I agree the SPL numbers cited by Andy seem quite high.
This is mostly subjective but relevant: Yesterday, I spent time assessing the desired max SPL of a system I'm working on for/with a friend. He has a cheap radio Shack SLM that correlates reasonably to REW readings with a calibrated mic. We turned volume to max on some rock & blues tracks and saw peaks reaching high 80s at our seated position ~3.5m away. This was very loud, even though undistorted. We didn't keep it at this level for long. I doubt the dynamic range was 20 dB. Perhaps 15. Likely means at the standard 1m measurement point, peaks were well over 100 dB.
This SPL level would be hard to reach, imo, without moving to the high end of drivers suggested this far, and probably not needed for the majority.
Last edited:
Standard listening levels are those typical of a cinema and are just below the levels that would cause hearing damage with long term exposure. One can listen louder for short periods without damage. It is a common level for listening and working with music in the foreground though some will opt for 5 or 10 dB less while other opt for 5 to 10 dB more and suffer hearing damage over time. It seems a reasonable target for a home high fidelity speaker to be fully clean.
Yes I prefer listening around cinema levels when listening to music in the foreground at home.
@andy19191 Thanks for the clarification, yes, that kind of SPL is going to influence many of the design goals.
I believe broader audience would like to know whether or not you are going to design active x/o filters. Do you plan to buy
the drivers or is that supposed to do someone else?
I shall design crossovers for the implementations I build and document the procedures for obtaining the coefficients. The coefficients themselves will vary with the tolerances of the drivers and probably room location (varying the mid/woofer crossover for on-wall vs in room location likely to be more effective than overall room EQ but subject to study) so probably won't publish coefficients.
There are different kinds of digital filters and ones that mimic analogue crossovers are popular since they have lower computational costs. I will likely implement one or two of these for interests sake but would hope others would take a lead in developing them. It is likely some of the active crossovers hardware already owned by people will require this type of crossover due to lack of resources and/or limited access to setting the coefficients.
There are also analogue active filters which seem popular among some audiophiles. I have no wish to use these anymore in my builds but would be interested in getting involved if others wish to do so.
Not sure I wholly understand the question. We will obviously have to establish a good track record if driver manufacturers are going to donate drivers to support our designs. Got a long way to go before that.
I will need to buy a pair of woofers but have several drivers on shelves I would like to use for some of the tweeter/midrange assemblies I hope to build. I will likely use a 5" midwoofer and 1" tweeter I have on the shelf in order to test and check the mounting of the tweeter/midrange assembly to the woofer cabinet particularly if passive isolation is adopted. I will buy the drivers for the twin waveguide assembly version. If an easy build/nice looking version with a (likely) 5" midrange and (likely) 1" tweeter with integral waveguide and passive crossover is settled on I will likely buy the drivers for that but use an active crossover. I also hope to put a pair of previous generation KEF coaxial mids I have on the shelf in a single waveguide to see how things work out. This obviously isn't part of the group project though because the drivers are not available for DIY.
Well, good luck with the donation of drivers by the manufacterers. Irealist. Maybe distributors, but they have their own blogs or known designers like T.G. or elses, or sell kits.
Imho it is going in alll the directions at once. That is not good. You have to lead it as said and took decisions imho or it will end nowhere alas.
The only possibility I see if you don't want to found it or buy yourselfs the drivers, is to ask some members that have the drivers already of a small list you focus on, to show their measurement free air and in box with calibrated mic and if they have some T&S measured the rigth way with added mass enough. If they designed with them already and are nice enough to share their datas or .frd., etc.
Imho it is going in alll the directions at once. That is not good. You have to lead it as said and took decisions imho or it will end nowhere alas.
The only possibility I see if you don't want to found it or buy yourselfs the drivers, is to ask some members that have the drivers already of a small list you focus on, to show their measurement free air and in box with calibrated mic and if they have some T&S measured the rigth way with added mass enough. If they designed with them already and are nice enough to share their datas or .frd., etc.
Last edited:
Andy, I have been looking at your spec document. I posted it here for people who have not seen it on the google drive.
In my opinion, I think you should average together the specs for the KH420 and the 1237A. These two represent the current state of the art. A speaker which sits between them would also be state of the art.
