Optimal listening distance for open baffle speaker

[tommydoan84]

hi all,

Recently, I have tried out my 1st OB DIY project. Below is my system
- Full Range driver: SCANDINAVIAN AUDIO LABS SAL 08C08 SAL driver 6.5'' in 8'' frame
- Sub: Eminence Beta 15''
- Active Xover: miniDSP 2x4HD
- Amp for FR: PASS FW Aleph J
- Amp for Sub: PASS F4 (dual mono class A current amplifier)

My earlier attempt to OB turned out a complete failure. The sound is so thin and peaking on the highs. Bass is very shy and thin.

Not giving up on my project, I asked many questions to the forum and the original designer of SAL driver. I tried out many things, different baffle size/thickness, placement of FR driver at the baffle, different angle and placement of OB, material (MDF, wood), different style for sub (Dipole, U frame, H frame). Finally I'm getting close to a real success!!

After many trials and errors, experimenting, auditioning, modifying Xover points, tuning in different filter/slopes, volumes for each drivers, and lots, lots of saw dust, I finally hit the spot of which the sound becomes ALOT fuller, transparent, more weight, and dynamic.

However, not until last night, I suddenly struck GOLD when I did a random experiment and WoW the difference to the sonic quality is DAY and NIGHT!! What I did is I moved my listening position further back to 20 ft apart from the speaker placement instead of 10 ft as my original listening position.

I don't know if words can describe my experience last night. All i can say is the music is so ALIVE, the speakers really disappear in the back ground and all that was left in the room is just the "music". The sound is so full, transparent filling up the entire room and hitting directly to my eardrums with every distinct notes of every instruments and vocals. The bass is so natural and full. I don't even need to use any EQ, bass boost on the sub. The lowbass, midbass, mid range, and highs are INCREDIBLE. Feels like I rediscover music all over again for the first time!! This is the 1st time that I think DIY experience really brought me to my knees. I did build alot of class A amps, preamps in the past, it does improves my system by many folds but none has brought back the experience that I had from last night.

Sorry for the long message, the key take-away and latest discoveries for my music epiphany is moving away about 20ft away from the speakers.

My questions to all with alot more experience than I ever had is there a general rule or technical reasons behind for OB speaker placement and more importantly, listening distance from the listener away from the speaker to really achieve BEST optimal sound from OB speakers ? For me it's 6ft from speakers to back wall and 20 ft away from listener to the speakers. Is there other workaround to sit around 10ft away from speaker and still achieve optimal sounding as with 20ft away ?

Please see the picture for my current OB project. I'm using H-frame sub and a separate solid wood panel for FR. The speakers are placing 6ft away from the back wall.

- 1st picture is my room size, speaker placement and listening position
- 2nd picture shows how far I sit from speaker (10ft)
- 3nd picture shows how far I sit from speaker (20ft) to achieve the optimal sound!

Thanks,
Tom
1st pix

2nd pix

3rd pix (Ultimate SOUND!!)

 

[CharlieLaub]

Perhaps at position 2 there are one or more room modes that help to offset response problems. You position 1 was almost exactly in the middle of the room, and could have been in a room response null. There are programs that help you find where room modes exist and plan accordingly.

Post some measurements and explain in detail how you have used your miniDSP unit to effect crossover and frequency response shaping/EQ/etc. You could also post an in-room measurement at your listening position.

You could also try to keep seating position 2 but move the loudspeaker system forward to about where your old position 1 was located (including the subs). Why are your subs behind the main panels?

 

[CharlieLaub]

Also, I don't know what the designer told you, but the loudspeaker would be much better if implemented as a 3-way with a much narrower baffle and the SAL right at the top or even with the frame protruding past (above) the top edge of the baffle. A 12" or 15" high efficiency driver can be used between where the H-frame response is dropping into the frame's resonance and where the SAL is rolled off too far.

A small fullrange driver in a large baffle does not make a good OB speaker. It's the mistake I see OB builders making over, and over, and over, and over....

 

[scottjoplin]

Have you seen this Charlie? New OB Project: The Djinn

 

[tommydoan84]

hi Charlie,

Thanks for your response. Any feedbacks are welcome. I will find some time take some in room measurement with REW/UMIK and miniDSP. I do not use any EQ for Sub and FR. I only use Xover at 400 mhz LR24 db/oct and adjust the output volume of sub/FR for them to blend nicely.

What do you suggest for the baffle size of the 3-way OB speaker ? Are you suggesting a super tweeter with a cap filter on top of the FAST system ? I have not used super tweeter since I don't find a need for it just yet. I have tried many baffle size, small, large in vertical and horizontal and does not get good results. The sound is think and peaking at the highs. Until, I stumble upon this one with H frame. Placements of H frame a bit behind the FR is just my random placement no reasons behind it really.

