Linn Owners

Time of Flight

hannson.design

Newbie
Wammer
Oct 31, 2018
108
76
33
HiFi Trade?
  1. No
I can’t for the life of me figure out why the distance to the listener is needed for the calculation of “Time of Flight” in SO+. I’ve always been under the impression that any time delay compensations were calculated by considering the relative offsets of the drive unit acoustic centres. So why would we need to provide a distance to listener measurement? Any ideas?

 
  • Like
Reactions: akamatsu

Paulssurround

Space Explorer
Wammer
Oct 11, 2018
3,991
4,295
148
Vancouver
AKA
Paul
HiFi Trade?
  1. No
My take of it is that when you create music, the sound waves in a particular listening room are going to create room modes, or boomyness with the bass.

You can’t change the laws of physics, and as you walk around the room, you will probably find areas of the room that has boomyness.

This boomyness can often be at the listening position.

Space optimisation doesn’t eliminate the room modes, but it can move the boomyness away from the listening position, therefore giving a much better sound quality

 

timster

Linnguist
Wammer
Oct 9, 2018
2,955
2,459
148
Wiltshire, UK
HiFi Trade?
  1. No
I can’t for the life of me figure out why the distance to the listener is needed for the calculation of “Time of Flight” in SO+. I’ve always been under the impression that any time delay compensations were calculated by considering the relative offsets of the drive unit acoustic centres. So why would we need to provide a distance to listener measurement? Any ideas?
There's inter-speaker time of flight (between drive units) and extra-speaker time of flight (between speaker and listening position). If the listening position is off-set from the centre, the time of flight from right speaker to listening position will differ from that of left speaker to listening position. SO+ can compensate for that using Exakt jiggery-pokery.

My take of it is that when you create music, the sound waves in a particular listening room are going to create room modes, or boomyness with the bass.

You can’t change the laws of physics, and as you walk around the room, you will probably find areas of the room that has boomyness.

This boomyness can often be at the listening position.

Space optimisation doesn’t eliminate the room modes, but it can move the boomyness away from the listening position, therefore giving a much better sound quality
This is the case for SO in genera, nNot only SO+. But I'm not sure that's how it does it. AFAIK, it mitigates the modes with the filters, but I don't think SO+ actually moves them. I'd like to know if that was the case - Philbo?

 
Last edited by a moderator:
  • Like
Reactions: akamatsu

Philbo

Newbie
Wammer
Apr 26, 2017
72
91
38
HiFi Trade?
  1. No
I can’t for the life of me figure out why the distance to the listener is needed for the calculation of “Time of Flight” in SO+. I’ve always been under the impression that any time delay compensations were calculated by considering the relative offsets of the drive unit acoustic centres. So why would we need to provide a distance to listener measurement? Any ideas?
Not sure I understand your confusion.  The time of flight we correct for within each Exakt loudspeaker is to ensure the linear phase output from each individual drive unit arrives at the listening position at the same time.  To calculate the path differences from each drive unit to the listener, and therefore calculate the required delays, we need to know the position of the loudspeaker and the position of the listener (we already know the position of the drive units within the loudspeaker).

My take of it is that when you create music, the sound waves in a particular listening room are going to create room modes, or boomyness with the bass.

You can’t change the laws of physics, and as you walk around the room, you will probably find areas of the room that has boomyness.

This boomyness can often be at the listening position.

Space optimisation doesn’t eliminate the room modes, but it can move the boomyness away from the listening position, therefore giving a much better sound quality
This has nothing to do with time of flight delays.

 

DavidHB

Well-Known Wammer
Wammer
Oct 9, 2018
2,632
1,773
148
Isle of Wight, UK
AKA
David
HiFi Trade?
  1. No
I can’t for the life of me figure out why the distance to the listener is needed for the calculation of “Time of Flight” in SO+. I’ve always been under the impression that any time delay compensations were calculated by considering the relative offsets of the drive unit acoustic centres. So why would we need to provide a distance to listener measurement? Any ideas?
I used to think that too, until I was corrected by Philbo. And he's the expert.

David

 

hannson.design

Newbie
Wammer
Oct 31, 2018
108
76
33
HiFi Trade?
  1. No
Not sure I understand your confusion.  The time of flight we correct for within each Exakt loudspeaker is to ensure the linear phase output from each individual drive unit arrives at the listening position at the same time.  To calculate the path differences from each drive unit to the listener, and therefore calculate the required delays, we need to know the position of the loudspeaker and the position of the listener (we already know the position of the drive units within the loudspeaker).
Thanks Philbo. I think my confusion is due to the fact that I don’t understand!

My thinking goes along the lines of:

Without some form of time alignment and if each drive unit reproduces a transient at the same time, the listener will hear the transient multiple times depending upon on the distances between each drive units acoustic centres. If the listener moves closer to the loudspeaker the time delays between each drive unit will remain the same, only the time of Flight will be reduced.

