Open baffle or no baffle for nearfield full range monitors?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello all,

I'm hoping to get a pair of Jordan Eikona 2 drivers from the groupbuy here on diyaudio: http://www.diyaudio.com/forums/group-buys/282230-new-jordan-eikona-2-group-buy.html
I wish to use them to make full range nearfield monitors. Going for a very pure sound, something in between headphones and normal speaker use is probably the best description.
Listening distance (ear to driver) will be 50 to 70 cm, drivers suspended in the air at ear height without any reflecting surfaces nearby (and wall behind the drivers is about 6 meters away), no box or porting etc, will make the frequency flat by high quality EQ on my computer.

Originally I thought I would do best to do away with any open baffle and simply use the drivers "naked" with some minimal suspension system (was thinking something based on mic stands with boom arm so I can position and angle the drivers any way I want).
But now I'm questioning if doing away with a baffle is a good design because of several reasons. First, will the surround of the driver cause edge diffraction which is a form of distortion and will create a new wavefront all around the edge making the "point source" a lot bigger?
And secondly, a driver is not a perfect dipole. I picked up a small full range driver and listened carefully from all angles and the back sounds completely different than the front, not only in the treble (of which there is about none in the back) but also in the mids. And from the side there isn't anything like perfect cancellation of frequencies.

Now my questions are the following:
Does anybody know if this "compromised" audio from the back can reach my ears at the front through methods like edge diffraction from the surround edge or any part of the driver construction at the back (perhaps combined with reflections at the back of the driver from parts of the driver construction / magnet system), or simply bend around the driver / surround by omnidirectional wave behavior at lower frequencies? (at which frequencies would this behavior start?)
Now a baffle with nice rounded edges will reduce or eliminate edge diffraction from the surround edge, but how large does it need to be to "block" for instance any audio from the back above let's say 500Hz?
And what I also don't understand yet is if and how the effect of baffle size changes with listening distance from the baffle.

Btw to make clear, I'm not looking at baffle use for EQ / Qt purposes. I will do all my EQ-ing of the driver on the computer.
Also, if I'll use a baffle I prefer it to be as small as required for sound quality purposes, and make it round with the driver in the middle, no need for me to smooth out the bass dropoff below baffle size with baffle design, a well defined point will only make EQ-ing easier it seems to me.

Many thanks for any insights anybody can share on this!
 
Now a baffle with nice rounded edges will reduce or eliminate edge diffraction from the surround edge, but how large does it need to be to "block" for instance any audio from the back above let's say 500Hz?

Infinite baffle completely separates frontal wave from back wave emission.
And the baffle needs to be solid, i.e. sound-proof

Once you EQ the driver to be linear, the distortion will rise because the driver excursion for reproducing the bass will modulate the motion for the mid and high frequencies, so it's better to use dedicated drivers if you fell you're going to surpass the extimated limit of the driver.
I guess that 10 W would be the maximum allowable for such driver in the (low) bass region.

*the speakers should be suspended with rubber bands and not just laid on something*
 
Thanks for the replies!

But my thinking was that normal rules apply less because I'm listening at such a close distance.
50cm listening distance will give me 12db more volume than 2m listening distance, and 18db more volume than a somewhat normal 4m listening distance.
So if I have to EQ 18db extra in the bass at 50cm listening distance for a "naked" driver to be linear, the driver works just as hard as a very large ported speaker at 4m listening distance.
And I don't think I need to EQ 18db extra to get the bass linear with a "naked" driver, but less. So I would actually gain volume or less distortion and intermodulation distortion, including in the bass, relative to a big ported speaker at 4m listening distance.

Eikona_2_Freq.jpg

Here is the frequency response of the Eikona 2 driver.
I'm assuming this was measured anechoic and with a "naked" driver?
If so, I would need to EQ around 10db at 50Hz (I'm fine with 50Hz don't need it to be linear any lower). So after EQ-ing I still have about an 8db advantage over a big ported speaker at 4m listening distance for 50Hz reproduction.
So I'm not going to exceed the designed limits of the driver at all it seems to me (also do not need ear shattering levels, moderate levels are fine).

Of course the driver itself acts as a baffle as well. But as by my original message, I was wondering if I'd do well to still make some kind of additional baffle for sound quality purposes. And what are the formulas for baffle size and it's influence on frequency? Also I'm very interested to learn if baffle size and distance are linked in some way to the baffle influence on frequency. I mean, a 50cm baffle at 50cm distance is relatively as big as a 3m baffle at 3m distance. Instinct tells me that distance should matter as well as baffle size regarding baffle frequency effects but maybe I'm wrong about this?
Could someone perhaps point me to a good website which explains all this in a somewhat understandable manner?
 
Hi,

You don't understand that baffle loss throws away speaker excursion,
not just response. You have absolutely no hope of EQing it down to
50Hz if its mounted on a small baffle / no baffle, even at 0.5m.

The above response is certainly not a naked driver.
Its into half space on a very large open baffle.