Of course, one can go back and forth all days with all theories, simulation and formula, plots, curves as what I had done some in the earlier phase with some of my OB model but none can beat the real experience. I got so frustrated with my earlier trials so I just threw all the numbers and plots away and trying different ideas/combination to pick out ones that catch my ears attention. H frame sub sounds the BEST to me in term of LF. I used to combine both Sub and FR in the same baffle with lots of EQ on LF and HF but does not get good result until i tried last combination as you saw in the picture of my original post. I just use Xover at 400Mhz and filter of LR-24db for Sub and FR and that's it.

Below are the few models that I have tried and does not give me any good results

Regards,
Tom

Also, I don't know what the designer told you, but the loudspeaker would be much better if implemented as a 3-way with a much narrower baffle and the SAL right at the top or even with the frame protruding past (above) the top edge of the baffle. A 12" or 15" high efficiency driver can be used between where the H-frame response is dropping into the frame's resonance and where the SAL is rolled off too far.

A small fullrange driver in a large baffle does not make a good OB speaker. It's the mistake I see OB builders making over, and over, and over, and over....

 

[scottjoplin]

Looking at your picture perhaps try moving the subs slightly forward of the SALs and move back to the original listening position. They are too far apart considering your crossover frequency

 

[CharlieLaub]

What do you suggest for the baffle size of the 3-way OB speaker ?

I attached a pic of how I suggest you consider reinventing the OB as a 3-way (H-frame plus baffle with lower 12" and upper SAL drivers). The "X" means remove this part of the baffle, and would produce a wider lower part with a narrower "arm" sticking up where the SAL driver would be mounted. The "arm" does not need to stick up as far as I have drawn, but that would be OK if it did. Aim for 35"-40" height at center of SAL driver, which is about ear height when seated. You can't implement the 3-way with a single miniDSP 2x4 and you would need another 2 channels of amplification. You probably would want to support the magnet of the SAL with a fin perpendicular to the baffle, since the arm is barely wider than the driver (the narrow baffle is intended) and would not be mechanically strong. If you want to give that approach a try, let me know. It will be much better than the SAL-planar-OB + H-frame.

I also attached a rough guesstimate of the frequency response on axis of the SAL driver in the OB pictured in your first post. Don't pay attention to the high end roll-off above 10kHz, the sim is not accurate in this region. Notice that the response is it VERY non-flat, with a peak at 500Hz, a hole around 1kHz, and falls off below 500Hz. This is what you can expect from an open baffle system like this (on axis). The big hole at 1kHz is a product of the "small driver in a large baffle" construction. You would not have this if you use the modified baffle I described in the last paragraph.

hi Charlie,

Thanks for your response. Any feedbacks are welcome. I will find some time take some in room measurement with REW/UMIK and miniDSP. I do not use any EQ for Sub and FR. I only use Xover at 400 mhz LR24 db/oct and adjust the output volume of sub/FR for them to blend nicely.

This is what I suspected. No EQ means you are not correcting for the very tilted response of the SAL in the OB. Same for the H-frame. You need to do some learning about what the response from dipole systems looks like - it's NOT FLAT AT ALL, so just throwing a LP+HP crossover on there is really not resulting in anything good (as you have found out). It is no surprise at all that you felt the sound was "thin", because it was! Making some measurements with a mic, even some basic ones, can be very helpful and that is what I suggested it.

Even if you wanted to keep the large OB from the first post, it's possible to model the response of the driver in the baffle using software (I guesstimated it and attached the plot of the FR). Once the model is accurate you can devise a ballpark EQ correction that will get the response flattened out properly, to a first order approximation. This will be much less ideal that what you can get if you measure the response of the driver, in the baffle, and then design some proper correction/EQ for it, but it will be better than just a HP/LP crossover like you have now. If you give me the dimensions of the baffle and the driver location on the baffle I can do some quick modeling that will give you an idea of the frequency response that the speaker is producing. The same goes for the H-frame. What is the height and width of the opening, and how deep is each side (baffle to opening on each side)?

Finally, while that Pure Audio Project speaker looks awesome, the MFM type driver arrangement is bad for the fullranger response. It's back to that small driver in the middle of a large baffle kind of approach.

 

[tommydoan84]

Scott,

Tried it. Sounds bad. I think I hit the "magic" number and placement of FR and sub by accident with the baffle size I have and the room size of my space.

I tried to move sub forward higher than FR or move FR 1 feet backward, all the magic, 3D sense has completely lost in both listening positions 1 and 2.

