Tall Story Tests
So I have a few things to share on the tests I have been making.
The first is the image. It is a white noise test which shows the raw response of the tall panels, no eq and no filtering. The results are clear, a ‘saddleback response with some eq required to flatten the response.
I attach 2 MP4 recordings of the panels. Both were recorded raw with no eq of any description. These were recorded in a room so there are some obvious room reflections evident, also some noise floor at the start of each recording is very evident. Sorry about that but I could not record at the level I would have liked or I would not be popular.
You may need to download the free VLC media player to listen to them.VLC – Download VLC Media Player for Windows, Mac, Android & iOS
I hope you find the results interesting and please when listening to the MP4 files have in mind Veleric’s warning about the potential for misleading results. This is very definitely a work in progress.
Burnt
So I have a few things to share on the tests I have been making.
The first is the image. It is a white noise test which shows the raw response of the tall panels, no eq and no filtering. The results are clear, a ‘saddleback response with some eq required to flatten the response.
I attach 2 MP4 recordings of the panels. Both were recorded raw with no eq of any description. These were recorded in a room so there are some obvious room reflections evident, also some noise floor at the start of each recording is very evident. Sorry about that but I could not record at the level I would have liked or I would not be popular.
You may need to download the free VLC media player to listen to them.VLC – Download VLC Media Player for Windows, Mac, Android & iOS
I hope you find the results interesting and please when listening to the MP4 files have in mind Veleric’s warning about the potential for misleading results. This is very definitely a work in progress.
Burnt
Attachments
I'm no golden ear or anything close to it, so anything I describe may sound strange as I don't know all those "audiophile" terms that are commonly used.
Now I have never heard that music before on the first track posted, so I don't know if its some audiophile recording or not. So to me it sounded a bit strange. The singers voice sounded somewhat disconnected from the instruments. What I mean is that there seemed to be some artificially generated reverb that to me was very evident in the voice that seemed to separate or disconnect the voice from the instruments playing as if they were done separately and then joined together in a mix. Secondly the wood block hits (I assume thats what they were) had very striking attacks and yet on the decay I could hear what sounded like space or air around the decaying attack. Very interesting and pleaseing. On the loudest passages of the voice there appeared to be some noise distortion mudding up the vocals. Then towards the end, there seemed to be what I originally thought was a tenor sax playing but it sounded very muted as if someone had put a mute on the bell...but then it occurred to me that that was probably just a synthesized sax from a keyboard rather than the real thing...dummy me. The spaciousness of the recording just seemed a bit weird to me as if there was something very artificial going on...not unpleasant per se, but definitely not like a real live event.
Hopefully that makes sense to somebody that knows this recording and its pedigree. Let me know if I'm way off base. I only listened to it once through some cheap ear buds on an HP laptop. What I did find interesting is that I could actually hear the details I described above which I think is a testament to the clarity or naturalness of the recording of the speaker playing, i.e., the speaker made the flaws very obvious to my ears...just what I want a speaker to do.
Now I have never heard that music before on the first track posted, so I don't know if its some audiophile recording or not. So to me it sounded a bit strange. The singers voice sounded somewhat disconnected from the instruments. What I mean is that there seemed to be some artificially generated reverb that to me was very evident in the voice that seemed to separate or disconnect the voice from the instruments playing as if they were done separately and then joined together in a mix. Secondly the wood block hits (I assume thats what they were) had very striking attacks and yet on the decay I could hear what sounded like space or air around the decaying attack. Very interesting and pleaseing. On the loudest passages of the voice there appeared to be some noise distortion mudding up the vocals. Then towards the end, there seemed to be what I originally thought was a tenor sax playing but it sounded very muted as if someone had put a mute on the bell...but then it occurred to me that that was probably just a synthesized sax from a keyboard rather than the real thing...dummy me. The spaciousness of the recording just seemed a bit weird to me as if there was something very artificial going on...not unpleasant per se, but definitely not like a real live event.