Some of the specs listed below for the two products are probably best understood as attributes of the finished system, but were design targets during the design/development phase. For example, Self Generated Noise. I am not aware of any driver manufacturer publishing this spec for their drivers, so how could the design team select drivers based on this requirement? System weight is another attribute. I very much doubt that the engineers were willing to compromise other aspects of the speaker in order to keep the weight under some arbitrary limit, especially considering the application. I very much believe that whatever the final weight came out to be, that became the spec.
As I said, if you sort of averaged together the requirements, you would have a very good starting point.
In my opinion, I think you should average together the specs for the KH420 and the 1237A. These two represent the current state of the art. A speaker which sits between them would also be state of the art.
Some of the specs listed below for the two products are probably best understood as attributes of the finished system, but were design targets during the design/development phase. For example, Self Generated Noise. I am not aware of any driver manufacturer publishing this spec for their drivers, so how could the design team select drivers based on this requirement? System weight is another attribute. I very much doubt that the engineers were willing to compromise other aspects of the speaker in order to keep the weight under some arbitrary limit, especially considering the application. I very much believe that whatever the final weight came out to be, that became the spec.
As I said, if you sort of averaged together the requirements, you would have a very good starting point.
Detailed Spec (as of 2/17/2025)
| Spec | Target | Neumann KH 420 | Genelec 1237A |
| Free field frequency response | 28 - 20 kHz ± 2dB | 37 - 20 kHz ± 2.5 dB | |
| Linearity 100-10kHz | ± 1.2 dB | +1 -2.5 dB | |
| Self generated noise | < 20 dB(A) @ 10 cm | < 10 dB(A) @ 1m | |
| Max SPL in full space | 116.4 dB, 3% distortion, 1m, averaged over 100-6kHz | 118 dB, half space, 1m, distortion ???, averaged over 100 - 3 kHz | |
| Max bass SPL | 109.9 dB, 3% distortion, 1m, averaged 50-100 Hz, half-space | ||
| Max short term SPL with IEC weighted noise (IEC 60268-5) at 1m in typical listening conditions (?) | 109 dB(C) | 112 dB, long term RMS, 1m, 100 - 3kHz, distortion ??? | |
| Other IEC specs(?) | |||
| Woofer | 10” | 12” | |
| Midrange | 3” waveguide | 5” waveguide | |
| Tweeter | 1” waveguide | 1” waveguide | |
| Crossover frequencies | 570 Hz and 2 kHz | ||
| Cabinet loading | Bass reflex, front ports | Bass reflex, front ports | |
| Dimensions (h, w, d mm) | 645 x 330 x 444 | 680 x 400 x 380 | |
| Weight | 35 kg | 42 kg |
The detailed spec is incomplete and inconsistent and put out there to get people contributing. It would benefit from a couple more current midfield monitors, a few more entries such as amplifier power mentioned a few posts earlier. It is useful in bringing out the type of thing involved if one wants to get quantitative about design.
The most relevant file to the success of the project is likely the task list which is also incomplete and put out there to get people thinking about the questions that need answering before a design can come together. To me this is the fun interesting stuff with the design largely following from what is found out. But is it the fun interesting stuff for others?
So far I have sought to play the role of administrator for a group project rather than some form of leader (beyond putting together a rough spec) in the hope people would pick up and run with some of the tasks. So far this hasn't happened. Perhaps it is too early. Perhaps people need to see more of a project before considering getting involved. Perhaps a leader is needed to provide clarity in the early stages. Perhaps organising a group project like this is unlikely to work.
If a group doesn't seem to be forming organically around a rough spec what should be the next move?
As mentioned earlier I am prepared to commit to some cabinet vibration analysis and some detailed acoustics. That should provide an outline cabinet design and a reasonable acoustic spec in terms of low frequency performance.
Does anyone else want to commit to some relevant tasks over the next month or two?
Does anyone want to take the role of leader?
What else should we be asking/doing to get things going?
The most relevant file to the success of the project is likely the task list which is also incomplete and put out there to get people thinking about the questions that need answering before a design can come together. To me this is the fun interesting stuff with the design largely following from what is found out. But is it the fun interesting stuff for others?
