AFMG Network Forum

2brain · Member

Is it time RASTI was retired and STIPA or real STI replaced it?

Even customers now seem to have read IEC 60268-16 and expect actual system result measurements in STIPA but they can only get RASTI or %Alcons predictions from EASE4.2.

Are there any plans to 'upgrade' predictions to STIPA or proper STI?

Bruce · Member

"Proper STI" has been available since the release of EASE 4.0. In order to calculate it you need the AURA module where you are able to calculate full-length response files for listener seats or echograms for each mapping point. You then have a choice of the statistical measures like %ALCons or the impulse response based measures like STI.

PaulMalpas · Member

Bruce is right, we can potentially get an STI from a ray-tracing derived decay curve with AURA, but most model runs just don't justify the hours it takes to get any results out of AURA, or the days it might take for me to determine just what AURA is doing in order for me to believe what I am seeing is valid.

The great thing about the statistical part of the EASE software - and the only real reason myself and anyone else I have discussed this with uses EASE - is that it is very easy to understand just what ease is doing with the calculation. A direct level plot, based on the speaker directivities - check! A reverb level based on the volume, RT and total acoustic output - check. A noise level, which if fixed relative the Total SPL calculated on a pre-run lets you set the S:N - check. All the relevant parts for an empirical %Alcons calc, followed by an empritical translation by the Farrell-Becker to STI. Now, this may be called RASTI in the software, but it isn't, as RASTI is a measurement method that applies weghting characeteroistics to Modulation Transfer Indeces calculated for 9 mod frequencies over the 500Hz and 2kHz octave bands. EASE does not go near MTI at this stage, though it could do so, based on statistical acoustic decays only.

And thsat is the point/. If EASE used an MTI approach at teh statistical stage, we could get a rough STI, simulated RASTI, simulated STI-PA or whatever we wanted, without waiting hours for a ray-tracing. As most jobs are a case of comparing up to 5 cases in each of a dozen or so areas, this is the only practical way to proceed when trying to set the RT needed in each case to meet an STI stanard.

So, please, no misleading mentoion of RASTI and no more Peutz/Farrell-Becker approximations.

I've said my piece now!

Paul