AFMG Network Forum
AFMG Network Forum Index -> EASE 4 -> questions about the STI
Post new topic  Reply to topic View previous topic :: View next topic 
questions about the STI
PostPosted: Tue Sep 29, 2015 9:33 am Reply with quote
gggffgg
Member
 
Joined: 29 Sep 2015
Posts: 2
Location: Papouasi




Hello,

I have many questions about the validity of STI in EASE, here is the first:

- For example, I have a space with a volume of 4000 m3.
I add a measurement of reverberation time: 1.8 seconds in the "Room data"

I add speakers, and I calculate the STI 60268-16 (2011) Ed. 4 with Signal masking and Noise Levels.

I get an average of 0.48 STI.

- Now I open it locally to an adjacent room of the same size (I open some doors), so I have a EASE model with a volume of 8000 m3.
This local second has no speakers.
The reverberation times measured in the first local change little, 2.0 seconds.

I get an average of 0.52 STI ...

- Why STI calculations 60268-16 (2011) depends on the volume and absorption coefficients deducted from my reverberation time measurements data locked in Room Data ?!

- My STI values have increased because I have increased the volume, surfaces in the model and indirectly the absorption coefficients of these surfaces, I think this is very worrying.

- What can you tell me about the validity of the calculation in this case?

- Have you some details about the STI calculation used by EASE?

thank you in advance
View user's profile Send private message
PostPosted: Tue Sep 29, 2015 12:00 pm Reply with quote
thomas
Member
 
Joined: 16 Mar 2011
Posts: 73




Hi Pierre,
a first, rough, thought is:
Reverberation time is important for speech intelligibility, but nor the only factor.
Another one is the level of the reverberation field.
When you open the doors, then the total absorption area increases and the level in the reverberation field decreases.
But the direct sound and the early reflections remain the same, or nearly the same. (All assuming the sound source and listener positions remain in room 1.)
So the SNR will propably increase. And STI is mainly calculated from SNR.

I think when you compare total level, L7, L50, total level - L50, C50 for both situations, you should get an imression of what happens.

Kind regards
thomas
View user's profile Send private message
PostPosted: Tue Sep 29, 2015 1:20 pm Reply with quote
gggffgg
Member
 
Joined: 29 Sep 2015
Posts: 2
Location: Papouasi




Indeed, with a Sabine / Eyring theory, the absorption area increases and therefore the reverberant field decreases.
However the area concerned by the PA in real life (understanding "in situ") changed little (0.2 seconds more) as the reverberant field.

And the fact that the STI increases due to a simple opening on an adjacent volume is strange (especially if the RT increased slightly).

In complement here is another example:

I have a 400 m corridor.

For simplicity I make a model of 200 m
I place speakers in the ceiling.
I force RT (measuring in situ)

I get a STI 0.49.


Now if I make a model of 400 m, with the same PA as the previous model (quantity and position) and the same RT
I double the volume and surfaces.

And I get a STI 0.52 ... Yet locally, under the speakers, we have the same configuration (same RT,Same speakers, Same level).

Thank you very much for your answers !


Last edited by gggffgg on Sun Dec 27, 2015 10:29 am; edited 2 times in total
View user's profile Send private message
PostPosted: Wed Sep 30, 2015 2:00 am Reply with quote
thomas
Member
 
Joined: 16 Mar 2011
Posts: 73




Hi Pierre,
Quote:
Indeed, with a Sabine / Eyring theory, the absorption area increases and therefore the reverberant field decreases.
However the area concerned by the PA in real life (understanding "in situ") changed little (0.2 seconds more) as the reverberant field.


No, when you couple two identical rooms, the absorption area increases real (at least when you have more absorption in the second room then on the removed partition wall). In theory (Sabine) the reverberation time increases, because the absorption area is less than doubled.

But anyway, I just calculated a little bit (I wanted to see, if I can use simulation for a proof for my earlier statement).

room: small: 20 m x 10 m x 7 m / large 40 m x 10 m x 7 m
floor/walls: alfa = 0,1, ceiling alfa=0,6
1 speaker (sphere) in the middle (z=6 m)
1 seat (somewhere)

simulation: aura mapping 2d

results (all for 1 kHz 1/3 oct.):
total SPL small: 90.6 dB large: 89,4 dB
D50 (defintition): small: 53,2% large: 62,%

calculated Levels with total SPL & D50:
L50 (0..50ms): small: 87,9 dB large: 87,3 dB
L50..inf ("reverberation level"): small: 87,3 dB large: 85,1 dB

The early reflection level decreases by 0.6 dB, wheras the late level decreases by 2.2 dB. So STI has to be better in the large room.

And indeed, all parameters for the prediction of intelligibility (center time, C50, D50, STI) are better for the large room, than for the small one.
And: T: small: 1,4 s large: 1,7 s !

Kind Regards
thomas
View user's profile Send private message
AFMG Network Forum Index -> EASE 4
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
All times are GMT  
Page 1 of 1  

  
  
 Post new topic  Reply to topic  


Powered by phpBB © 2001-2003 phpBB Group
Theme created by Vjacheslav Trushkin
Variation by CodeWeavers and AFMG