So far I have sought to play the role of administrator for a group project rather than some form of leader (beyond putting together a rough spec) in the hope people would pick up and run with some of the tasks. So far this hasn't happened. Perhaps it is too early. Perhaps people need to see more of a project before considering getting involved. Perhaps a leader is needed to provide clarity in the early stages. Perhaps organising a group project like this is unlikely to work.
If a group doesn't seem to be forming organically around a rough spec what should be the next move?
As mentioned earlier I am prepared to commit to some cabinet vibration analysis and some detailed acoustics. That should provide an outline cabinet design and a reasonable acoustic spec in terms of low frequency performance.
Does anyone else want to commit to some relevant tasks over the next month or two?
Does anyone want to take the role of leader?
What else should we be asking/doing to get things going?
When asking a group of people for a volunteer to step forward, sometimes all the smart people take a step back, leaving one person as the "volunteer". Everyone took a step back about a week ago @andy19191 ... 😉 Yes, you are the leader (IMO). This project desperately needs someone to assume the responsibility of being the final decision authority.
Tasks
I will look into the following two tasks. Give me a few days to put something preliminary together that we can chew on and discuss
I have a 3-way project in the pipeline, but other projects and work keep getting into the way, unfortunately.
anyhow, I am very much interested in this project and would also contribute - I'll just need to find some time for it.
(I will also try to finally complete my port optimizer spreadsheet during the next days, by the way!)
I already made a small proposal in post #285, regarding different baffle variants:
https://www.diyaudio.com/community/...-box-coffin-group-project.423501/post-7928192
Regarding the task document excerpt hifijim quoted above:
I would be interested to study possibilities to have a flat and classic looking "monkey coffin" baffle, that is (partially) made of melamine foam and thus eventually combining effects of a wide baffle (for lower frequencies) and absorptive surfaces (for higher frequencies), maybe mimicking the advantages of small baffles and/or big roundovers.
anyhow, I am very much interested in this project and would also contribute - I'll just need to find some time for it.
(I will also try to finally complete my port optimizer spreadsheet during the next days, by the way!)
I already made a small proposal in post #285, regarding different baffle variants:
https://www.diyaudio.com/community/...-box-coffin-group-project.423501/post-7928192
Regarding the task document excerpt hifijim quoted above:
I would be interested to study possibilities to have a flat and classic looking "monkey coffin" baffle, that is (partially) made of melamine foam and thus eventually combining effects of a wide baffle (for lower frequencies) and absorptive surfaces (for higher frequencies), maybe mimicking the advantages of small baffles and/or big roundovers.
Last edited:
I also think it is interesting to consider how to make a classic sharp edged box that isn't plagued by significant edge diffraction.
The simplest way or at least most common is through increased directivity, a waveguide like in both the Neumann and Genelec speakers mentioned. This could be good enough from an audibility and frequency balance perspective, like in most of the affordable KEF speakers. If the target is a wider pattern then that will not be so effective.
A very narrow baffle will cause diffraction loss to be more spread out, it smooths out the baffle step by spreading it over a wider range. You cannot get this same response through another method unless it also is diffractive in nature, I can't imagine how that would practically be approached either..
If you can successfully absorb the off axis radiation there will be no edge effect, there will be absorption loss but not diffraction loss if this makes sense, so they will not be equivalent to each other. The position and surface of the absorber would also be important as it could be reflective enough at some frequencies to behave like a baffle or like a soft diffractive edge if it was sharp enough. There is only one shot at absorption in this configuration as the sound won't bounce around progressively losing more energy like inside a cabinet.
If there was enough reflectivity in the absorber material it could be shaped like a simple waveguide, like the old dunlavy's.
It is possible to simulate these sort of things with BEM but the results are very dependent on the damping factor which is hard to make true to life, but it does show what is possible if a material with those properties could be found.
The right kind of felt has been shown to work in this kind of configuration, but actually finding the right kind of felt that works and is available for everyone, is not expensive or ugly if exposed is not easy. Most felt that looks good is reflective.
Thanks fluid!
I was thinking about (melamine) foam covered with nice looking fabric.