Hopefully that makes sense to somebody that knows this recording and its pedigree. Let me know if I'm way off base. I only listened to it once through some cheap ear buds on an HP laptop. What I did find interesting is that I could actually hear the details I described above which I think is a testament to the clarity or naturalness of the recording of the speaker playing, i.e., the speaker made the flaws very obvious to my ears...just what I want a speaker to do.
And just to make this perfectly clear, I wasn't listening to the speaker, I was listening to the recording that the speaker was playing. The speaker to me wasn't adding anything, rather it was letting the flaws of the recording pass through it and representing it as it was made...to me thats what a good speaker is about, listening to the recording and not the speaker itself. Good speakers make you want to hear whats there, not what the speaker trys to make it sound like...does that make any sense?
All you have to do is find that recording track and compare it to Burnts track.
You are listening to a recording playing from a set of speakers being recorded again and listening to them from a set of headphones. LOL
How do you know the speakers weren't adding anything? What are the flaws of that recording? If you've never heard that music before how would you know the flaws in that track? How do you know the speaker was representing the recording as it was made/intended? Its not making any sense just hype. LOL
I guess its would have helped if I had included track details for comparison.
DML Test 1 is “Iv’e done it again” by Grace Jones from the 1981 Album ‘Nightclubbing’, details here Nightclubbing (Grace Jones album - Wikipedia)
DML Test 2 is a cover of Massive Attacks Unfinished Symphony’ by London Grammar called “Bitter Sweet Symphony” . It was recorded live at BBC Maida vale Studios. The track is available via iTunes, Tidal etc. London Grammar - Wikipedia
The tracks were recorded on an iPhone and then converted to audio MP4 to allow me to upload to the site.
My intension was just to demonstrate that usable bass down to circa 40hz was possible with this panel ratio, nothing more.
I have my thoughts on areas the panels need work on but for now I will leave it up to anyone who wants to comment.
DML Test 1 is “Iv’e done it again” by Grace Jones from the 1981 Album ‘Nightclubbing’, details here Nightclubbing (Grace Jones album - Wikipedia)
DML Test 2 is a cover of Massive Attacks Unfinished Symphony’ by London Grammar called “Bitter Sweet Symphony” . It was recorded live at BBC Maida vale Studios. The track is available via iTunes, Tidal etc. London Grammar - Wikipedia
The tracks were recorded on an iPhone and then converted to audio MP4 to allow me to upload to the site.
My intension was just to demonstrate that usable bass down to circa 40hz was possible with this panel ratio, nothing more.
I have my thoughts on areas the panels need work on but for now I will leave it up to anyone who wants to comment.
Last edited:
You are right, the recording has the following flaws .
- Its from sub-optimised panels with obvious frequency response flaws
- Its recorded in a room with wall reflections and if you listen closely a double decker London bus adding bass
- Its recorded on an iPhone microphone
- Its been through a couple of conversions to get it to a format and a size acceptable to the site upload limits.
- all headphones each have a sound
....and I only intended to show with these tracks that you can get bass, although I am grateful for any constructive feedback on any aspect of the test.
@geosand
Thank you for your very detailed feedback which I found most interesting.
Yes track 1 is a studio recording and all ambiance effects are added in the mix.
Yes its a keyboard, not a sax. You have good ears and your descriptive ability needs no jargon,
I have included a link in the post above to more information on the album and the artist.
Thanks again
Burnt
Last edited:
Let me clarify something. Could your main 4ft.X2ft. panels reach down to 40hz? Or was it just that the bass output was less?
@DMLBES The bigger panels rolled off a lot higher, circa 80 hz from memory. The output at 40hz was very low, that’s why I used subs. Similar experience in my other panels although they did go a little lower, circa 50 hz
I will measure them in the morning as it’s late here and then post the plot.
Burnt
I will measure them in the morning as it’s late here and then post the plot.
Burnt
Last edited:
Thanks Burnt. So thats what the rumbling was that I heard. I didn't mention it because I wasn't sure what it was. Its funny how some can listen to a recording and make out the clearly delineated "issues" in the recording and others just say, oh it must be the speakers or the recording methods or what not. To me its a testament to a good speaker and recording when you can clearly hear the issues in the recording itself. I don't mention such things as a high end rolloff or low end weakness, because those are an obvious part of the playback chain. At least no engineer I know purposely tries to roll off high/low ends to be clearly audible, they do so for other reasons and strive to make them as inaudible as possible. Frankly, I am astonished at the sound quality and subtle details I heard despite the re-recording chain and my simple basic listening chain AND the fact I only listened to it once, for what 3 mins? Your speakers really impressed me, even more so given the limitations I stated.
