Startseite
Wählen Sie die deutsche Sprachversion Sélectionnez votre langue! New HEAD acoustics UK subsidiary New HEAD acoustics Italy subsidiary Choose the Japanese version New HEAD acoustics Korea subsidiary Choose the Chinese version
  Products & Solutions    Telecom    Products    Analysis Systems Overview    
  ACQUA Extension Modules (ACOPT) Overview    ACOPT 28
ACOPT 28 - Option SNRI & TNLR Calculation According to ITU-T G.160 (App. II, Amd. 2) (Code 6855)
This option offers a new single measurement descriptor (SMD) which allows to conduct measure- ments for calculation of values according to ITU-T G.160 (Appendix II, Amendment 2, 08/2011) as summarized in the list below. Moreover, this option provides a new menu item in the ACQUAlyzer "Calculation" menu which allows to apply this calculation method to the signal currently displayed in the time window.
Main Values:
SNRI (Signal-to-Noise Improvement): The difference of signal-to-noise ratios between processed and unprocessed signal.
NPLR (Noise Power Level Reduction): The difference of background noise levels between processed and unprocessed signal in the signal parts with short speech pauses. Thus this metric estimates the noise level reduction during speech.
 
 
TNLR (Total Noise Level Reduction): The difference of background noise levels between processed and unprocessed signal in the signal parts with no speech activity.
DSN (Difference between SNRI and NPLR): Describes the balance between TNLR and NPLR. Should be around zero.
Additional Single Values:
SNR High / Mid / Low: Signal-To-Noise ratio for speech frames with high, mid and low speech power acc. to ITU-T G.160. Available for unprocessed and processed signal.
ASL Clean: The active speech level of the clean speech signal acc. to ITU-T P.56.
Delay Processed vs. Unprocessed: Delay calculated with cross-correlation between processed and unprocessed signal.
Delay Processed vs. Clean: Delay calculated with cross-correlation between processed and clean signal.
Delay Unprocessed vs. Clean: Delay calculated as the difference of the (Proc vs Unproc) and (Proc vs Clean).
 
 
ASL Unprocessed / Processed: The active speech level of the noisy speech signals are calculated over the signal segments with low / mid / high speech power (ITU-T P.56 due to additional background noise not applicable here).
Speech activity: Speech activity acc. to ITU-T P.56, calculated with clean speech signal.
SNR Processed / Unprocessed: Alternative calculation of SNR; uses the active speech and background noise levels of processed and unprocessed signal.
BGN Level Processed / Unprocessed: The background noise level of the processed / unprocessed signal, determined in parts with no speech activity.
Note: All ACOPTs require the latest ACQUA version.
Subscribe to our newsletters!
 
 
I would like to receive the newsletter. I have read the privacy policy and accept it. I am aware that I can revoke this consent at any time (information on how to unsubscribe are included in every e-mail).
   
 
HEAD acoustics GmbH

Ebertstraße 30a
52134 Herzogenrath
Germany

E-mail: info@head-acoustics.de
Find your local contact



© 2019 HEAD acoustics GmbH. All rights reserved.
Download Information Material
top of page