If the drive units are time aligned, the listener will hear only one transient sound. If they were to move closer to the loudspeaker, they would still only hear one transient sound even though the time of flight has been reduced.

I can understand why Time Alignment would be desirable, but Time of Flight is a mystery because the difference in acoustic centres are already known and the Time of Flight is not needed to calculate the filter time delays.

 

Briain

DJ Duggie Dinbins
Wammer
Oct 10, 2018
108
142
63
Edinburgh
AKA
Briain
HiFi Trade?
  1. No
Hi

If you sit close to a tall pair of multi-unit speakers - like my 350s - your ear will be approximately level with the treble units. Just considering two of the units for now (the treble unit, which is about 37 inches above floor level, and the lowest bass unit, which is approximately 12 inches above floor level) if you were to take a tape measure and measure the distance between your ear and the treble unit, then again measure between your ear and the lower bass unit, you'll note quite a large difference in the measurements (I blame Mr. Pythagoras), which means that the output of the bass unit will arrive later than that of the treble unit (as sound travels at approximately 1 foot per millisecond, you can work out the time difference from the difference in the two physical distance measurements that you have just taken) so that is what the time of flight delay settings address; in a two way speaker, it would be the delay applied to the signal fed to the nearest unit such that the sound from it and furthest unit arrives at the same time (though obviously, if you have 5 unit speakers like 350s, there are in-between delays applied to the units in-between the two that I used in the above example, such that the sound from all of them arrives - at your ear - at exactly the same time).

Of course, if you were to move your seat further away and take both measurements again, the difference between them all will be smaller (due to the angles between the two units we are measuring to being smaller; Mr. Pythagoras strikes, once again) so that means you would require a smaller difference between the delays applied to all the signals, which is why time of flight calculator (in Konfig) needs to know where all the units are (above floor level, and obviously, each model of speaker has different unit spacings and different heights above floor level), where your head height is at (in terms of being above floor level, not its state of sobriety), and the distance between yourself and the speakers; changing any one of these parameters will result in changes required to the various delays (clearly, Mr. Pythagoras only had a single unit in each of his speakers, so this pesky little side effect of his revelations likely never occurred to him, whilst head banging away to the Hawkwind of his day).

Bri :)

NB I am using 350A (so no Exakt and thus no time of flight tweaks are available) but I also sit 4.7 metres from them, so the differences in time of flight error (between the units) will be less severe than for someone who sits a lot closer to them. That said, I do have a 345 sub sitting in the rear corner of the room, so just 3.3 metres behind me, so I have had to use a miniDSP to add about 3.9 milliseconds delay to the sub feed (Pysuboras theorem?) as that was an absolutely massive (4.70 m - 3.36 m = 1.34 metres, or about 3.9 ms) time of flight error which needed fixing (to eliminate the 'precedence effect', which was pretty darned horrible thing to experience; a very big thank you to Phil Budd for alerting me to the existence of the precedence effect), so even my clonky old in-Exakt system has some form of time of flight compensation.

 
Last edited by a moderator:
  • Like
  • Thanks
Reactions: Russ-R and akamatsu

Briain

DJ Duggie Dinbins
Wammer
Oct 10, 2018
108
142
63
Edinburgh
AKA
Briain
HiFi Trade?
  1. No
Hi

Not at all; I was merely clarifying what you had already said by using a tall speaker example, whereby folks might [hopefully] more easily be able to visualise - in their 'mind's eye' - the angles and distances that we are dealing with when sitting very close to the speakers, compared to the angles and distances that we are dealing with when sitting further away from the speakers (and in my case, really quite far away from them; many would say the further that they sit from my speakers, the better that they sound; though when they site using totally different city as being the optimal distance, I sometimes wonder if I need to consider improving the system) and indeed, I wanted to also add a bit about what I do with my (closer by) rear sub, as that is a good example of using time of flight compensation in a non-Exakt system (in my case, the main speakers are fed from the XLR outputs of my source, but I'm having to digitise the 'spare' RCA analogue outputs from my source, add the required 3.9 ms overall delay, then convert it all back to analogue before sending it onwards to the sub).

Bri :)

 
Last edited by a moderator:

hannson.design

Newbie
Wammer
Oct 31, 2018
108
76
33
HiFi Trade?
  1. No
Thanks Timster and Briain, your explanations have shown me where I was going wrong.

I was thinking in terms of the time delayed acoustic centres being on a 2 dimensional flat plane, whereas if we think about them being on a 3 dimensional arc with our ears being at the centre of the sphere, I can see the need for the speaker to listener distance. Easy really, when you know the answer!