Moved the Sub back to 4ft away from fall and SAL 6ft away from wall and stand away 20 ft from FR, the magic starts to happen again. I think I need to pattern this magic number or something

-Tom

Looking at your picture perhaps try moving the subs slightly forward of the SALs and move back to the original listening position. They are too far apart considering your crossover frequency

 

[tommydoan84]

Thank you so much for the thorough reply and explanation Charlie! I need to take some time to re-read it again to absorb all the new knowledge and suggestion and get back! It's fun to try out different things but at least I get to somewhere at least I know the sound is "RIGHT" for the 1st time after many unsucessful attempts. I was trying to EQ FR curve earlier by measuring SPL curve on REW with miniDSP and Umik ohmi mic to get as flat response as possible but someone told me flat curve with EQ does not mean to have good sounding and 3D! Since then I throw all EQ away and mostly use my ears to make judgement

Thanks again for the reply Charlie. Appreciate the input
Regards,
Tom

I attached a pic of how I suggest you consider reinventing the OB as a 3-way (H-frame plus baffle with lower 12" and upper SAL drivers). The "X" means remove this part of the baffle, and would produce a wider lower part with a narrower "arm" sticking up where the SAL driver would be mounted. The "arm" does not need to stick up as far as I have drawn, but that would be OK if it did. Aim for 35"-40" height at center of SAL driver, which is about ear height when seated. You can't implement the 3-way with a single miniDSP 2x4 and you would need another 2 channels of amplification. You probably would want to support the magnet of the SAL with a fin perpendicular to the baffle, since the arm is barely wider than the driver (the narrow baffle is intended) and would not be mechanically strong. If you want to give that approach a try, let me know. It will be much better than the SAL-planar-OB + H-frame.

I also attached a rough guesstimate of the frequency response on axis of the SAL driver in the OB pictured in your first post. Don't pay attention to the high end roll-off above 10kHz, the sim is not accurate in this region. Notice that the response is it VERY non-flat, with a peak at 500Hz, a hole around 1kHz, and falls off below 500Hz. This is what you can expect from an open baffle system like this (on axis). The big hole at 1kHz is a product of the "small driver in a large baffle" construction. You would not have this if you use the modified baffle I described in the last paragraph.


This is what I suspected. No EQ means you are not correcting for the very tilted response of the SAL in the OB. Same for the H-frame. You need to do some learning about what the response from dipole systems looks like - it's NOT FLAT AT ALL, so just throwing a LP+HP crossover on there is really not resulting in anything good (as you have found out). It is no surprise at all that you felt the sound was "thin", because it was! Making some measurements with a mic, even some basic ones, can be very helpful and that is what I suggested it.

Even if you wanted to keep the large OB from the first post, it's possible to model the response of the driver in the baffle using software (I guesstimated it and attached the plot of the FR). Once the model is accurate you can devise a ballpark EQ correction that will get the response flattened out properly, to a first order approximation. This will be much less ideal that what you can get if you measure the response of the driver, in the baffle, and then design some proper correction/EQ for it, but it will be better than just a HP/LP crossover like you have now. If you give me the dimensions of the baffle and the driver location on the baffle I can do some quick modeling that will give you an idea of the frequency response that the speaker is producing. The same goes for the H-frame. What is the height and width of the opening, and how deep is each side (baffle to opening on each side)?

Finally, while that Pure Audio Project speaker looks awesome, the MFM type driver arrangement is bad for the fullranger response. It's back to that small driver in the middle of a large baffle kind of approach.

 

[CharlieLaub]

Scott,

Tried it. Sounds bad. I think I hit the "magic" number and placement of FR and sub by accident with the baffle size I have and the room size of my space.

I tried to move sub forward higher than FR or move FR 1 feet backward, all the magic, 3D sense has completely lost in both listening positions 1 and 2.

Moved the Sub back to 4ft away from fall and SAL 6ft away from wall and stand away 20 ft from FR, the magic starts to happen again. I think I need to pattern this magic number or something

-Tom

Tom, don't take this the wrong way but:

When you don't understand how something can happen: MAGIC

When you understand how something can happen: SCIENCE

Learn how your OB system is functioning and measure the response it is producing. Then take a scientific approach to making your loudspeaker sound good by designing a customized EQ curve. You will then be able to enjoy the sound from your speakers all over your room, not just at some "magic" position.

 

[cspirou]

Siegfried Linkwitz talks about this with respect to his speakers

Room Acoustics

 

[Tweet]

Tom, this might be of some supplementary help to you as well as Charlie's advice.

