Lii Audio F15 open baffle pics

[Anton Mullet]

,

Edited July 20, 2023 by Anton Mullet

[Anton Mullet]

On 22/02/2023 at 10:53 AM, oztayls said:

Mine are not F15's, but Lii Fast 8s. Baffle is a cross between the winged type shown by the OP and a barrel shape. Well, it was once a barrel after all lol. I found that no baffle step correction was needed. Construction was a tad complex, because the barrel was first deconstructed and then the staves reshaped to form a tighter curve, so it was a bit like making a new barrel from scratch. It was a bit of an experiment as I didn't know what to expect from the use of an old oak shiraz barrel. Was it worth the result? Hell yeah! The best sounding speakers I've every heard. The bonus is that after 4 years, there is still a distinct shiraz aroma in the room as I left the backs raw! They are easily driven by a home-made 2-3W Gerald Curtis 6CL6 amp that runs as a pentode without any negative feedback from the output transformers. The two 6CL6 high-bandwidth pentode video tubes have enough gain in this circuit to take a line level input and amplify it quite respectably. It drives these and the two subs without any problem in what is quite a large room.

 

I've added some old Richard Allen tweeters, upward firing and crossed with a single 2uF cap. They work very well. 

 

I have some W15 woofers, so I plan to make a pair of Betsy  type baffles for those sometime soon. I have a pair of 250W Yung plate amps to drive those. 

I was trying to give feedback on you about the speedy and friendly speaker stuff. Will work it out.

Anyway I saw this and Ive got  a pair of F15's sitting in their boxes for 3 or 4: years now  but still haven't mounted them on a baffle yet. The wine barrel idea is great. It's got me thinking.

[oztayls]

I have some F15s as well Anthony! Also some w15s. We have a dilemma  

[EddieT]

On 2/8/2021 at 1:16 PM, davewantsmoore said:

 

A driver with no baffle gets quiter as the frequency goes down (ie loses low frequency response) ..... as the baffle gets bigger, this effect diminishes.

 

All speakers need some sort of filter to make it play flat (ie. to lift the falling bass response) .... the size of the baffle just becomes part of that correction filter.

 

So if you had a small baffle.... you would need to lift the bass more (using another way, eg. a crossover/response shaping network).... than if you had a larger baffle.

 

So the point is that there will be no "correct" or "optimum" size.

What guy is ever going to admit to having a "small baffle" haha

[davewantsmoore]

27 minutes ago, EddieT said:

What guy is ever going to admit to having a "small baffle" haha

Something about "it's not the size of the baffle, it's what you do with it....."

 

 

Using a big baffle to get a flat(er) frequency response at LF, is a bit like putting bigger wheels on your car to make it go faster.   You could just push the accelerator pedal harder, or have a larger engine, or.....

In the end it's all a balancing act, depending on the goals.... but there are very good reasons why "modern" open baffles are very narrow (basically as narrow as the drivers will allow) or use no baffle at all.

[EddieT]

11 minutes ago, davewantsmoore said:

Something about "it's not the size of the baffle, it's what you do with it....."

 

 

 

hahahahaha Thanks, laughs have been few and far between for quite a while!  the other was also interesting thanks.

Edited September 2, 2023 by EddieT

[MattyW]

I always though it was both about the size of the baffle and what you do with it.

[andyr]

4 hours ago, EddieT said:

What guy is ever going to admit to having a "small baffle" haha

 

Here are my 'zero baffle' spkrs, Eddie:

 

 

 

As @davewantsmoore said:

 

 

 

... hence, with no baffle - I need massive amounts of boost below about 500Hz (plus a cut above) to get the mid/bass units flat.  The miniDSP unit I use to provide the XOs for my active rig also provided this EQ.

 

Edited September 2, 2023 by andyr

[davewantsmoore]

58 minutes ago, andyr said:

massive amounts of boost

People always seem to talk about it in these terms, which I think really paints some totally undeserved connotations.

 

Quote

massive

It's only 6dB/octave more than a normal woofer.

ie. you are only "undoing" the equivalent of a 1st order high pass filter.

 

Quote

boost

For those who think "oh I don't want to be boosting the signal too much, there is something wrong with doing that"

 

We can just imagine the EQ the other way around, eg. if you used the appropriately shaped "low pass" filter to do the correction .... then the EQ is "cut only".  There's no "boost".

What you have, in reality, is a 6dB loss in acoustic efficiency per octave.... that you need to compensate for somehow.