I just can’t believe I’ve struggled with that one for so long  :doh:

 
Last edited by a moderator:
  • Like
Reactions: timster

Briain

DJ Duggie Dinbins
Wammer
Oct 10, 2018
108
142
63
Edinburgh
AKA
Briain
HiFi Trade?
  1. No
I guess it's really 4 dimensions as it involved time, too (so spacetime: https://en.wikipedia.org/wiki/Spacetime) but one really needs the 5th dimension to then fully appreciate it (red wine, so that - one assumes - must be 'spacetimewine', but I can't yet find the Wikipedia page on that concept, though; maybe we should compose a particularly silly one and see how long it lasts)! :)

 
Last edited by a moderator:
  • Haha
Reactions: akamatsu

hannson.design

Newbie
Wammer
Oct 31, 2018
108
76
33
HiFi Trade?
  1. No
Do you own KEF LS50s?  That would help explain the struggle...
Ha ha 😂. No, I’ve got a pair of Katalyst Akudoriks which I absolutely adore. By only thinking in 2D the distance to listener measurement didn’t make sense, but I’ve got it now. I’ve probably been a flat earther for far too long 😂

 

Baba Yaga

Wammer
Wammer
Feb 12, 2019
763
632
113
Baba Yaga‘s hut
AKA
BY
HiFi Trade?
  1. No
Can we take this a step further, please?

The ‚Pythagoras explanation‘ works for let‘s say particals leaving the driver at the same time and arriving at slightly different times at the ear hence the need for correction, I get that.

However, we are talking about a superposition of waves (longitudinal ones for that matter) with different wavelengths (and frequencies), so even if the first signal arrives at the same time (which I think is important for localization) the next ones will fan out (or not?). So how does the correction help here? I must be missing something...

edit: I think I can answer myself, every frequency needs to be shifted differently 

 
Last edited by a moderator:
  • Like
Reactions: akamatsu

Paulssurround

Space Explorer
Wammer
Oct 11, 2018
3,991
4,295
148
Vancouver
AKA
Paul
HiFi Trade?
  1. No
Not sure I understand your confusion.  The time of flight we correct for within each Exakt loudspeaker is to ensure the linear phase output from each individual drive unit arrives at the listening position at the same time.  To calculate the path differences from each drive unit to the listener, and therefore calculate the required delays, we need to know the position of the loudspeaker and the position of the listener (we already know the position of the drive units within the loudspeaker).

This has nothing to do with time of flight delays.
I messed up.

I need to stop posting late at night before I go to sleep, and actually read the OP more carefully    9_9

 

fredbatch

Newbie
Wammer
Nov 1, 2018
205
333
83
North Wales
HiFi Trade?
  1. No
Another way of putting it:

t=d/v

Not forgetting:

v=f.λ  

...in support of Briain and Pythagoras:

spkr-room.JPG

 
Last edited by a moderator:
  • Thanks
Reactions: akamatsu

akamatsu

Michael
Wammer Plus
Oct 9, 2018
7,221
8,110
183
Point Roberts, WA, USA (Vancouver)
AKA
Michael
HiFi Trade?
  1. No
I've been scratching my head about "time of flight optimisation" for the past couple of weeks. This thread has answered my main questions, which of course has revealed another question. In my house, the floor has a slight downward slope from left speaker to right speaker. My right speaker is set 11mm higher off the floor than the left. This is to have the speakers be at the same level with each other on the horizontal plane. I'm pretty sure SO+v2 uses the height of the speakers off the floor to calculate time of flight optimisation. But SO doesn't know that my floor is sloped, so it thinks that the right speaker is higher than the left, which it isn't. I can correct time of flight by inputting the heights as the same. My question is how would this affect the room mode and boundary optimisations? I would think minimally as we are dealing in frequencies 80Hz and below.
 

HollyJohnsen

Newbie
Wammer
Dec 20, 2021
192
314
83
Germany, Waldeck
HiFi Trade?
  1. No
I would think minimally as we are dealing in frequencies 80Hz and below.



Thinking about that would be an option. For a few days I have been experimenting with SO+ and comparing without SO+ with settings saved by me. In theory, there should be problems with the bass without SO, but in practice I can not determine it, the bass does not cover anything, remains melodic, does not boom. On the contrary, the music wins without SO+, it becomes so much more lively, more audible, clearer and more detailed. I will continue to study ....
 

Moomintroll

Fictional Character - not to be taken literally
Wammer
May 20, 2014
3,516
4,525
148
HiFi Trade?
  1. No
I've been scratching my head about "time of flight optimisation" for the past couple of weeks. This thread has answered my main questions, which of course has revealed another question. In my house, the floor has a slight downward slope from left speaker to right speaker. My right speaker is set 11mm higher off the floor than the left. This is to have the speakers be at the same level with each other on the horizontal plane. I'm pretty sure SO+v2 uses the height of the speakers off the floor to calculate time of flight optimisation. But SO doesn't know that my floor is sloped, so it thinks that the right speaker is higher than the left, which it isn't. I can correct time of flight by inputting the heights as the same. My question is how would this affect the room mode and boundary optimisations? I would think minimally as we are dealing in frequencies 80Hz and below.
Easy enough to test?

’troll
 
  • Upvote
Reactions: akamatsu

Latest posts

Forum statistics

Threads
113,444
Messages
2,451,263
Members
70,783
Latest member
reg66

Latest Articles