Your looking at about 20dB baffle loss at 50Hz and very poor SPL.

I'd recommend a tallish, thin, well rounded box, very well built,
about 10L, fitted with a Peerless 830878 3-1/2" passive radiator,
then EQ away to your hearts content. Sealed is also possible, 5L.

rgds, sreten.
 
Last edited:
Hi,

You don't understand that baffle loss throws away speaker excursion,
not just response. You have absolutely no hope of EQing it down to
50Hz if its mounted on a small baffle / no baffle, even at 0.5m.

The above response is certainly not a naked driver.
Its into half space on a very large open baffle.

Your looking at about 20dB baffle loss at 50Hz and very poor SPL.

I'd recommend a tallish, thin, well rounded box, very well built,
about 10L, fitted with a Peerless 830878 3-1/2" passive radiator,
then EQ away to your hearts content. Sealed is also possible, 5L.

rgds, sreten.

Aaah.. Thank you for that information!!
I was just beginning to suspect that. The open baffle math didn't make sense at all with the given frequency response of the Eikona 2. I was calculating the Eikona 2 as having a 15cm "baffle" as a naked driver, and indeed came to about 21db difference at 50Hz compared to a ~2m open baffle.
Ok, so that published frequency response is already open baffle.. That changes a lot.

That brings me to another question.
Baffle is half space. If you were to attach it to a tube (or horn) of 1m length, that would in a sense behave as a 2m baffle, except that the front wave doesn't see half space but nearly full space. Would that give 3db/oct instead of 6db/oct dropoff below true baffle size? (15cm in the case of the driver)
And I'm suspecting a simple tube would give a lot of resonance which would still be audible at listening position even when the end of the tube is 3 times as far away from the ear as the driver, so would a widening tube / horn fix that? (I was thinking making the angle such that it would be "hidden from sight" from the perspective of my ear)
I still really want to explore my options of getting away from a box.
 
Of all speakers I've heard (a lot, have been listening to studio monitors and "audiophile" speakers for about 20 years now), I'm left with a sense that all speakers have a box colouration of some sort. I want to get away from that. It's not as bad as crossovers, those I've really learned to hate (including well executed phase correct active crossovers).
Ports are of course especially bad. But I've had no luck either with closed boxes. For instance I've owned the Klein+Hummel O300. It is a closed box and doesn't suffer from any port colouration, but it does the opposite it actually limits the driver excursion in some way which is horrible at higher volumes. But even besides this maybe it's the speaker with the least audible box colouration I've heard but it still has it a little bit.

So I don't want crossovers / multiple drivers, I don't want a box, and I don't want a flat baffle. (oh and I also don't want headphones, mainly because of losing the head related transfer function)
I think I'm in trouble haha ;)
 
How nearfield ?

I had a pair of tang band w3-871s on an open baffle, 6" between frames.

At 1' it was magical and had bass.

I was thinking 50cm to 70cm (about 2 feet).
Ok so you listened at one foot and got bass from a 3" driver! And it was magical :) Good to hear!
I'm hearing too there's something very special to be found listening really nearfield to a single driver without a box. Right now using 3.3" drivers I took out of some old speakers to test my idea a bit, one of the drivers has a rotten surround so I can't listen well, and I don't even have them on a baffle just bare. And yet still I hear something very special (bit EQ by ear helped though)..

Can you give me detail on your baffle size? And did you listen often to it that close or was it intended for something else?
 
Maybe?

Years ago I made an ambiophonic demonstrator.
3" drivers more than 8" between speaker frames.
Got moldy in garage, tossed it and sold drivers.

But I used Roger Waters "amused to death".
I didn't need the ambio barrier because that was a q-sound ( crosstalk cancellation) recording.

Since you are 2', maybe a 4" driver.
Being 1' from the drivers, the ob bassrolloff wasn't happening.

Since you are 2', I'd assume 15" from speaker frame to ob edge.

But, you will have to try it, I may be off on my assumptions.
The further back, the more the drivers will hit the opposite ears.

Try a barrier.

Only problem is it hardly a relax on the couch / chair and listen to music.
Otherwise we would have studio monitors 3' from our ears.

I think most prefer a juke box with good sound everywhere.
 
Last edited:
Thanks for the info!

I do not want a barrier though as I think sound is supposed to hit both ears for a natural sound. Not doing this is one of the problems of headphones. Some people are adding crosstalk artificially with headphones to get more natural imaging but it still sounds very artificial compared to natural sound hitting the head and both ears.
I think there's a line there somewhere between behaving more like headphones and behaving more like speakers. I'd like to flirt with that line a little bit but stay within speaker / nearfield monitor range.