 

1 hour ago, andyr said:

'zero baffle' spkrs

 

One big issue related to the width of the speaker baffle (or just driver, if no baffle) is that we cannot use the driver above the "dipole peak" .... without encountering a large error in the SPL of the driver vs angle (aka. the "directivity").   Which leads to the "narrow baffle" design (to raise the dipole peak).

 

This is what wide open baffle speakers often get very wrong.... although one way to "solve" that is to go to very very wide baffle.

 

For an ~8" woofer, like in the pic (without a baffle).   The frequency limit is about 1khz.

 

 

This is why Linkwitz moved his Orion speaker from a 1" + 8" + 12".....   to a 1" + 4" +8" + 12" (to stop using the 8" too high in Hz).

[davewantsmoore]

Part of the reason for writing any of that, was to make the point that people often consider open baffle speakers to be a "simpler" design, than a box.

 

... but in many ways it is much harder.

[andyr]

 

I don't see any 'negative connotations', Dave - simply the magic of DSP.  

 

 

 

With the 4 mid/bass drivers I ended up with, I used:

• a +2dB LP shelf @ 500Hz
• plus a 5dB LP shelf @ 250Hz
• and a -8dB HP shelf @ 500Hz
• with a shallow 3dB peak @ 500Hz,

Plus:

• a 3dB reduction in the tweeter levels
• and a 5dB reduction in sub levels.

 

1 hour ago, davewantsmoore said:

For an ~8" woofer, like in the pic (without a baffle).   The frequency limit is about 1khz.

 

It's actually a 6 1/2" SB Acoustics 'Satori" driver, Dave.  Its FR is flat to well over 3500Hz - I have the 24dB L-R XO @ 2800Hz.

 

EDIT:  Sorry below this ... is a repeat of some earlier text which SNA is not letting me get rid of!  (Despite several tries.   )

 

Edited September 2, 2023 by andyr

[davewantsmoore]

2 minutes ago, andyr said:

I don't see any 'negative connotations'

People read their own negative connotations into it (ie. into what you said) .... based on "massive boost", and some sort of underlying fear of "boosting the signal a lot".

They shouldn't fear that, of course .... but many people inherently feel that it "must be something bad".

 

2 minutes ago, andyr said:

With the 4 mid/bass drivers I ended up with, I used:

• ...

Sure... this includes a lot of other "stuff" .... where as what I was talking about was the difference between a dipole and a monopole (or an open baffle, and a box).

 

2 minutes ago, andyr said:

It's actually a 6 1/2" SB Acoustics 'Satori" driver Dave.  Its FR is flat to well over 3500Hz

If you look at this chart from SL site... Note that this is just the impact of the "dipoleness".

 

 

You can make the frequency response flat above the peak at 0 degrees (looking at the speaker front on) .... but this EQ won't work (also make it flat) for the response at other angles.   It is a big distortion in a dipole response.

 

So your driver cannot be flat on all axis, this high in frequency.... which is an example of what I was saying about the dipole peak being a problem (to manage / avoid, etc.)

[andyr]

Whatever, Dave ... but the REW graph taken after the miniDSP magic shows the response is what I want - a slight downwards tilt to the right.

 

Made even better to listen to after taking all EQ out of the miniDSP and instead using Roon Convolution (set up by a friend).

 

[davewantsmoore]

11 hours ago, andyr said:

Whatever, Dave ...

You're welcome.  0_o

 

11 hours ago, andyr said:

but the REW graph taken after the miniDSP magic shows the response is what I want - a slight downwards tilt to the right.

Ok.  If you look at the polar plot of the speaker (directivity) you will see that using a driver above the dipole peak, means the response is not flat on all axis.    It is essential to measure/simulate speakers on all axis, as it has an immense contribution to the sound.

 

It's part of the reason why people go for "no baffle" (or small)... as opposed to the traditional wide.   To raise the frequency where this issue becomes a problem.

 

11 hours ago, andyr said:

Made even better to listen to after taking all EQ out of the miniDSP and instead using Roon Convolution (set up by a friend).

Yeah, I had a similar discussion with him too.

 

11 hours ago, andyr said:

Made even better to listen to after taking all EQ out of the miniDSP and instead using Roon Convolution (set up by a friend).

Roon and miniDSP sound identical, when using the same EQ .... and unsurprisingly, different when not.    (Just in case people will get the idea that there is something inherently better or worse about these, when applying the same EQ).