Could you provide the general details of the speaker again...excitors, panel material, size/shape?
And thanks for going to all the trouble to do and post the recording. I'd love to hear more...
geosand
Oh ok thanks....Now I am pretty much tempted to alter the width of my panels from 20inch down to maybe 15inches. I was thinking it could be like a bow effect where longer thinner bows flex more so its easier to string while the thicker but shorter bows are harder to flex so its harder to string.
My 23inch height X 15inch width sub panel can reach down to 50hz. It can go 40hz but at lower spl levels. So if I increase the height to make it longer hopefully it will flex/bend easier and produce more lower bass down to 40hz at adequate volume levels.
Panel Response
@DMLBES. So much for my memory. The big panels fall away from 250k down. Response below.
Just a couple of brief comments on the plot.
1. It’s an averaged FFT plot using white noise as an input.
2. There is a noisy PC in the room acting as a server which has a lot of low frequency noise which is why the drop in output appears to level off and flatline. The panels don’t suddenly level out at 60 Hz 15 dB lower than the rest of the spectrum, they keep going down.
Good luck with your experiment
Burnt
@DMLBES. So much for my memory. The big panels fall away from 250k down. Response below.
Just a couple of brief comments on the plot.
1. It’s an averaged FFT plot using white noise as an input.
2. There is a noisy PC in the room acting as a server which has a lot of low frequency noise which is why the drop in output appears to level off and flatline. The panels don’t suddenly level out at 60 Hz 15 dB lower than the rest of the spectrum, they keep going down.
Good luck with your experiment
Burnt
Attachments
Last edited:
That was kind of you to say geosand, I am glad you like them.
I will put together a post that gives the detail for you. They are a very simple build.
Thanks again for the positive comments, much appreciated.
Burnt
Hi all,
I just starting with DIY speaker building and am very intrigued by the DML design's I am reading about on this topic and in the topics on techtalk.parts-express and audiocircle (links below).
As far as I have been reading the best material to use is a sandwich plate of Carbon Fiber (CF) with a Nomex Honeycomb Core (NHC). But this material is not often used due to availability and price.
I have the option to create my own sandwich plate for a low cost due to some connections I have.
Here are my main questions:
1) I don’t understand why the NHC is the best core material to use in a sandwich plate because it has a very high sound absorption. Birch plywood has a very low sound absorption and is used very often as DML material without the CF. Which one would be better to start with?
2) I am planning to make a 2.1 set-up with DML panels. My panel for the Low end of the spectrum or as sub van have a width of 90” and a height of 15”, now my question is of some of you have experience with the 1 to 6 ratio and what pros and cons it have (planning to use 4x DAEX30HESF-4 or DAEX32EP-4 not sure which is better).
3) The location of the exciters has different options, the Manacor, the 2/5 – 3/5 ratio and several other options, but I haven’t seen a conclusion which is best, if the answer is that I have to test, I am planning to that, but if I know what option seems to be the best it is good to know where to start.
Thank in advance for the help,
Kr,
DIY Flat Panel Speaker Love -
Techtalk Speaker Building, Audio, Video Discussion Forum
NXT.......rubbish??....THINK AGAIN!
I just starting with DIY speaker building and am very intrigued by the DML design's I am reading about on this topic and in the topics on techtalk.parts-express and audiocircle (links below).
As far as I have been reading the best material to use is a sandwich plate of Carbon Fiber (CF) with a Nomex Honeycomb Core (NHC). But this material is not often used due to availability and price.
I have the option to create my own sandwich plate for a low cost due to some connections I have.
Here are my main questions:
1) I don’t understand why the NHC is the best core material to use in a sandwich plate because it has a very high sound absorption. Birch plywood has a very low sound absorption and is used very often as DML material without the CF. Which one would be better to start with?
