SOFTWARE AMPLITUDE LIMITATION IN DIGITAL SIGNAL PROCESSING APPLICATIONS
Original Publication Date: 1995-Jul-01
Included in the Prior Art Database: 2002-Mar-25
Wolfgang Meier: AUTHOR [+2]
The main focus of this idea is the reduction of harmonics and subharmonics of a software realized digital amplitude limiter. In RF base stations where the modulation devi- ation has to be limited, it is important to minimize harmonics which can influence the neighbour channels.
MOFOROLA Technical Developments
SOFlWARE AMPLITUDE LIMITATION IN DIGITAL SIGNAL PROCESSING APPLICATIONS
by Wolfgang Meier, Norbert Riittger and Jiirgen Schmidt
1. GENERAL input audio. that has to be limited and the limit levels +- U-limit. The number of saved digital The main focus of this idea is the reduction of samples is dependent ofthe rate, the audio is sampled. harmonics and subharmonics of a software realized
digital amplitude limiter. buffer-length = number of samples =
Time x sample rate
In RF base stations where the modulation devi- ation has to be limited, it is important to minimize Example: Time = Sms harmonics which can influence the neighbour sample rate = 8kHz channels. bufferPlength = 40
2. DESCRIPTION The figure shows also the behaviour ofan hard- ware limiter or software limiter which cuts the sam-
2.1 GENERAL pleat the defined limit value (at level +- U-limit). This means that the limited output audio includes a Figure 1 shows the content of an audio input large number of harmonics.
buffer with a defined length of time. It shows the
I +lJJimit ---_-----
F- i j
i ! i ! i ! i ! i ! i ! i !
! i i ! i
i ! i ! i ! j ! i ! I ! i ! i ! i ! ; ; I ]
I , i ! : ( ! ! ,~ t I I Ii ! ! I i ! ! ! j I mndifid samnlaa ii I I j i ! 1 l !
! i i ! I i j j j I I
i ! ! I i I i! !A i
'ir? I /
input audio !
Fig. 1 Input audio buffer (digital samples) with hard limiting
0 hlot"r"lL!. Inc. 1995 23 JulylYYj
audio samples (see Figure 2):
1. step -find the sample with the maximal amplitude value of the complete input buffer (example: -U-max-sample).
2. step -compare the absolute value of the maximal value
(K-max I) with the absolute value of the upper limit value
If the maximal sample value is higher than the upper limit value, a limitation process is necessary.
If the maximal sample value is lower than the upper limit value, no limitation process is done for this input buffer contents.
3. step (limitation process starts) -compute the correction factor (always positive)
The difference between the soft limiter and the hard limiter is that no cutting of the audio samples is done. Instead there is a modification of the audio samples that have to be limited.
There are two methods possible with this solI limiting method.
1. Sofl limiting with one correction factor: -this method uses a correction factor which is computed with the maximal absolute value of the audio samples, stored in an input audio bulk (see 2.1 General; 2.3.Description of the limit- ing process with one correction factor)
2. Sofl limiting with two correction factors: -this method uses two different correction fac- tors which are computed with the maximal positive and negative values of the audio sam- ples, stored i...