Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Planars & Exotics ESL's, planars, and alternative technologies

DIY planar magnetic + open baffle woofer array
DIY planar magnetic + open baffle woofer array
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 23rd December 2016, 11:45 PM   #21
Few is offline Few  United States
diyAudio Member
 
Join Date: Apr 2004
Location: Maine, USA
Yes, the reduction of boundary reflections and more uniform SPL with distance were two key reasons why I went with this design. My house has a highly reflective interior and there's a huge range of listening distances that I'd like to cover, realizing things can't be optimized for all distances simultaneously. A very tall dipole was my attempt at a solution. Once the 9' tall Christmas tree is gone I can test my idea of aiming the dipole null at the point along the reflective side wall that would otherwise send a big reflection at the main listening spot.

When I did the conventional ground plane measurements the mic was about 2 m away from the speaker and I tilted the speaker to aim its main axis at the mic. I'm sure I didn't do it very accurately, though. In any case, I ended up with a much noisier impulse response than I was getting with the mic a few feet in the air. Lots more "junk" after the main impulse and before the first clear reflection. That was, of course, the opposite of what I was going for.

The measurements I made using the large wall of windows as a vertical "ground plane" yielded much better results. Less hash after the impulse and the magnitude response was quite free from reflection artifacts above the frequency where room modes are going to dominate no matter what you do.

So maybe I just did a poor job with the ground plane attempt. You commented that you find it most useful below 500 Hz. That's not what I would have expected. The ground plane approach won't eliminate room modes, so I was thinking of it being most useful at higher frequencies where it could buy a bit more reflection-free time after the arrival of the main impulse.

I used to have some Radio Shack PZM microphones that were ideal for ground plane measurements (apart from their unknown frequency response!) but I've just been putting the business end of my mic as close as possible to the flat surface and crossing my fingers.

Few
  Reply With Quote
Old 23rd December 2016, 11:56 PM   #22
mwmkravchenko is offline mwmkravchenko  Canada
Mark Kravchenko
diyAudio Member
 
mwmkravchenko's Avatar
 
Join Date: Jan 2002
Location: Perth Canada
DIY planar magnetic + open baffle woofer array
My comment about ground plane being useful below 500 hertz should have been a little clearer. Useful if done outside as a true ground plane measurement.

You can do wide range ground plane measurements. But it is hard to interpret them when comparing them to other methods.

Like I said. You picked the deep end of the measurement pool to start in.

But your at least floating and not sinking!

The tip I gave you about doing nearfield measurements and then adjusting your farther measurements mic position until you get the closest results id pretty much the way to learn what works best.

And you are completely right in trying to find the place placement within your current listening room. Not so nice for family to be listening to test signals!

I use a test box from Smith and Larson. It can do a variety of measurements using a great deal of different test signals. Like the LMS system it can send a discrete sine wave and then setup a twin t filter to reject any other noise. Great for getting good measurements within a not so great measuring environment. But oh what a pain to be listening to! A 100 plus point or even a 512 point measurement takes forever and gives you awesome resolution in your results. But it is pretty awful to be near it.
__________________
Mark
www.kravchenko-audio.com
  Reply With Quote
Old 24th December 2016, 12:36 AM   #23
Few is offline Few  United States
diyAudio Member
 
Join Date: Apr 2004
Location: Maine, USA
Ah, got it. Yes outdoor ground plane measurements would be great. I live in the rural part of a rural state that'll be snow-covered until mid-April so I'll have to do what I can indoors for now.

My goal is to optimize the miniDSP settings for best psychoacoustic effect in my particular room, rather than to optimize the engineering of the woofers, for example, so I think I can make do with in-room measurements. I just need to accumulate more experience measuring speakers that don't act like point-sources.

Thanks for your advice, Mark! It's helped me clarify my thinking.

Few
  Reply With Quote
Old 24th December 2016, 12:50 AM   #24
mwmkravchenko is offline mwmkravchenko  Canada
Mark Kravchenko
diyAudio Member
 
mwmkravchenko's Avatar
 
Join Date: Jan 2002
Location: Perth Canada
DIY planar magnetic + open baffle woofer array
We have about 14 inches of snow right now. And a big storm on it's way.

No ground plane measurements for a while here either. Can be done if I have to. I lay out a bunch of sheets of plywood and away we go.
__________________
Mark
www.kravchenko-audio.com
  Reply With Quote
Old 24th December 2016, 01:27 AM   #25
Few is offline Few  United States
diyAudio Member
 
Join Date: Apr 2004
Location: Maine, USA
Non-Alaskan Americans shouldn't gripe to Canadians about winter. Should've checked my facts first!

Few
  Reply With Quote
Old 24th December 2016, 01:33 AM   #26
mwmkravchenko is offline mwmkravchenko  Canada
Mark Kravchenko
diyAudio Member
 
mwmkravchenko's Avatar
 
Join Date: Jan 2002
Location: Perth Canada
DIY planar magnetic + open baffle woofer array
A little whine, and a little cheese and we get a party right!
__________________
Mark
www.kravchenko-audio.com
  Reply With Quote
Old 24th December 2016, 03:13 AM   #27
kach22i is offline kach22i  United States
diyAudio Member
 
kach22i's Avatar
 
Join Date: Oct 2009
Location: Michigan
Wow those are tall.

Awesome.
__________________
Tubes and Martin Logan's
  Reply With Quote
Old 24th December 2016, 10:05 AM   #28
wesayso is offline wesayso  Netherlands
diyAudio Member
 
wesayso's Avatar
 
Join Date: Jul 2007
What a lovely looking pair! Beautiful!

If I may add a few thoughts... measuring an array up close is never going to work. You need distance to get the sound to integrate, especially on the bass arrays. Running around the room to get reflection free measurements will also be near to impossible.
What I would do is set them up for your listening position and take measurements there. With these tall arrays I'd like to be at least 3 meter from them.
As you have chosen open baffle, play with toe-in/out and see what it does to reflection levels in your measurements.
Are you using REW? If so, check the impulse for blips after that impulse spike, and check the next tab, the filtered IR set to "dB FS" in the upper left corner.

What you're looking for is the magnitude of the reflections. Now this graph won't tell the whole story unless you use it's filtering capabilities, it will get you started.
It should look something like this:
Click the image to open in full size.
A nice, well defined peak with a big drop immediately afterwards. Try to have those reflections down as far as you can, preferably more than -20 dB for the first ~20 ms. These are measurements from my (sealed) full range line arrays.

Keep in mind that the stuff behind the microphone can also reflect back. You need some space behind the microphone. I didn't have that space behind the listening position and used damping panels to absorb that energy. I also needed a damping panel on a side wall, as I do not have that dipole magic helping me out.

Once you have that lower reflected energy (we still didn't see the lower end, as that would require to use filters in the filtered IR tab) we can see how the measurement looks.

Personally I don't use the normal gating. There's a valid reason for that. A high note only lasts a couple of milliseconds. So everything after that isn't the high end of the speaker anymore. Actually, the high end is already done and you're only adding the room influence and frequency decay into the measurement. It would be different if we had an actual anechoic room to play with.
I use another form of gating, called a frequency dependent window. This gates the measurement very short for high frequencies and gradually longer for longer, lower frequencies. This leaves out way more room influences. The frequency dependent window can be found behind the IR window button, you can set it in cycles or width in octaves. Don't use any gating combined with this.

Be advised REW is smoothing these frequency dependent windows if I compare it to some other tools.

Base your correction, be it PEQ or FIR on a 6 cycle window for starters. Do not use 6 cycles for phase correction though, you would be trying to EQ room positions, that's not going to work in a large area. Start simple and EQ the frequency curve when viewed with that frequency dependent window of 6 cycles. Every plot without that window set will automatically look way better, but more important, your sound will improve.

There are other ways to try and exclude the room in your measurements. There were some excellent examples in the RePhase thread only days ago. Look here for more info:
rePhase, a loudspeaker phase linearization, EQ and FIR filtering tool

I hope this gives you some new alternatives to try. It worked very well for me as another line array builder. Line arrays behave different, your open baffle will behave different from my sealed arrays. You've got to find a way that works for you. Maybe this helps. My thread as linked is full of my personal journey trough measurements and DSP, it may be a fun read for you. It is huge though. Sorry for that .
__________________
Use Science to design your speakers and they will sound like a piece of Art...
  Reply With Quote
Old 24th December 2016, 11:57 PM   #29
Few is offline Few  United States
diyAudio Member
 
Join Date: Apr 2004
Location: Maine, USA
Yes, tall is a good word for them. I had the same reaction when I started assembling the frames to hold the four planar magnetic panels. And then I realized I needed legs and feet, besides. Ugh.

Wesayso:

Thanks for the thorough contribution. I started reading that rePhase thread awhile ago when I began to familiarize myself with the program. I haven't yet gotten very far through it, though, so I'll have to focus my efforts on the description of your adventure.

Also, I had read somewhere that REW has a frequency dependent window but when I didn't see it among the other window options I forgot to follow up on it. Your hints as to where to find it will be very helpful.

My main listening seat is about 9 m from the speakers, as you suggested. I'll need to wait for Christmas to pass before I can experiment more seriously with speaker angle and its effect on side wall reflections. The Christmas tree is standing precisely where the worst side reflection will come from.

I was a little confused by your comments about using filtering in the IR response. Were you suggesting that filters be used or not? I don't have any experience with impulse response filters so that may be the cause of my confusion. Perhaps the thread you linked to will answer my questions. In any case, thanks again for the advice.

On a more general note, my sense of Floyd Toole's suggestions (without going back to read his book and check my memory!) is that some form of reflection-free measurement (such as gating) should be used above around 400 Hz +/-. Below that frequency room modes should be thought of as part of the system's response so ungated measurements are necessary. That was the philosophy I was using as a starting point. I'd be interested to hear whether others have found it to be a good approach or whether there are pitfalls I should be aware of.

Few
  Reply With Quote
Old 25th December 2016, 09:29 AM   #30
wesayso is offline wesayso  Netherlands
diyAudio Member
 
wesayso's Avatar
 
Join Date: Jul 2007
Hi Few,

I was commenting about the filtered IR tab. An IR is a representation of the measured response, a chosen interpretation of it. There's much more info hidden in it than an IR graph shows. It can be dissected to see that info by using filters on it. You can find some on the left bottom in the filtered IR tab. It lets you set a filter to see the data in a narrow filtered section of the IR.

Just like you can gate it to exclude later reflections. Only we get to see a very narrow slice. For instance 8000Hz 1/3. That will show you what's really going on around 8 KHz. If you click in the IR tab next to it, you can even see the cycles of information at that frequency.
In the filtered IR tab it shows you the real attenuation at that frequency, not the interpreted overall representation. Realise it's hard to explain for me as it would be easier if we were sitting next to each other and I could show you this stuff in real time.

Instead of gating before the first reflection I suggested to use the frequency dependent window.

What would happen if you gate an IR before your first big reflection? Say the first big reflection happens at 4 ms. That would mean you'd see about one cycle of usable info at 250 Hz. But lets see how many cycles you have at 1 KHz. Once cycle at 1 KHz lasts 1 ms.
This would mean the 4 ms gated measurement has 4 usable cycles at 1 KHz.

Let's move even further up in frequency, what happens at 8 KHz. One 8 KHz cycle would last 0.125 ms, meaning you could finish 32 cycles of 8 KHz in that 4 ms gated time frame. But REW doesn't use an infinite number of cycles in it's measurement. So actually you are looking at data that shows what happens long after 8 KHz stopped "singing".
So at 250 Hz the speaker just started to sing, at 1 KHz we may see some useful data and at 8 KHz its supposed to be finished long ago. Meaning at this frequency you see what the speaker has done and all of the little stuff that happens afterwards.

If we take Toole's hint, we would need about a 10 to 12 ms of reflection free zone before we see any useful data at 400 Hz (If I want to see about 5 to 6 cycles).

What's the difference with using a frequency dependent window?
At 250 Hz, using 6 cycles the gating window would be 24 ms. At 1 KHz that window would be 6 ms long. And at 8 KHz the window that is used is is only 0.75 ms long.

So look at it as a sliding window. To see the early response of that wave front. You are gradually letting more of the room into the measurement as you go down in frequency. But the correction based on this window is acting where it happens, not long after that wave front stopped "singing".

That blip in the IR we see as a reflection (the reason we gated at 4 ms), at what frequency is that one happening? We would need to look at the spectrogram window to find it there. Or use those 1/3 octave filters again in the filtered IR tab to get a clue. Chances are it's only happening at a higher frequency.
See how the frequency dependent window basically lets you look inside the measurement and shows the early wave front as it happens? (at it's origin) Gradually you let the room into your analysis.
Now if you succeed in my proposed clean up of the IR for the first 20 ms, we would surpass that target Toole mentioned.
If we get a clean first 20 ms, we would have a clean 6 cycles of data down to 300 Hz.
I'll tell you up front, it's not that easy to hunt down reflections, but every bit of clean up you can do there helps. The less DSP is needed to get to our goal.

"The speakers will only sound as good as the room they are in" is often said. So make sure your speakers work with the room.
__________________
Use Science to design your speakers and they will sound like a piece of Art...
  Reply With Quote

Reply


DIY planar magnetic + open baffle woofer arrayHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
DIY planar magnetic DickHertz Planars & Exotics 17 12th September 2014 08:00 AM
help designing a magnetic planar woofer baffleman Planars & Exotics 14 18th April 2014 04:01 PM
Open baffle woofer line array concept. oon_the_kid Full Range 14 19th October 2011 11:02 AM
"Box" design and bracing for open baffle woofer array ryoshu Multi-Way 0 17th August 2008 06:07 PM
DIY Open baffle / Sealed woofer design preiter Multi-Way 11 25th August 2006 05:31 PM


New To Site? Need Help?

All times are GMT. The time now is 12:00 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 15.00%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright ©1999-2017 diyAudio
Wiki