2) I am planning to make a 2.1 set-up with DML panels. My panel for the Low end of the spectrum or as sub van have a width of 90” and a height of 15”, now my question is of some of you have experience with the 1 to 6 ratio and what pros and cons it have (planning to use 4x DAEX30HESF-4 or DAEX32EP-4 not sure which is better).
3) The location of the exciters has different options, the Manacor, the 2/5 – 3/5 ratio and several other options, but I haven’t seen a conclusion which is best, if the answer is that I have to test, I am planning to that, but if I know what option seems to be the best it is good to know where to start.
Thank in advance for the help,
Kr,
DIY Flat Panel Speaker Love -
Techtalk Speaker Building, Audio, Video Discussion Forum
NXT.......rubbish??....THINK AGAIN!
RFN,
Welcome to the DML club! My thoughts are below.
The main advantageous feature of Nomex is that it is lightweight. Combined with Carbon fiber skins, it provides a structure with very high ratio of stiffness to weight, which in turn provides higher SPL for the same power.
Plywood, birch or other, has decent, but not great stiffness to weight ratio, and provides what I view as adequate (but far from impressive) volume level. On the plus side, in my experience, plywood can provide as flat and as broad a frequency response as any panel material that I have tested so far. I have not actually tested carbon/nomex, so I don't know for sure that it provides as broad a frequency response, but I expect it can.
If you do try Carbon/nomex, I suggest trying panels no more than about 3 or 4 mm thick overall. Thicker panels might work too, but will have to be very, very large cover the low frequency range.
Others my disagree, but my suggestion would be to use a conventional subwoofer and use the panels for, say 150 Hz and above.
There really is no universal "best" position for exciters. The ideal location will depend on the properties of the panels, how they are supported (framed), which exciters they are, what result you hope to achieve, and on and on...
That said, I have actually been surprised at how little the exciter location really matters. Generally, placing them in the vicinity of the center (anywhere within about the middle 20% of the panel) works almost as well as any other place within that same region, in my experience. But, I do suggest you test at least a few different positions, mainly to be sure that you don't randomly stumble onto a particularly poor location by dumb luck.
Hope this helps.
Welcome to the DML club! My thoughts are below.
The main advantageous feature of Nomex is that it is lightweight. Combined with Carbon fiber skins, it provides a structure with very high ratio of stiffness to weight, which in turn provides higher SPL for the same power.
Plywood, birch or other, has decent, but not great stiffness to weight ratio, and provides what I view as adequate (but far from impressive) volume level. On the plus side, in my experience, plywood can provide as flat and as broad a frequency response as any panel material that I have tested so far. I have not actually tested carbon/nomex, so I don't know for sure that it provides as broad a frequency response, but I expect it can.
If you do try Carbon/nomex, I suggest trying panels no more than about 3 or 4 mm thick overall. Thicker panels might work too, but will have to be very, very large cover the low frequency range.
Others my disagree, but my suggestion would be to use a conventional subwoofer and use the panels for, say 150 Hz and above.
There really is no universal "best" position for exciters. The ideal location will depend on the properties of the panels, how they are supported (framed), which exciters they are, what result you hope to achieve, and on and on...
That said, I have actually been surprised at how little the exciter location really matters. Generally, placing them in the vicinity of the center (anywhere within about the middle 20% of the panel) works almost as well as any other place within that same region, in my experience. But, I do suggest you test at least a few different positions, mainly to be sure that you don't randomly stumble onto a particularly poor location by dumb luck.
Hope this helps.
Don't know if you have seen this thread?
Carbon fiber + nomex honeycomb sandwich DML panel construction project
So Burnt which type of panel design to you prefer better?
Free floating or Frame and spline?
nesha,
Very interesting, thanks for posting this. I didn't know anybody makes guitars like that, but it makes a lot of sense to me. I have already tried making a DML panel with nomex core and wood veneer skins. It seems promising, but not perfect. Definitely louder than plywood. But I did not feel very confident in my fabrication process. But a concept I'm planning to work more on. Maybe I can learn from the guitar makers!
Eric