Fane Sovereign 12-250TC project measurements

[ppataki]

One of my friends has built a pair of floorstanding loudspeakers using 12" Fane 12-250TC fullrange drivers. I am posting this on his behalf and with his permission.

Cabinet is sealed with the below parameters:

Made of 19mm baltic birch plywood using internal bracing (unfortunately I do not currently have the pictures of the inside of the cabinet)

The final result looks like this:

The objective measurements will follow below, also with the mdat file attached

Subjectively the sound is typcial to fullrange drivers:
- very 'live sound', you feel like you could talk to the musicians in the room
- extremely holographic presentation, each instrument has its own 3D space
- extremely precise soundstage
- very heavy DSP is needed to flatten the response (all the listenings were done with DSP - don't even try these without it, you will cry )

Now the measurements:

FR with zero correction (L/R channels)
1/12 smoothing

FR with full correction

Corrections were done using a combination of 'pre-EQ' low and high shelves followed by parametric EQ-ing using CraveEQ (Transparent Phase Mode) based on REW generated filters and then CurveEQ for further flattening of the response
Also tried Dirac Live but the resulting sound became a bit dull and lifeless compared to the above method

Now when listening to music using the above curves the bass was not there......
After measuring the LR channels combined the reason for that became apparent:

There is a huge cancellation dip between 50 and 100Hz when the two speakers are measured together
Normally we shall not touch this since unwanted ringing might occur if we try to compensate but we had no choice and applied some further filters to correct the summed response which looks like this after correction:

By listening we could not notice excessive ringing, I suspect due to the size of the room (70 sqm)
Bass came to life and the 'order of the world' is restored

Distortion looks OK too in my opinion:

Step response is pretty nice too:

Wavelet diagram:

Overall my friend is extremely happy with the results!

I cannot upload the measurement since it is too big but here is the link to it:

https://drive.google.com/file/d/17ZQfNaIoeTs3OJc9DGobv_K5fRjhnc7H/view?usp=sharing

I would encourage anybody to try a big fullrange driver (with adequate DSP) - the listening experience is worth the effort

 

[Rednaxela]

What a cool project! Bookmarked.

The speakers are looking very nice. I have some experience with their bigger brother. There are really a lot of things wrong with these Fanes but I can totally see how an end result like this can be incredibly enjoyable to listen to.

Just for my understanding, you corrected for a flat in-room response at the MLP?

 

[ppataki]

Just for my understanding, you corrected for a flat in-room response at the MLP?

Yes, that is correct

I have some experience with their bigger brother.

The bigger brother will arrive today, actually
Allegedly the last available pair in Europe since Fane unfortunately stopped producing the 15" version.....

What a cool project! Bookmarked.

Thank you

 

[ppataki]

And here we go, some pictures attached about the 15-300TC model that just arrived this morning
I have put there my hand for scale only....it is HUGE!!

For the time being I have no plans with them, they will take some rest in the garage and then time will tell...
I am still having an awful lot of fun with my MCLA speakers

EDIT: added a video with a 9V battery

New video by Peter Pataki

photos.app.goo.gl

 

[ppataki]

@Rednaxela
You might be interested in this; this is another project I did in my living room before moving to my current MCLA project

DIY Audio Nirvana 12" + Dayton Reference FAST project

I am sharing with you yet another DIY project we have done with my neighbor (who is a carpenter...:)) much earlier this year It is a FAST (Fullrange Assisted Subwoofer Technology) system consisting of 12" Audio Nirvana Classic fullrange speakers and Dayton Audio RSS315HF-4 subs I could not find...

www.audiosciencereview.com

 

[voodooless]

That is a pretty cool project! Were any off-axis measurements done?

Can you elaborate a bit more on how you fixed the bass response without changing the location of the speakers?

How about a multitone measurement? Would be interesting to see the IMD on such a speaker? Large cones have the advantage that they don't need to move a lot, so it might actually be not bad.

 

[ppataki]

That is a pretty cool project!

Thank you!

Were any off-axis measurements done?

No, only MLP measurements were done
But I can tell you that these type of speakers do beam like hell so I could only recommend them for 1-seat listening position scenarios (that is the case for me and also for my friend so we don't mind that)

Can you elaborate a bit more on how you fixed the bass response without changing the location of the speakers?

Just applied correction filters to this measurement (this is from my original post):
This is LR channels combined

In REW we generated filters and applied those in CraveEQ and the results you can see in the original post
The way we did that is to set target to 65dB, set LC cutoff at 25Hz with 24dB/octave slope then set Filter Tasks like this:

At the end REW did not have to introduce any boosting filters
And then added a 10dB boost in CraveEQ to compensate (no clipping in the signal chain since volume control is at the very beginning of the chain in Jriver using its 64-bit internal volume)
As mentioned in my post we were a bit afraid to use this method since it generally introduces excessive ringing but in my friend's case it was not audible, probably due to the size of the room

So signal chain is like: 64-bit volume control --> pre-EQ shelves (high and low) - this is to lift the two ends of the FR so it is easier to work with that in REW in the subsequent steps --> per channel EQ-ing using CraveEQ based on REW filters --> per channel CurveEQ to further flatten the response --> LR combined EQ-ing using CraveEQ based on REW filters to compensate for the cancellation dip --> 'Room curve' up to taste using CraveEQ
All this is in DDMF Metaplugin and btw it can work in a multichannel environment too (that is what I am doing in my current system)

As mentioned in my post, we tried Dirac Live too but the sound was rather lifeless and the same goes to using Convolution too

How about a multitone measurement?

I will need to ask my friend to do that. Which multitone measurement in REW do you have in mind?

 

[voodooless]

No, only MLP measurements were done
But I can tell you that these type of speakers do beam like hell so I could only recommend them for 1-seat listening position scenarios (that is the case for me and also for my friend so we don't mind that)

I understand that. But off-axis will be heard back in reflections, so it's still relevant to see if it's at least smooth even if it's beamy.

Just applied correction filters to this measurement (this is from my original post):
This is LR channels combined

In REW we generated filters and applied those in CraveEQ and the results you can see in the original post
The way we did that is to set target to 65dB, set LC cutoff at 25Hz with 24dB/octave slope then set Filter Tasks like this:
View attachment 215522
At the end REW did not have to introduce any boosting filters
And then added a 10dB boost in CraveEQ to compensate (no clipping in the signal chain since volume control is at the very beginning of the chain in Jriver using its 64-bit internal volume)
As mentioned in my post we were a bit afraid to use this method since it generally introduces excessive ringing but in my friend's case it was not audible, probably due to the size of the room

Seems like a mostly bog-standard EQ. Low Q filters should still be fine regarding ringing. I'm surprised how well the holes are filled up. Lucky

I will need to ask my friend to do that. Which multitone measurement in REW do you have in mind?

Something like this should be interesting:

 

[Rednaxela]

very 'live sound', you feel like you could talk to the musicians in the room

This is something I experienced myself too when playing around with the fifteens. Their effortlessness and this live feeling are pretty addictive and seem to make up for a lot of their (on-paper) flaws.

Which made me wonder, after you EQ-ed your twelves into submission so brutally, how much of this high-sensitivity full range tricone Fane magic remained in your opinion?

Also, have you considered MMM + a little house curve? Nothing outrageous, just a mild deviation of ruler-flat?

 

[ppataki]

Which made me wonder, after you EQ-ed your twelves into submission so brutally, how much of this high-sensitivity full range tricone Fane magic remained in your opinion?

It not only remained, but got much better!
The flatter you make these the better the experience will be
In my living room I have had 6 different one-way systems in the last 2 years so I think I do have some experience with these.... (I only use 1-way systems btw.)

MMM + a little house curve?

Yes, house curve is there but not on the measurements
We do not however use those 'standard' house curves but we set it totally up to taste
Currently my house curve is:
Low shelf +4.5dB at 100Hz with Q=1
High shelf -1.5dB at 20kHz with Q=0.2
Again, this I believe totally depends on your taste, the room and the equipment
I have used different house curves on my previous systems

 

[voodooless]

Have any plots of the eventual EQ curve?

 

[ppataki]

Have any plots of the eventual EQ curve?

No, but I will ask him for the screenshots from DDMF Metaplugin

 

[ppataki]

Something like this should be interesting:

I have done this on my MCLA system. It looks like this:

Is this what you are looking for?
If yes, I will ask my friend to do it with his Fanes

 

[voodooless]

I have done this on my MCLA system. It looks like this:

View attachment 215555

Is this what you are looking for?
If yes, I will ask my friend to do it with his Fanes

Yeah, that looks useful!

 

[ppataki]

Yeah, that looks useful!

Cool, I will ask him to do the same
btw I did this with both channels driven at the same time
What distortion value is good enough here? I guess my 11.8% is pretty high, isn't it?

 

[voodooless]

What distortion value is good enough here? I guess my 11.8% is pretty high, isn't it?

I don’t think that is the real number though. Looks like roughly 1% < 200 Hz, much less above.

 

[ppataki]

I don’t think that is the real number though. Looks like roughly 1% < 200 Hz, much less above.

How do you tell the 1%?

 

[voodooless]

How do you tell the 1%?

It’s actually even less . Your fundamental is about 94dB, distortion product at 50, making the difference 44 dB, which makes about .63%. Above 5kHz it’s worse, about 1.2%. Still not bad I would say.

 

[ppataki]

Now I got it, thank you!

 

[Digby]

What is the allure of the big full range drivers in comparison to, say, a large coaxial speaker. They obviously take a lot of EQ to tame, so is it just a case of cost or are there other differences that had you preferring a full range driver?