore control the peak frequency of the bass response. In addition, since the shape of the filter frame/baffle has an effect on the sound as well, primirily on the 3000-5000 Hz region in my experience, changing the design of the filter frame part, or modifying how "concave" or pronounced the "bowl shape" of the baffle is, enables a bit of control over the frequency response in the highs. However, the higher in the frequency range you go, the more difficult it becomes to control the sound, and the more you depend on the characteristics of each individual driver and the peaks or dips introduced by specific ear pads, earlobe shapes, etc.
5) Controlling the acoustic chamber behind the driver:
The last major aspect of tuning a headphone is dealing with what happens behind the driver, All three of the headphone designs that I'll present, despite all being fundamentally different in terms of design, offer an ability to control the rear acoustic chamber in four main ways, by allowing me to:
- Control the volume/shape of the chamber – This is possible by utilizing various custom designs of the chamber walls, including 3D printed adapters that can be inserted into the default chambers, or by lining the walls of the chambers with materials of varying degrees of thickness.
- Control the reflective/absorbtive properties of the walls of the chamber – The methods mentioned in the previous point can be used for this purpose as well. Lining the walls of the chamber with a certain type of felt, or foam, will produce different results than leaving the walls exposed, with a smooth plastic surface. Various irregular patterns can be designed and featured on the 3D printed inserts too.
- Control the degree to which the chamber is opened at the end opposite to the driver – This aspect can be controlled by using various filters, felts, foams or fabrics, in combination with grills of different designs and various degrees of openness covering the ear cup holes. Of the four points mentioned here, this one has the most dramatic effects in most cases, allowing to entirely transform (and sometimes ruin) the sound of a headphone just by adjusting this single variable. Anyone who has used an open-back headphone like the K701's and covered the grills with their hands, knows how big the effects on the sound can be. Using less opened grills doesn't just "choke" the amount of airflow available to the driver, but it can also lead to more reflections from whatever is blocking the opening. For that reason, I generally like to keep the rear chambers as opened as possible, with very transparent grill designs (at least 60% open area), and then use acoustic filters to control the airflow, if that is necessary to begin with. In some cases, it isn't, and leaving the grills alone leads to good results. As for the effects on the sound, in my experience, blocking the rear grills on V1 and V3 models leads to a gradual increase in the 2-4 kHz region, while gradually making the bass tighter. Blocking the grills past a certain degree leads to an exaggerated mid-high region, gradually spreading the effect over a wider frequency range, combined with a gradual reduction in the low-end performance, leading to an overly stiff and rolled off bass. At the same time, with some drivers, leaving the rear overly opened can lead to a somewhat bloated low-end, lacking in control. It's a matter of fine balance, but luckily, it's very easy to find that balance, as the effects of using acoustic filters to block the ear cup openings are predictable and consistent, just like when using them on the baffles. Additionally, blocking or opening the ports on the back of the drivers themselves, which are usually covered with acoustic filters (paper, mesh, felt...), can lead to similar effects, but IMHO it's best not to touch those filters, unless absolutely necessary, because it could lead to irreversible damage.
- Control what happens within that volume of air in the chamber – Precisely, dealing with the standing waves in the chamber behind the driver, since sound waves are obviously emitted from the back of the driver as well. For this purpose, if required, certain porous or fibrous materials can be used to fill the volume of the chamber to a certain extent, in order to break up the standing waves. Using certain types of foam, or even better, stuff like polyfill, specifically designed to deal with this issue in speaker boxes, is usually very effective, simple and affordable. This isn't particularly useful for tuning the sound balance, but instead for dealing with certain possible issues (narrow peaks at specific mid to high frequencies) that could arise with certain headphone designs, especially those with relatively long, cylindrical rear chambers.
So what does this mean?
It means that it's entirely possible to have a single base headphone design that can work with several different drivers, which require slightly different environments in order to function optimally, because now all of the crucial sound tuning elements can be modified while keeping the majority of the headphone components untouched.
Because of this, it's possible to not only fine tune the sound of a headphone, and achieve different sound signatures when using a single driver, but even after building one configuration of a headphone, it's possible to keep most of the "expensive headphone parts" the same, put an entirely different driver in it, and still have the ability to adapt that headphone to function properly with that new driver, by simply changing the "cheap headphone parts". This is great because it's not only fully reversible, but is also far more efficient and affordable than changing the entire main baffle, or the entire headphone. The best part is, the amount of headphone drivers available is enormous. Even if we just limit our choices to the "Grado mod drivers", the 46 mm diameter ones, there's multiple high quality choices on the market. Expanding to the entire aftermarket driver industry, there are literally hundreds of choices at various price points. And at the end, there are the OEM drivers, since pretty much every major headphone company sells their own proprietary drivers as spare parts, so it's possible to build custom headphones using them as well.
It also leaves us with some interesting implications regarding the potential life-times of such headphones, the possibilities to gradually change or upgrade them in the future, replace parts with new updated and improved versions, replace them with parts produced out of higher quality materials, maybe made with some newly developed additive manufacturing technology that's yet to become available. The possibilities are literally endless. Of course, theoretically, it still takes a lot of work to develop a modular design that actually works in practice, and it's a far more difficult task than developing a single-configuration headphone. But, as I've said at the start, my idea has been to design fully modular, DIY-friendly and future-proof headphone platforms that could serve me well for years, and I think this comes quite close to what I've had in mind. Does it require more work than simply going into a store and buying a headphone? Sure. But it's very fun to play around with, and that's what headphone DIY should be all about. Especially when it's possible to end up with great results, while potentially saving money.