 

 

[Grizaudio]

10 hours ago, andyr said:

Whatever, Dave ... but the REW graph taken after the miniDSP magic shows the response is what I want - a slight downwards tilt to the right.

 

Made even better to listen to after taking all EQ out of the miniDSP and instead using Roon Convolution (set up by a friend).

 

Hi Andy, firstly congrats on giving OB speaker building a go.

 

….. are you able to share free space (outside) on and off axis system response for your speaker @ say 10deg increments? I.e Raw system response with only crossovers/crossover EQ applied. 
 

Did you manage to get measurements happening after we PM’d? 

[Grizaudio]

5 minutes ago, davewantsmoore said:

Ok.  If you look at the polar plot of the speaker (directivity) you will see that using a driver above the dipole peak, means the response is not flat on all axis.    It is essential to measure/simulate speakers on all axis, as it has an immense contribution to the sound.

Great advice 

[davewantsmoore]

 

5 minutes ago, Grizaudio said:

10deg increments

 

It's likely taking this thread too far off topic.

 

All dipoles have the characteristic I described (so it will be there too) ... and it's just a balance in the design between "wide baffle" and "no baffle".    The dipole peak is a nasty wrinkle in the sound power that you don't get in a box speaker..... and so moving it out of the way (narrow baffle) is almost essential ..... where as the benefit of the old school "big baffle" way (get more bass) can be handled other ways.

 

1 big driver (even on a relatively small baffle), doesn't really address the issue, as it's still too big for WL > 500Hz .... but as far as 1 big driver on a OB goes, there's nothing that can be done.

Edited September 2, 2023 by davewantsmoore

[Grizaudio]

12 minutes ago, davewantsmoore said:

Roon and miniDSP sound identical, when using the same EQ .... and unsurprisingly, different when not.    (Just in case people will get the idea that there is something inherently better or worse about these, when applying the same EQ).

Roon is far more powerful. Not saying necessarily different sounding. 
 

……but different sounding can happen, when running low frequency FIR correction. 
 

If correction is being applied using FIR, Roon will easily process accurate high TAP counts at low frequency, whereas MiniDSP potentially won’t have the TAP count available which is especially true for 20-30hz accurate FIR a correction etc. 

Edited September 2, 2023 by Grizaudio

[davewantsmoore]

7 minutes ago, Grizaudio said:

Roon is far more powerful.

Yes, it can do EQ that miniDSP mostly can't.

The point was more simply different EQ will sound different ... and the same EQ will sound the same (whether implemented on Roon, miniDSP or something else).

 

7 minutes ago, Grizaudio said:

……but different sounding can happen, when running low frequency FIR correction. 
If correction is being applied using FIR, Roon will easily process accurate high TAP counts at low frequency, whereas MiniDSP potentially won’t have the TAP count available which is especially true for 20-30hz accurate FIR a correction etc. 

We have done this dance before....  people seem to wildly misunderstand the implications what you've (and Andy) just stated.

 

If you create your mixed/linear phase LF filter and apply it with Roon (which miniDSP, for example) cannot do .... and then you create a different filter in miniDSP that achieves the same SPL.  (ie. the result is matched very closely).   They will sound identical.    The point being that the phase part of the filter is totally unimportant, aside from the impact it has on the frequency response.

 

You can read 3 feet of text books to convince yourself that what I've said here must be true .... but it's much more intuitive and practical to just do the simple listening tests, and you will see that it is (unsurprisingly) not audible.

 

Eg. calibrate system to linear frequency response, and linear (low frequency) phase response

Add filter to system, which distorts phase (only) back to "minimum phase" (or whatever) response.

Measure outputs to ensure no mistakes have been made, etc.

Turn filter on and off while listening.

[Grizaudio]

7 minutes ago, davewantsmoore said:

Yes, it can do EQ that miniDSP mostly can't.

The point was more simply different EQ will sound different ... and the same EQ will sound the same (whether implemented on Roon, miniDSP or something else).

 

We have done this dance before....  people seem to wildly misunderstand the implications what you've (and Andy) just stated.

 

If you create your mixed/linear phase LF filter and apply it with Roon (which miniDSP, for example) cannot do .... and then you create a different filter in miniDSP that achieves the same SPL.  (ie. the result is matched very closely).   They will sound identical.    The point being that the phase part of the filter is totally unimportant, aside from the impact it has on the frequency response.