I think that would pass the good looking test, but I don't know the surface properties of it and if it is similar across manufacturers. I think that would need a test to see. You have made some good practical tests like this in the past, if you are open to try another one 🙂
There seems to be enough positive interest at this early stage for it to be reasonable to start a group project. Acting as leader was not my intention since it conflicts with my exit strategy mentioned earlier should the group project fade away as many do. My two initial proposed tasks fit nicely with testing the software I am developing so not too much would be lost for me if the project faded away. Leading would be effectively committing to building the speaker in the summer if I don't move home or later if I do. There is a fair chance the downsize will not include a workshop/garage with construction/assembly then taking place in the home. This could be part of the design but is not a natural fit with maximising the technical performance of a DIY speaker. It would likely widen the appeal though for people looking for a first build.
I need to think about it for a few days. If I am to lead then the first task is to further clarifying the project in terms of objectives, likely tasks, likely timescales,... which fits well with thinking about it.
For those of us that listen with grilles on shallow waveguides, roundovers, foam,... to reduce diffraction can be built into the grille while keeping the sharp edges of a monkey coffin. Grilles the thickness of the JBL L100s could do a fair bit. On the other hand it would lose the thin picture frame around the grille aesthetic of retro British monitors which would likely need a less thick grille to look right.
A wide baffle at low frequencies is going to be wide without something like multiple drivers with different relative phases to shape the wavefront. I think the difference in wide vs narrow baffle is likely best embraced as a feature.
I need to think about it for a few days. If I am to lead then the first task is to further clarifying the project in terms of objectives, likely tasks, likely timescales,... which fits well with thinking about it.
For those of us that listen with grilles on shallow waveguides, roundovers, foam,... to reduce diffraction can be built into the grille while keeping the sharp edges of a monkey coffin. Grilles the thickness of the JBL L100s could do a fair bit. On the other hand it would lose the thin picture frame around the grille aesthetic of retro British monitors which would likely need a less thick grille to look right.
A wide baffle at low frequencies is going to be wide without something like multiple drivers with different relative phases to shape the wavefront. I think the difference in wide vs narrow baffle is likely best embraced as a feature.
Another question to be resolved is how much performance degradation do we accept due to the grill? My preference would be a grill design that looks good, but is only meant for casual listening... serious listening would require grill removal.
This project, like any project, needs someone willing to build a prototype and/or finished cabinet. It needs someone willing to buy drivers for testing in that prototype or finished cabinet.
That seems to be the wrong way round? The intention is for a well designed grille to improve performance compared to the case with it off and all edges exposed. I doubt the effect will be particularly significant due to the wide baffle. It is one of the things we ought to quantify as part of the project.
With a passive crossover I can see there may be a case for choosing to optimise for grille on or off when seeking to use bumps and dips from edge diffraction to cancel bumps and dips from the drivers and crossover. The flexibility of an active crossover lessens this. Effectively removes if one is prepared to switch between grille on and grille off coefficients.
In a domestic setting I don't like to see exposed drivers and when tiny fingers are out and about or I am being particularly clumsy the speakers are safer with grilles on. Of course many like to have the drivers on display so it would likely be unwise to require a grille to be on. Two sets of coefficients for active crossovers and choose to optimise for one the other with passive?
Anyway designing a grille to improve the performance looks like an interesting and relevant task to me. Sticking felt and/or plastic to the cabinet to achieve similar effects and then sticking a grille over it works for me but perhaps less so for those that want the grille off. Perhaps it can be done in a way that is visually attractive for those that want drivers exposed.
An acoustically transparent grill system that avoids edge diffraction was created by Linn for a speaker model dubbed Kaber in the late '80s. The grill was entirely frameless, just a bit of cloth with an elastic.
It consisted of a groove around the perimeter of the front baffle, around the top, sides & bottom of the baffle, less than an inch from the front. Stretchy grill cloth cut to the right size/shape with an elastic around its perimeter like a bath cap fit neatly into the groove perfectly flush with the baffle.
I used this in a couple of builds way back, but lost pics of...
Let's see if I can find an illustrative pic of the
Ok, this is about the best single image I found:
It consisted of a groove around the perimeter of the front baffle, around the top, sides & bottom of the baffle, less than an inch from the front. Stretchy grill cloth cut to the right size/shape with an elastic around its perimeter like a bath cap fit neatly into the groove perfectly flush with the baffle.
I used this in a couple of builds way back, but lost pics of...
Let's see if I can find an illustrative pic of the
Ok, this is about the best single image I found: