Paulssurround

Adding Correct Toe-In Rotation Angle To SOv2 Is Critical To Success

Recommended Posts

7 hours ago, Baba Yaga said:

Well, yes, wandering in the dark...

I assume TOF (SO) deals with the different drivers while Phase alignment (Exakt) deals with the phases of the various frequencies from one driver, correct?

If so, TOF deals with all frequencies in some way or the other, including mids and tweeter. Boundary optimization deals with the power response provided by the room. From a certain frequency that may be irrelevant though (that seems to be what Philbo is stating), but we don‘t know what that limiting frequency is that they use in their model.
The phase alignment provided by Exakt seems to handle all drivers with all frequencies. 

Phase alignment is handled by Exakt. The signals from the woofer and mid driver are delayed so as to reconstruct the musical signal in correct phase alignment. This is well explained on the Linn website.

Time of flight is handled by SO+. How this works is more of a secret. I really can't explain it more than Petelinn did above.

Your last paragraph is as I understand things as well.

Share this post


Link to post
Share on other sites
Super Wammer

Exakt resolves phase errors in the drivers and through the crossover point(s).

It also time delays a driver (all of the frequencies coming from that driver) to allow for the physical difference in the source of the sound in the speaker (typically in a 2 way, the dome on the tweeter is closer to the listener to the point where the mid-bass cone is connected to the voice coil - so the tweeter is delayed for a few micro seconds).

SO+ may be doing some other time manipulation, but the above is correct for the Exakt filter design because the microsecond delay time has to be input to the filter design, should the designer choose to do so.

  • Like 1
  • Upvote 1

Share this post


Link to post
Share on other sites
On 09/07/2020 at 05:10, akamatsu said:

https://docs.linn.co.uk/wiki/index.php/Akubarik_Aktiv#Positioning

The following in contained within the above linked document; "The Linn 3K Array used in the Akubarik loudspeaker offers exceptional dispersion characteristics. This means that there should be little or no need to toe-in your loudspeakers."

When I picture in my mind the Akubariks that I don't have yet, I always imagine them square with the room. Fortunately, I find imaging of the sound stage rather annoying.

@akamatsu In your case I would at least try minimal toe-in on your new Akubariks, when they arrive. Mine are toed in 1,5 mm (0,059 inches).

  • Upvote 1

Share this post


Link to post
Share on other sites
Posted (edited)
55 minutes ago, Tendaberry said:

@akamatsu In your case I would at least try minimal toe-in on your new Akubariks, when they arrive. Mine are toed in 1,5 mm (0,059 inches).

I spoke with the dealer yesterday. He said he uses a little bit of toe-in for Akubariks. And thank you for replying with your helpful advice. I look forward to giving it a try.

Edited by akamatsu

Share this post


Link to post
Share on other sites
7 hours ago, Tendaberry said:

... I would at least try minimal toe-in on your new Akubariks, when they arrive. Mine are toed in 1,5 mm (0,059 inches).

My 'bariks were originally set up by my dealer without toe-in, which IIRC was the original Linn recommendation. They stayed like that for about two years.. Then, when the Space Commander came to visit and sprinkle his magic SO v1 stardust over my system, the first step was to adjust (by not very much) the speaker positioning and add a little more toe-in than you are using. (My listening position is only about 2,5m fro the speaker baseline.) Even before the changes to the SO, the toe-in improved the spatial presentation of the system more than somewhat, and I have kept it for my current SO v2 optimisation.

David

  • Like 1

Share this post


Link to post
Share on other sites
Posted (edited)

What surprises me is that SO2 requires to enter the toe-in value in angular degree. How would one measure it? With a geometry set square? Wouldn’t it be much easier to just have to enter the distance between the speaker’s housing rear outer and inner corner to the rear wall?

Yes, I know how to calculate the required toe-in value in angular degree. In a right-angled triangle the tangent sine of alpha is defined as the length of the opposite divided by length of the adjacent hypotenuse. In this case the length of the opposite is the difference between the distance of the speaker’s housing rear outer and inner corner to the rear wall and the length of the adjacent hypotenuse is  the speakers’s width on its rear side. Hence alpha (the required toe-in value in angular degree) can be calculated as the arctangent arcsine of the opposite divided by the adjacent hypotenuse. I’d be seriously interested in a rough estimation what percentage of Linn’s dealers and customers are able to either measure or calculate the toe-in value in angular degree correctly.

My Akubariks are toed-in 5 mm (roughly 0.2 inches). I leave it up to you to calculate the toe-in value in angular degree. ;)

Edited by BB1
The speaker’s width is the hypotenuse, not the adjacent
  • Like 1

Share this post


Link to post
Share on other sites
1 hour ago, BB1 said:

My last sentence in my previous post wasn’t aimed at you, David. Obviously we’ve posted at the very same time.

No problem. I'd already picked that up. The general point, as I am sure you agree, is that toe-in, like other aspects of speaker positioning, does require a certain amount of trial and error. 

I also agree with you about the way in which toe-in is specified in SO v2. Your suggestion makes sense to me, provided that speaker front baffle widths are recorded in Linn's speaker database. Do you happen to know whether this is the case? 

David 

Share this post


Link to post
Share on other sites
Posted (edited)
9 hours ago, BB1 said:

What surprises me is that SO2 requires to enter the toe-in value in angular degree. How would one measure it? With a geometry set square? Wouldn’t it be much easier to just have to enter the distance between the speaker’s housing rear outer and inner corner to the rear wall?

Angular degree seems to be the only universal method. 
E.g. inner and outer rear corner are the same for my 350, so it won‘t work here.
Using the front corners might be better but does still not work for all speaker shapes, take e.g. Kef Blades or Vivid speakers.

Calculating the angle for the 350s is a bit tricky due to their curved shape (which on the other hand makes them so appealing to me).

And of course one has to use the correct reference point for measuring the angle. It has to be the one that is stated in the SO2 dialog when selecting a speaker, in my case the ‚Outside top front corner‘. The speaker is then thought to be rotated with the reference point being fixed.

For clarification I attached a picture for my right speaker. Hope I got the angle of 10 degrees correct with this approach 😀

474627DF-F041-4133-91E8-AF81D44B5A6F.jpeg

Edited by Baba Yaga

Share this post


Link to post
Share on other sites
2 hours ago, weinconsi said:

... may this helps you - made it easy for me to measure the angles.

Bildschirmfoto 2020-07-15 um 13.44.56.png

Yes, that would do it. Great idea. Wish I had one of these devices. 

I used the laser on my laser measure device to follow the laser beam down the middle of my Akubariks to the front wall from the front baffle of the speakers, at the middle of the 3K array. I couldn’t measure from the side of the speaker cabinets, given that my speakers have curved sides.

Then I used a T-square to get a right angle from the wall, and used a protractor to measure the difference in the angle of the speakers from the T-square. 

I am not sure how precise I was able to be, but feel confident enough with the results

Share this post


Link to post
Share on other sites

Use a compass. Need I explain how?

  • Haha 1

Share this post


Link to post
Share on other sites
16 hours ago, DavidHB said:

our suggestion makes sense to me, provided that speaker front baffle widths are recorded in Linn's speaker database. Do you happen to know whether this is the case?

Yes, the „Speaker Metrics“ contain the front baffle width.

LinnAkubarikSpeakerMetrics.xml
<SpeakerMetrics version="3">
	<Manufacturer>linn</Manufacturer>
	<Model>akubarik</Model>
	<Width>0.21</Width>
	<XYMeas>Outside top front corner of cabinet</XYMeas>
	<ZMeas>Bottom edge of plinth</ZMeas>
	<DriveUnits>
      ...

Share this post


Link to post
Share on other sites
On 15/07/2020 at 10:44, Baba Yaga said:

And of course one has to use the correct reference point for measuring the angle. It has to be the one that is stated in the SO2 dialog when selecting a speaker, in my case the ‚Outside top front corner‘. The speaker is then thought to be rotated with the reference point being fixed.

No, any arbitrary reference point of the speaker can be chosen for measuring the angle. In the following sketch the blue triangle matches the way you’ve measured alpha. The red triangle is identical to the blue one and uses the same reference point on the speaker’s side. And the purple triangle is similar to the red one, i.e. all three angles of both triangle are identical, but it moves the reference point to the outside top rear corner.

Sketch_1.thumb.png.79197a4687fe5893c026f583beabefbd.png

In the next sketch, the upper green triangle is similar to the purple one and therefore it is also similar to the blue one in my previous sketch:Sketch_2.thumb.png.600b9522349965743326865e9a7066e3.png

The sine of alpha is a / c, where c is the width of the speaker’s rear side. The length of a can be calculated as L2 - L1, where L2 is the distance from the outer top rear corner to the front wall and L1 is the distance from the inner top rear corner to the front wall.

Therefore sin (alpha) =  a / c = (L2 -L1)  / c and alpha = acrsin ( (L2 -L1)  / c).

SO2 doesn’t know the width of the rear side, but the width of the front side. Hence its calculation of the toe-in angle must be based on the position of the outside top front corner (as indicated by the lower green triangle in above sketch). By just adding the option to enter the distance between the inner top front corner and the front wall in SO2, it could calculate the correct toe-in value automatically.

On 15/07/2020 at 10:44, Baba Yaga said:

Using the front corners might be better but does still not work for all speaker shapes, take e.g. Kef Blades or Vivid speakers.

Agreed, the reference point for e.g. KEF Blade 2s is „Front centre of plinth“. But wouldn’t it be of great benefit if Linn would offer entering a second reference point for automatically calculating the toe-in if the speaker geometry allows for it?

Please note that I’ve corrected an error my previous post. At practical toe-in values the effect of this error is negligible since the lengths of the adjacent and the hypotenuse are nearly the same. If e.g. "a" is 5 mm the correct angle is 1.6372°  whereas the incorrect angle is 1.6366°. And if "a" is 3 cm, the  correct angle is 9.87° and the incorrect one 9.73°. 

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.