 

You can read 3 feet of text books to convince yourself that what I've said here must be true .... but it's much more intuitive and practical to just do the simple listening tests, and you will see that it is (unsurprisingly) not audible.

 

Eg. calibrate system to linear frequency response, and linear (low frequency) phase response

Add filter to system, which distorts phase (only) back to "minimum phase" (or whatever) response.

Measure outputs to ensure no mistakes have been made, etc.

Turn filter on and off while listening.

I don’t disagree, especially if you are SPL frequency matching. I was just highlighting there is a capability difference between systems, and if one was trying to apply complex FIR using minidsp it won’t happen. Some minidsp will perform FIR but will lose accuracy vs a high Tap solution. 
 

…… so you only have the option to absolutely match response. 
 

I suspect this is where the audible difference lies. I.e filters are different. 

Edited September 2, 2023 by Grizaudio

[andyr]

2 minutes ago, davewantsmoore said:

 

Roon and miniDSP sound identical, when using the same EQ .... and unsurprisingly, different when not.    (Just in case people will get the idea that there is something inherently better or worse about these, when applying the same EQ).

 

 

I have no doubt you are correct Dave; the key point is "the same EQ".

 

My miniDSP nanoDIGI is not powerful enough to do more than just IIR filters.  Whereas the PC which runs Roon ... uses 32k taps.

 

33 minutes ago, Grizaudio said:

Hi Andy, firstly congrats on giving OB speaker building a go.

 

Thank you Griz.  

 

After having forced my wife to have to put up with large Maggie panels for 25 years ... I wanted something with minimum visual impact!  

 

33 minutes ago, Grizaudio said:

….. are you able to share free space (outside) on and off axis system response for your speaker @ say 10deg increments? I.e Raw system response with only crossovers/crossover EQ applied. 

 

No sorry ... I don't have this data.  (I don't have any garden!   )

 

33 minutes ago, Grizaudio said:

Did you manage to get measurements happening after we PM’d? 

 

No, I'm afraid I haven't managed to get any believable REW measurements.  

 

[davewantsmoore]

16 minutes ago, andyr said:

No sorry ... I don't have this data.  (I don't have any garden!   )

You could look at this (> 500Hz) in a room , with reasonable resolution.

 

25 minutes ago, Grizaudio said:

capability difference

That's the thing, init.

 

If the "mixed phase" aspect of the filter makes no difference... then calling this "a capability difference", gets peoples minds somewhat on the wrong track (thinking that long low frequency filters are an "improvement").

 

... I understand what you mean though (yes, technically it is a "capability.... difference"   .... but hopefully you underhand my point.   Otherwise, next minute you get the entire internet thinking they need to use said long filters to improve their speaker.

 

24 minutes ago, andyr said:

The key point is "the same EQ".

 

Sure... these statements are all different.

 

1 When I changed the EQ the sound quality changed

2 When I took the EQ out of miniDSP and put the (same) EQ into Roon, the sound quality changed

3 FIR can achieve a better (sounding at LF) result than IIR (when the amplitude result is the same)

 

If you said #1 ... then sure.   But people easily hear #2 and #3....  and 2 and 3 are not normally true.

[EddieT]

On 2/9/2023 at 4:51 PM, andyr said:

 

Here are my 'zero baffle' spkrs, Eddie:

 

 

 

As @davewantsmoore said:

 

 

 

... hence, with no baffle - I need massive amounts of boost below about 500Hz (plus a cut above) to get the mid/bass units flat.  The miniDSP unit I use to provide the XOs for my active rig also provided this EQ.

 

Cheers they look good which is half the battle (Ive seen some ugly OB designs).  I do like free to air Im no bass head but gonna experiment with tone wood in a Betsy style OB for an old pair of Goodmans 212's.

I read once that there no such thing as an open baffle speaker unless you're playing them in a park  i.e. the room becomes the baffle.

Edited September 7, 2023 by EddieT

[EddieT]

Found this whilst searching more info if anyone is interested.

https://jelabs.blogspot.com/2012/06/je-labs-open-baffle.html

 

A series amount of info about the who what where when why and how of Open Baffles (plus a lot of other good hifi stuff)

[lasseo]

On 02/09/2023 at 7:56 PM, davewantsmoore said:

If you look at this chart from SL site... Note that this is just the impact of the "dipoleness".

 

If I'm interpreting the chart correctly, I would be able to equalise the respose up to about ka = 4. Above that the off-axis vs on-axis deviate (in shape) so you can't equalise yourself out of trouble.