15" from speaker frame to ob edge is a lot though.. that's 38cm, if I'd use the Eikona 2 which is btw effectively a 4" (10cm effective driver diameter) which has a 15cm frame size that would make 38+15+38 = 91cm total baffle width. If I'm listening in an equal triangle this would make the baffles touch well before 61cm (2 feet).
I'll do some calculations on optimal baffle size..
 
you'd be surprised.
The 2 speakers with a barrier, the sound is directly in front of you.
Moving speakers back to the stereo triangle made the center voice a 3' wide haze less detailed blob that also was less loud (maybe 3db).
But I see what you mean so far as how something is mixed.
You can always put both speakers on 1 baffle.



You could play with how far apart you want the drivers also.
 
Well I see no other way than a flat baffle indeed..
I just put my little test full ranges in tubes 30cm long with 10cm diameter, open back, and while this indeed gives more low mids and bass etc, it also sounds like it is in a tube. Pleasant with some music though, it resonates and gives weight etc but in no way natural and light / quick.
I'm guessing also a tube / horn that gets wider will still impart a sonic signature. Probably only a flat baffle will not impart sonics that depart from neutral and only give a frequency related effect.

Calculated the following for baffle size and angle.
In an equal triangle setup with 65cm sides and my ears at the lines of the equal triangle gives 50cm ear driver distance, 75cm (round) baffle size where the baffles touch where the drivers are pointing directly on axis along the equal triangle line (and my ears).

So with a baffle size of 75cm I'd get a 6db/oct additional drop off starting at 34300/(75*2) = ~229Hz relative to the published frequency curve.
So at ~57Hz it would be -12db relative to the published frequency response.
Compared to the same driver at 4m distance in an infinite baffle it would still be 6db louder at 57Hz because 50cm distance is 18db louder than 4m. So driver performs relatively 6db better at 50cm with a 75cm baffle.
If the 4m away driver uses a port / transmission line system it would add about 6db at 57Hz without extra effort for the driver, but this comes at a cost of bad impulse behavior / bass quality. So I could EQ +18db at 57Hz 50cm listening distance and get perfect quality bass that is linear at the same level and stress to the driver as the ported bass at 57hz 4m listening distance.
So with a 75cm baffle I see no problem using the Eikona 2 driver this way.

But.. then I read about edge diffraction a bit more.
And it seems the edge of the baffle will cause a lot of diffraction all over the spectrum. So a round baffle isn't a good design.. In fact there doesn't seem to be a great baffle design possible without edge diffraction. I'm not just worried about the effects on frequency, but more about that this means the baffle is "audible" and doesn't disappear from the ear's "sight". Which was one of my main points to begin with, disappearing speaker.
So maybe this is not what I'm looking for.. Maybe a half sphere baffle would still be better? Maybe a closed box that is a near perfect sphere? (saw a nice ikea blendle design..) Nature seems to want either a closed sphere box or an infinite baffle / flush mounting.
 
Hi,

You don't understand that baffle loss throws away speaker excursion,
not just response. You have absolutely no hope of EQing it down to
50Hz if its mounted on a small baffle / no baffle, even at 0.5m.

The above response is certainly not a naked driver.
Its into half space on a very large open baffle.

Your looking at about 20dB baffle loss at 50Hz and very poor SPL.

I'd recommend a tallish, thin, well rounded box, very well built,
about 10L, fitted with a Peerless 830878 3-1/2" passive radiator,
then EQ away to your hearts content. Sealed is also possible, 5L.

rgds, sreten.

There is a way to use the drivers "naked" but with very little OB losses. How? Put them up to your ear(s)! We can thank the proximity effect for that. This is pretty well known for microphones, but applies just the same for loudspeakers.

Take a woofer, suspend it in free air a meter or two away, and give it a listen. No bass, right? But then put your ear right up to the diaphragm and then what do you hear? You hear more bass! This is the proximity effect in action. NOTE that this will only be evident with drivers that actually have a response that extends into the bass region...

Even with a modestly sized OB the response below 100Hz or so is influenced by how close you are to the baffle. When you are close to the "front" of the speaker cone the SPL from it will be louder than the "rear" SPL because the pathlengths are appreciably different. This causes front SPL >> rear SPL, meaning there is not as much front-to-back cancellation compared to when you are farther away and the pathlengths are relatively similar. What you hear is trending towards infinite baffle (closed box) response.

You can hang the drivers so that they are positioned just to either side of your head/ears and you will be getting the on-axis, near-infinite-baffle response beamed right in there, not to mention you will only need very little amplifier power to drive them. It's won't look very pretty, but you will be in the nearfield and the room response (including room resonances) will be imperceptible and you will have practically no "baffle step" or "open baffle" losses. The sound will be similar to when using headphones, however, and some of the benefits of the open-baffle loudspeaker's room interaction will be lost.

Keep in mind this will not "augment" the bass. For a full ranger driver the infinite baffle bass output is typically rather anemic. But because your ear(s) are very close to the driver(s) their SPL output level will be much lower, meaning you can boost the low end more for the same overall excursion (compared to a typical listening distance of 2m). This is why a tiny headphone driver can still reproduce bass - it's right at your ear.

This would be a nearfield speaker in the truest sense of the word.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.