Acoustic Panels and Bass Trapping a new lounge

[playdough]

On 30/12/2023 at 5:34 PM, almikel said:

Hi Glen,

 

one great example of an awesome IB enclosure is William Cowan's IB design/build with 8 x 18" drivers:

https://www.cowanaudio.com/ib.html

 

I've been a long-time fan of William Cowan's contribution to DIYers, and I've learned a bunch from his site.

William is clearly a Tom Danley fan - William was building tapped horns and Unity horns a long time ago...both Danley creations...

 

William's tapped horns:

https://www.cowanaudio.com/th.html

 

William's Unity horns:

https://www.cowanaudio.com/unity.html

https://www.cowanaudio.com/finale.html

 

...Danley further developed the Unity horn into the Synergy horn - a 3 way "point source" horn targeted at the pro world (where Danley focusses).

 

Paul Spencer's PSE144 design is an extension/improvement on Danley's Unity design for a 2 way point source horn.

 

Circularly, I understand Lambda Acoustics (who made the unity horn for DIYers but no longer exists) has some connection with Acoustic Elegance, the driver manufacturer...I can't confirm this, but nice if true given I run Acoustic Elegance TD18s under my PSE144s.

 

cheers,

Mike

G'day Mike

HNY Mate, interesting topic of conversation is William Cowen's exploits.

Thanks for posting. the IB (8 x 18's) built into the timber home is interesting although it would rattle the nails out of the boards, honestly

Much better using enclosures (Tapped Horns) and isolating/decoupling them from the house as much as possible would be preferable. 

Actual efficiency of enclosures over IB also tip the scale as well. 

 

Looking toward exploring, designing and building a soffit mounting Unity type multiway horns in future with some form of acoustic/physical connection to the bass. Finding the larger the horn area/apeature, closer to omnidirectional frequency transition, the better. Luckily I have a room that lends it'self to this theory. 

I'm a dreamer   

 

 

Regards

Matt  

 

 

Edited January 2, 2024 by playdough

[davewantsmoore]

On 2/1/2024 at 10:35 AM, playdough said:

Probably need to be dragged into DL as my experience with "auto EQ and room correction" haven't been the best

It can't do anything about the underlying speaker.... so the speaker needs to be well designed, and the target curve you're asking for needs to be sensible in the context of the speakers underlying directivity / diffraction.

 

On 2/1/2024 at 10:35 AM, playdough said:

Made for those not proficient in understanding of their Room Acoustics and how it effects the HiFi System outcomes at the listening position.

But this is exactly what DL does.   (By comparing a whole bunch of measurements, and computing mixed phase correction).

 

Of course, if you have a "preference" that ties in with the specific room, then it's up to you to adjust the target curve to massage the result.

 

Just a tool of course, but...

[playdough]

Really dig what this Bloke is up to

Costs money although sometimes in doing that, is a fast track.

https://josephcrowe.com/products/video-training-series-crossover-design?fbclid=IwAR1d8fN7voO1JCPZx8MpVf9M2_nvnCvdjCgidSVsxgrhxxScj_Isjt2yU3g

[playdough]

1 hour ago, davewantsmoore said:

t can't do anything about the underlying speaker.... so the speaker needs to be well designed, and the target curve you're asking for needs to be sensible in the context of the speakers underlying directivity / diffraction.

Indeed Dave. HNY BTW

Speakers to this point are fairly well known *via years of being measured  , in this case controlled directivity/aperture horns only.

I'm unsure if diffraction has much of a part to play, I put it into the baffle step category and a very minor part of the overall reproduction. I tend not to overdramatise these points, particularly with open horns.

 

In the case here there is not really a baffle speaker for higher frequency in service.

There may be with the 21's as the point of omnidirectionality is reached to play in with the PSE 144 *crossover @ 250Hz, not so with the JBL 4732 *Crossover @ 120Hz 

Frequency response is reasonable and can be tweaked into a few dB either way of the target EQ, with a moving mic, pink noise and REW tone sweep

2 hours ago, davewantsmoore said:

But this is exactly what DL does.   (By comparing a whole bunch of measurements, and computing mixed phase correction).

 

Of course, if you have a "preference" that ties in with the specific room, then it's up to you to adjust the target curve to massage the result.

 

Just a tool of course, but...

The new power amp rig/DSP is going together, had a few hours yesterday and may be able to test power and prove operation of all the 8 power amp channels, later on.

As usual its very odd approach (that would horrify some) but the theory has been tested for quite some time now, just a full expansion of the approach.

Might buy in the "DL tool" later.

Preference is to ensure the hardware connections, interconnects, grounding and everything else that is hand made to be correct and working properly, first. The single ended system can be a challenge and with a multi channel even harder, correct grounding/earthing, is vital.

 

Finding the old fashioned moving calibrated mic test set up, used  in conjunction with REW and UMIK 1 to be extremely effective and powerful set of tools and may expand on that later. There is not much that cannot be done with these tools.

 

Personally just like to work out at what point to use the auto EQ tool after every other variable of the system/room is known and understood, in full including the Lounge it resides.

Getting there, slowly.

 

playdough.

 

 

 

[davewantsmoore]

28 minutes ago, playdough said:

baffle step category

 

Diffraction around the speaker driver (or speaker cabinet, or speaker horn) is the trickiest thing to get right in any speaker (IMVHO).

 

 

 

[playdough]

2 hours ago, davewantsmoore said:

 

Diffraction around the speaker driver (or speaker cabinet, or speaker horn) is the trickiest thing to get right in any speaker (IMVHO).

 

 

 

Yes, may be so with little speaker systems, at and around the tweeter/HMWoofer in a baffle mounted 2/3 way. The toys around here are not really in that category, not to say it can't happen.

 

 

https://www.tolvan.com/edge/help.htm

 

https://www.researchgate.net/profile/Dirk-Schroeder-4/publication/258098679_Real-Time_Hybrid_Simulation_Method_Including_Edge_Diffraction/links/54fd4d310cf270426d11c397/Real-Time-Hybrid-Simulation-Method-Including-Edge-Diffraction.pdf

 

https://www.aes.org/e-lib/browse.cfm?elib=5952

 

Some light reading on the matter 

[davewantsmoore]

42 minutes ago, playdough said:

Yes, may be so with little speaker systems

In some respects the difficulty rises with the size of the speaker

[playdough]

1 hour ago, davewantsmoore said:

In some respects the difficulty rises with the size of the speaker

Might be worth expanding on that Dave.  Always listening to any information regarding step/edge diffraction, especially yours !

 

I'm thinking, longer the outside edge length dimension with a flat surface baffle, the lower the frequency of anomaly transition.  

 

In saying that, keep in mind the magnitude of the anomaly will be lesser with more distance relative across the baffle to the drive element.

The lower the frequency, the more likely it would fall into frequencies that are highly interlegible by the human ear. 

 

There are a myriad of speaker baffle 3 dimensional shapes and types, all with their own way of dealing with dispersion and subsequent diffraction. 

Everything from loading the driver (front and rear) to diffracting edge of the driver enclosure. Mind boggles actually human ingenuity at work  all the speakers invented.

 

Obviously it's a complex topic and would like to hear your take on it. 

 

 

 

[davewantsmoore]

8 minutes ago, playdough said:

Might be worth expanding on that Dave.

As the dimensions rises, the frequency it occurs lowers.

 

The lower frequency is harder to measure accurately, and trends towards the sorts of ranges where you also have reflections from floors or walls.

 

8 minutes ago, playdough said:

I'm thinking, longer the outside edge length dimension with a flat surface baffle, the lower the frequency of anomaly transition.  

No... not the edge length (edge length is not directly relevant).    The dimension of the "baffle" (ie. width, height).

 

8 minutes ago, playdough said:

In saying that, keep in mind the magnitude of the anomaly will be lesser with more distance relative across the baffle to the drive element.

A wider baffle just moves the diffraction lower in frequency.   It doesn't make it any more or less.

 

8 minutes ago, playdough said:

Obviously it's a complex topic and would like to hear your take on it. 

Can't be avoided.

 

But avoiding unwarranted wiggles in the "power response" can be done my marrying the speaker directivity (baffle step/diffraction is just another form of "directivity") up with the "direct response" (ie. 0-degrees, or "listening window", or the "target curve" in a room correction system like DL).

 

The point (from 10 posts ago) being that selection of target response depends on the speaker directivity.... and also whether that directivity is well designed.

 

Fighting against the speaker design (whether that be a good speaker design or a bad one) when selecting a target curve, will not improve the sound.    Any "didn't get a good result" with something like DL, should be viewed as a "misused the tool", as opposed to a reflection of some inherent flaw with the tool.

 

 

[playdough]

Making cables using some nice hardware, Teflon coax/Reen RCA and Olex PVC/Neutrik.  No fakes and 8 Channels to make.

Edited January 7, 2024 by playdough

[playdough]

Went back and re did those solders after doing 18 more. Re wet an re positioning closer to the plug a mm.

Looks nice. Almost time for a "stereo outage" better get a radio going.

The amps run almost cold at idle, its' all had a good time soaking.  

Might go looking for a pair of normal 4 channel Power Amps later on, if these are  no good after 25 years under the bed. Used them as a Lad to compete in Car Audio. Have using similar for a while runs sweet.

Switching over to the PSE144 and miniDSP Flex 8 today + power amp 4 channels x 100wRMS ea. Each 21" driver will have it's own channel.

PSE will have another 4 channels. 

 

Nearly ready to plug it all in, after a clean up looks like an Electrical worksite.

Devise filtering @davewantsmoore  in room.  Fair platform to work with.

 

 

 

 

 

[playdough]

Seems to work. The miniDSP Flex 8 is dead quiet, nothing odd going on there straight off the bat. Ink black noise floor.

Running USB straight from the PC, works well with no other converters/boxes (Winner less is more) 

Found it easy to set up, intuitive and less janky than it's predecessor. Routing of the incoming signal to left and right no fuss. Would prefer a night mode for the interface, being picky.

No nasties at all, which was a relief, so the job went on

 

Pink noise and the moving mic to set the gains on the 8 channels after polarity checks done at length to the point of disassembly, using a AA battery on individual drivers and as pairs of drivers with the PSE mid woofers. Confirm actual direction of driver diaphragms, all correct.

 

Basic filters advanced into a true 4 way after some listening to music.

The far right and far left 21's, 0 to 225 Hz, R and L 21's, 0 to 70 Hz.

The 21's were creating some combing at frequencies above 120Hz, fixed, which solidified and nice centre image.

Mids in the PSE, 225 Hz to 1200Hz

CD's 1200Hz to 20KHz

Set a normalised EQ for all the 21's with the boost for sub F. 

The pair of additional bass amps go ok, sweet.

 

So, the PSE144 is in service after all that, making magic although not quite  perfect and will need some fine EQ work as things move forwards.

 

Comparison to the JBL 3 way Theatre Screen HF ??? 

Too early to tell

 

playdough

ps. never be intimidated with setting up filters/EQ in room with an 8 channel active, it's a normal thing.

REW Measurements to come.

 

 

 

[playdough]

mmmm, wondered why no magic, one of the compression drivers was down 6dB.  Removed it and it's a bit wiffy cooked ferrofluid.

Another is on it's way  , cheers @Paul Spencer

 

Still all good, the driver works, accurately and just crank it up to play at the same magnitude, fixed. 

 

Now there is that famous PSE144 solid holographic centre image, ahhhh  . Yea good, just have to rely on memory of the JBL now. Start in REW.

 

The 3 Alpine power amps draw 1.5A fairly steadily, each at 14.9v. 64.8watts total, plus 10w into the DSP, is a miserly 8 channels. With monitor and the PC, under 250w so at 30cents a KWh, 7.5cents an hour. Off grid tho. 

The 3 SMPS DC Drivers dip the LED Lighting when switching on all at once. Amps have a remote on switch. 

Little bit of heat, but should improve when the array is vertical. No fans.

 

 

 

 

 

Edited January 8, 2024 by playdough

[jimcan]

I am loving your enthusiasm Glen!

Just curious, you mention your crossover from the PSE's to mid bass is at 250Hz?

Mine are 350Hz, 24dB. I do use a Paul Spencer tuned passive between the CD and 4 mid drivers in the horn though.

 

It might sound like heresy but I have lately really been enjoying Spotify Premium as my source through these speakers.

The conventional wisdom here seems to be that this source is not good enough, but it works for me

[playdough]

15 minutes ago, jimcan said:

I am loving your enthusiasm Glen!

Working my way through it. One driver at a time. Gain testing to the mm with the mic sorts things out.

17 minutes ago, jimcan said:

Mine are 350Hz, 24dB. I do use a Paul Spencer tuned passive between the CD and 4 mid drivers in the horn though.

Yea, dunno, but set up a 250Hz high pass on a 48dB roll off. Measurements with the RTA , seems ok no hole in the response so far, no really wiggly jagged graph  with pink noise.

 

REW mic next, sort out some things. Tone sweeping.

Still sorting gain structure, but love the fine detail in the image. PSE/JBL, both sound similar but image differently. I love both the same, both excellent. The JBL 4732 is a Commercial tool, it works, it's big, sounds big and fat. The PSE 144 is an esoteric artwork in sonic excellence, it's all about image quality and the PSE is more faithful, particularly in image depth and holographic ability.

Narrower in dispersion than the JBL, improved image focus. 

A fair amount less interaction with the room, might show up in the waterfall graph. 

36 minutes ago, jimcan said:

It might sound like heresy but I have lately really been enjoying Spotify Premium as my source through these speakers.

The conventional wisdom here seems to be that this source is not good enough, but it works for me

Mate, music is music. As long as it has full bandwidth. It can be done with a mobile, but the data rate is excessive. No problem if on an NBN connection, CD quality is fine. I'm guessing Spotify Premium would stream CD quality. I can hear if the content is compressed. some things online sound ok, some not so.

I'm a lover of the CD;s are my favourite ! Playing high resolution audio off the local network or straight from SSD. Normally sounds better than recordings played online. Not against streaming music at all, youtube surfer.

Cheap as chips at flee markets, cd's, X Generation Heaven.

 

[playdough]

On 04/01/2024 at 5:52 PM, davewantsmoore said:

Fighting against the speaker design (whether that be a good speaker design or a bad one) when selecting a target curve, will not improve the sound.    Any "didn't get a good result" with something like DL, should be viewed as a "misused the tool", as opposed to a reflection of some inherent flaw with the tool.

Yes agree. Bad speaker designs, like my stand mounting the JBL BiRadials soon show up during listening and ejected  

The 21's and the PSE144 are an uncompromising match up (best I've heard a PSE144), having the 8 channels to drive them is also  a good thing. Some pleasing results so far. Might run the USB mic later on although there is a little hardware job to hide the power amp array first, need a welder, steel and motivation to do the last part.

On 04/01/2024 at 5:52 PM, davewantsmoore said:

The lower frequency is harder to measure accurately, and trends towards the sorts of ranges where you also have reflections from floors or walls.

Depends how low the frequency, generally below 100Hz can get a little tricky, but the moving mic shows it's prowess here, over tone sweeps.

Tricky as in room interaction, yes, but a fundamental response issue will show up as a constant across many measurements and dealt with in EQ or hardware swap out.

On 04/01/2024 at 8:49 AM, davewantsmoore said:

exactly what DL does.   (By comparing a whole bunch of measurements, and computing mixed phase correction).

 

Of course, if you have a "preference" that ties in with the specific room, then it's up to you to adjust the target curve to massage the result.

 

Just a tool of course, but...

Yea it's rather laborious doing it manually although after many different system set ups, it becomes a work process and fairly easy.

 

On 04/01/2024 at 5:52 PM, davewantsmoore said:

But avoiding unwarranted wiggles in the "power response" can be done my marrying the speaker directivity (baffle step/diffraction is just another form of "directivity") up with the "direct response" (ie. 0-degrees, or "listening window", or the "target curve" in a room correction system like DL).

I'll assume power response is actually fundamental response.

Idealy a HF speaker set should be able to maintain constant directivity/polar pattern, to the point of frequency omnidirectionality, case in point the JBL Screen array, PSE144, not quite, but very close.

I did change/update far right and far left filter on the 21's to the inner left and right to play a little voice matching with the PSE. Improved the centre image.  

 

Edge diffraction is to be fair in the case of hardware operated in this thread is not really a thing, if I was building something else, it may be. Overdramatised, back of mind thing.

[davewantsmoore]

2 hours ago, playdough said:

Depends how low the frequency, generally below 100Hz can get a little tricky, but the moving mic shows it's prowess here, over tone sweeps.

Oh no, I mean like below ~600Hz ....

 

It's almost impossible to measure properly in a room unless it is a giant room.

[playdough]

27 minutes ago, davewantsmoore said:

Oh no, I mean like below ~600Hz ....

 

It's almost impossible to measure properly in a room unless it is a giant room.

No, not really. 

Above measurement is in room, no difficulty.

What aspect of measurement are you eluding to Dave ? Some other measurement type, what examples have you had problems ? 

 

Measurements are relatively east and actually very accurate. Interpreting the information is not so easy and learnt through experience. 

I've not experienced the "nigh on impossibility" you describe.

 

 

 

[playdough]

If, those measurements were taken by someone else in another sound lounge, I'd be relatively impressed with what is able to be read from those graphs. Full range , solid fundamental very little harmonic ordered distortion, nice target toll off. Tone sweeping certainly shows a lot of information, the 3rd octave dip led display with dials and buttons is more fun.

[playdough]

Example of in room testing with the RTA. This is a raw (2 way filtered250Hz/1200Hz) PSE144, in room. Obviously a potent EQ session will make the target in this case. 

Edited January 9, 2024 by playdough

[davewantsmoore]

58 minutes ago, playdough said:

Above measurement is in room, no difficulty.

But this isn't the "baffle step" measurement.

 

58 minutes ago, playdough said:

What aspect of measurement are you eluding to Dave?

The "baffle step".   ie. diffraction / directivity.....  that's what we were talking about.

[playdough]

7 minutes ago, davewantsmoore said:

But this isn't the "baffle step" measurement.

 

The "baffle step".   ie. diffraction / directivity.....  that's what we were talking about.

Correct, Paul Spencer might be able to elaborate on that, it's his design.

Thankfully the speaker design is sound, known.

Measuring the dispersion pattern of the horn can be done, horizontal and vertical planes at intervals to generate a 3 dimensional graph. Diffraction, not really easy to measure, bit of a nut to crack there, maybe ?

Probably best done outside, on the side of a hill  

 

Mainly looking at the fundamental at this point. Magnitudes and filters are fairly well set but using no EQ yet, (shown in the RTA measurement) the horns sound reasonable for low level listening, coherent, especially outside the sweet seat and when in the kitchen.  

I did have 2.2uf caps in series, with the compression driver that did  a lot to pre EQ with a passive component at the driver. Almost flattens the >1200Hz response before EQ with DSP. 

The RTA Graph certainly shows the horn loading efficiency from the quad 4" , CDriver and the effectiveness of the DSP Filtering.

EQ will make it better, normalised. Still cleaning up the other set of speakers/rig and have music.  

 

 

[davewantsmoore]

4 hours ago, playdough said:

Correct, Paul Spencer might be able to elaborate on that, it's his design.

But, aren't you doing your own filters?

Both for the PSE? ... and for the crossover from PSE to woofer?

 

The diffraction correction is part of the filters.

 

It's no big deal, likely.... the only point was that it's hard to measure in a room, and that it becomes more and more difficult with larger speakers.

[almikel]

7 hours ago, davewantsmoore said:

But this isn't the "baffle step" measurement.

 

The "baffle step".   ie. diffraction / directivity.....  that's what we were talking about.

@playdough - apologies if this is sucking eggs, but baffle step compensation is used where a woofer is mounted on a flat baffle, and as the frequency drops, the sound starts to "wrap around" the enclosure reducing the energy/SPL of the direct sound, and baffle step EQ boost is applied to cancel out the natural rolloff caused by the baffle dimensions.

Rod Elliot has a good article on Baffle Step Compensation https://www.sound-au.com/bafflestep.htm

 

Baffle Step compensation isn't required for your PSE's - they're not mounted on a baffle.

Looking at the pictures of your room and speakers I wouldn't expect you have baffle step issues with your woofer enclosures that need to be compensated for...room issues will likely outweigh any baffle step issues...

 

cheers,

Mike

 

[davewantsmoore]

12 minutes ago, almikel said:

 https://www.sound-au.com/bafflestep.htm

If you look at Fig 1.

The sphere dimensions (300mm) correspond to ~1Khz.

 

The PSE dimensions (a horn is just an irregular shaped "baffle") correspond to ~300Hz (H) and 600Hz (V).

 

 

That being said, the original point wasn't about the PSEs... just that these things (diffraction from the speaker geometry) was more and more difficult to accurately measure the larger the speaker got.