Measurements are an integral part of loudspeaker design otherwise it is mostly a hit and miss affair with a touch of theory mixed in to confuse the issue. It is indeed a reality check because the room, baffle size, baffle placement of drivers, distance from walls, seating distance, room furniture, carpets, ceiling height, floor and windows, open doors, etc, etc all contribute to the acoustic environment in which you hope to design your loudspeaker. You may have noticed that changing your listening environment can radically change the sound of a loudspeaker, that's how dependent a loudspeaker is on the space around it and how compromised it is in its placement. It's a learning curve for us all and it is not the easiest of subjects to absorb with all its intricacies, twists and turns. Not to be preachy but loudspeaker design can be a pain in the bum.

C.M

Have a look at this site............

http://www.hifizine.com/2010/12/prototyping-4-way-open-baffle-speaker-with-the-minidsp-2x4/

 

[Tweet]

Tom, note there are three parts to the referred article, see the top third paragraph of the opening page of the link provided. I hope this a help to you.

C.M

 

[tommydoan84]

Thank you so much for the pointers, Tweet and cspirou. More homework for me to read up! Wish I knew DIYaudio earlier way back when I had all the time I need on my hand! It's hard to keep up with work, life, family, and hobbies balances at the same time these days

Regards,
Tom

Tom, note there are three parts to the referred article, see the top third paragraph of the opening page of the link provided. I hope this a help to you.

C.M

 

[tommydoan84]

Charlie,

I just got some free minutes tonight to read your post again in more details. Your suggestion sounds good and very tempting to me. I'd like to give it a try and see how it much better it compared to my current system. I understand for 3 way, I need two 2x4 miniDSP. I was waiting for Nelson Pass B5 Analog xover DIY kit to come out to build and replace the current 2x4 HD miniDSP to see if it improves my current setup. To be honest, i'm very happy with what I have now. I completely understand where you're coming from Charlie and I don't feel offended at all. I'm an engineer myself so anything and everything needs to have a solid reason behind with proof, number, data and explanation. It's my day to day job.

This is only my personal take. Please only take it with a pinch of salt and please don't feel offended Charlie I personally think music has a much higher sense than just number, data theory, simulation, and science according to my own experiences. We're talking about physical sound wave, air moving energy, and human mind psychological interaction here. I know some system has perfect response curve, distortion number and sounds mediocre where as some system has ugly measurement but does sound really good! That's the reason why some people are drawn to tube sound even though their measurement are way off chart for a reference "good sounding" system. I know it does not make sense with engineering but sometimes "good" is "good" and it's beyond words to describe. It's somewhat like religious enlightenment. All religions have the same universal language, once a soul is awakened, you are simply awaked. Plain and simple. You'll see through everything as they way it happens no more no less but there's no science, data, plots to prove that you are really awaked only you who can see and know. It's like life is nothing but a dream and you are suddenly awake from that dream and free from all the suffering, pain and sorrow, etc. Okay enough with this I can go on and on about this topic. It used to be part of my hobbies in the past!

I will definitely find some time to get a response curve with my listening position 1 and 2 with my current system to see what the response look like. I used to do it in my earlier phase but I'm tired of it after a while! However, i think it will help with the new experiment that we're doing here.

I wanted to try out your approach to see how much improvement it makes with my "accidentally sounds good" system that I have just stumbled upon.
My H sub measurement is in the speaker placment sheet in my first post. It's 18''x18''x16'' (H x W x Depth) I followed this number from Quarter wave paper from Martin http://www.quarter-wave.com/OBs/U_and_H_Frames.pdf

The baffle size that you were drawing with an X has height of 40 inch, width 20 inch, thickness 3/4'' plywood. I personally think this baffle is small and too thin not sturdy enough. Do you think I show find something that has more thickness ? like 1 inch or 1 1/2 inch ? what size of baffle should I try with your approach ?

My current baffle has 1inch 1/2 thick solid wood core, it's very heavy and sturdy. I also added concrete block at the base to keep it from vibrating

Thanks,
Tom

I attached a pic of how I suggest you consider reinventing the OB as a 3-way (H-frame plus baffle with lower 12" and upper SAL drivers). The "X" means remove this part of the baffle, and would produce a wider lower part with a narrower "arm" sticking up where the SAL driver would be mounted. The "arm" does not need to stick up as far as I have drawn, but that would be OK if it did. Aim for 35"-40" height at center of SAL driver, which is about ear height when seated. You can't implement the 3-way with a single miniDSP 2x4 and you would need another 2 channels of amplification. You probably would want to support the magnet of the SAL with a fin perpendicular to the baffle, since the arm is barely wider than the driver (the narrow baffle is intended) and would not be mechanically strong. If you want to give that approach a try, let me know. It will be much better than the SAL-planar-OB + H-frame.

I also attached a rough guesstimate of the frequency response on axis of the SAL driver in the OB pictured in your first post. Don't pay attention to the high end roll-off above 10kHz, the sim is not accurate in this region. Notice that the response is it VERY non-flat, with a peak at 500Hz, a hole around 1kHz, and falls off below 500Hz. This is what you can expect from an open baffle system like this (on axis). The big hole at 1kHz is a product of the "small driver in a large baffle" construction. You would not have this if you use the modified baffle I described in the last paragraph.


This is what I suspected. No EQ means you are not correcting for the very tilted response of the SAL in the OB. Same for the H-frame. You need to do some learning about what the response from dipole systems looks like - it's NOT FLAT AT ALL, so just throwing a LP+HP crossover on there is really not resulting in anything good (as you have found out). It is no surprise at all that you felt the sound was "thin", because it was! Making some measurements with a mic, even some basic ones, can be very helpful and that is what I suggested it.

Even if you wanted to keep the large OB from the first post, it's possible to model the response of the driver in the baffle using software (I guesstimated it and attached the plot of the FR). Once the model is accurate you can devise a ballpark EQ correction that will get the response flattened out properly, to a first order approximation. This will be much less ideal that what you can get if you measure the response of the driver, in the baffle, and then design some proper correction/EQ for it, but it will be better than just a HP/LP crossover like you have now. If you give me the dimensions of the baffle and the driver location on the baffle I can do some quick modeling that will give you an idea of the frequency response that the speaker is producing. The same goes for the H-frame. What is the height and width of the opening, and how deep is each side (baffle to opening on each side)?

Finally, while that Pure Audio Project speaker looks awesome, the MFM type driver arrangement is bad for the fullranger response. It's back to that small driver in the middle of a large baffle kind of approach.

 

[Juhazi]

tommydoan84, looks like you have a nice hobby!

But if you want to spend more time with you family and/or just listening to good sound in next years, please study and follow the fundamental things in last few posts!

p.s. Looks like matevan at Audiocirle likes the long and winding road too...

 

[tommydoan84]

Thanks for the feedback Juhazi.

Charlie,

Shouldn't we defeat the purpose of a full range driver for a 3 way system ? I think the reason why I chose SAL FR driver and open baffle is to avoid a 3 way system and the enclosed box design from the start. I wan to experienced something different with OB approach. I already had a JM Focal lab 3 way speakers in my system for a while that I'm very familiar with its sonic characters

SUB is an added support for FR open baffle due to nature lack of bass which I already found a cure for with H frame sub and the placement that I used earlier. The bass sounds very deep, open, and natural to my ears now. It's not shy at all as what I heard before. I think they call this approach FAST. I see some people add a super tweeter on top with L-pad or a single cap around 1uf to .33uf to support the high freqs that is missing form some FR driver. With the SAL driver that I had, it's covered enough all the high frequencies that required so far in most of the song that I have auditioned to.

-Tom

 

[Tweet]

Tom, Here are a few ideas for a baffle shape that may help you get pass the usual slab of wood that masquerades as a baffle.
OB gets you out of the same old box sold to us in the shops, into something that allows for artistic creativity in DIY, even though the same laws in physics still apply.
A good OB design simply leaves people somewhat amazed at what they are hearing when they reflect on what they have heard in the past.
I'll get off my soap box now, but here is a goodie in the link below.

NaO

A simple, powered subwoofer in a sealed enclosure working up to 150c/s is another alternative with a small OB baffle design, I use such a design in my TV room. It is the most musical and natural loudspeaker I have yet heard. Placement you will find (and have found) is most important with OB design, it's a tweaker's paradise in getting there (the crossover and all) but when you do you will have accomplished something you've always wanted, a beautiful sound system.

C.M

 

[augerpro]

I'm sure OB have their peculiarities, but mostly I think you just discovered the dominant factor that is the room itself and room vs speaker position vs listener position. Many angles to consider there, but since you have a dedicated room that you can move around in, try this first for fun: Stereo Speaker Placement for Optimum Sound - ecoustics.com

 

[gadut]

Tom,

have you tried without any xo for SAL driver? attached it directly to AJ, then you can fine tune beta 15 to get a nice blend in midbass area. better using plain baffle for SAL and beta 15

I wont expect beta 15 can produce any good low bass, you should get a real subwoofer 18" above with strong motor (high BL) and not too heavy cone, this is mostly found in pro audio subwoofer. then put it in H-frame.

until you try supertweeter, you will not know what have been missing from fullrange driver