• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1 
2 /* -----------------------------------------------------------------------------------------------------------
3 Software License for The Fraunhofer FDK AAC Codec Library for Android
4 
5 � Copyright  1995 - 2013 Fraunhofer-Gesellschaft zur F�rderung der angewandten Forschung e.V.
6   All rights reserved.
7 
8  1.    INTRODUCTION
9 The Fraunhofer FDK AAC Codec Library for Android ("FDK AAC Codec") is software that implements
10 the MPEG Advanced Audio Coding ("AAC") encoding and decoding scheme for digital audio.
11 This FDK AAC Codec software is intended to be used on a wide variety of Android devices.
12 
13 AAC's HE-AAC and HE-AAC v2 versions are regarded as today's most efficient general perceptual
14 audio codecs. AAC-ELD is considered the best-performing full-bandwidth communications codec by
15 independent studies and is widely deployed. AAC has been standardized by ISO and IEC as part
16 of the MPEG specifications.
17 
18 Patent licenses for necessary patent claims for the FDK AAC Codec (including those of Fraunhofer)
19 may be obtained through Via Licensing (www.vialicensing.com) or through the respective patent owners
20 individually for the purpose of encoding or decoding bit streams in products that are compliant with
21 the ISO/IEC MPEG audio standards. Please note that most manufacturers of Android devices already license
22 these patent claims through Via Licensing or directly from the patent owners, and therefore FDK AAC Codec
23 software may already be covered under those patent licenses when it is used for those licensed purposes only.
24 
25 Commercially-licensed AAC software libraries, including floating-point versions with enhanced sound quality,
26 are also available from Fraunhofer. Users are encouraged to check the Fraunhofer website for additional
27 applications information and documentation.
28 
29 2.    COPYRIGHT LICENSE
30 
31 Redistribution and use in source and binary forms, with or without modification, are permitted without
32 payment of copyright license fees provided that you satisfy the following conditions:
33 
34 You must retain the complete text of this software license in redistributions of the FDK AAC Codec or
35 your modifications thereto in source code form.
36 
37 You must retain the complete text of this software license in the documentation and/or other materials
38 provided with redistributions of the FDK AAC Codec or your modifications thereto in binary form.
39 You must make available free of charge copies of the complete source code of the FDK AAC Codec and your
40 modifications thereto to recipients of copies in binary form.
41 
42 The name of Fraunhofer may not be used to endorse or promote products derived from this library without
43 prior written permission.
44 
45 You may not charge copyright license fees for anyone to use, copy or distribute the FDK AAC Codec
46 software or your modifications thereto.
47 
48 Your modified versions of the FDK AAC Codec must carry prominent notices stating that you changed the software
49 and the date of any change. For modified versions of the FDK AAC Codec, the term
50 "Fraunhofer FDK AAC Codec Library for Android" must be replaced by the term
51 "Third-Party Modified Version of the Fraunhofer FDK AAC Codec Library for Android."
52 
53 3.    NO PATENT LICENSE
54 
55 NO EXPRESS OR IMPLIED LICENSES TO ANY PATENT CLAIMS, including without limitation the patents of Fraunhofer,
56 ARE GRANTED BY THIS SOFTWARE LICENSE. Fraunhofer provides no warranty of patent non-infringement with
57 respect to this software.
58 
59 You may use this FDK AAC Codec software or modifications thereto only for purposes that are authorized
60 by appropriate patent licenses.
61 
62 4.    DISCLAIMER
63 
64 This FDK AAC Codec software is provided by Fraunhofer on behalf of the copyright holders and contributors
65 "AS IS" and WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, including but not limited to the implied warranties
66 of merchantability and fitness for a particular purpose. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
67 CONTRIBUTORS BE LIABLE for any direct, indirect, incidental, special, exemplary, or consequential damages,
68 including but not limited to procurement of substitute goods or services; loss of use, data, or profits,
69 or business interruption, however caused and on any theory of liability, whether in contract, strict
70 liability, or tort (including negligence), arising in any way out of the use of this software, even if
71 advised of the possibility of such damage.
72 
73 5.    CONTACT INFORMATION
74 
75 Fraunhofer Institute for Integrated Circuits IIS
76 Attention: Audio and Multimedia Departments - FDK AAC LL
77 Am Wolfsmantel 33
78 91058 Erlangen, Germany
79 
80 www.iis.fraunhofer.de/amm
81 amm-info@iis.fraunhofer.de
82 ----------------------------------------------------------------------------------------------------------- */
83 
84 /***************************  Fraunhofer IIS FDK Tools  **********************
85 
86    Author(s):
87    Description: dit_fft ARM assembler replacements.
88 
89 ******************************************************************************/
90 
91 /* NEON optimized FFT currently builds only with RVCT toolchain */
92 
93 #ifndef FUNCTION_dit_fft
94 
95 /* If dit_fft was not yet defined by ARM-Cortex ... */
96 
97 #if defined(SINETABLE_16BIT)
98 
99 #define FUNCTION_dit_fft
100 
101 /*****************************************************************************
102 
103    date:   28.07.2005   srl
104 
105    Contents/description: dit-tukey-FFT-algorithm
106 
107 ******************************************************************************/
108 
109 #if defined(FUNCTION_dit_fft)
110 
111 
dit_fft(FIXP_DBL * x,const INT ldn,const FIXP_STP * trigdata,const INT trigDataSize)112 void dit_fft(FIXP_DBL *x, const INT ldn, const FIXP_STP *trigdata, const INT trigDataSize)
113 {
114     const INT n=1<<ldn;
115     INT i;
116 
117     scramble(x,n);
118     /*
119      * 1+2 stage radix 4
120      */
121 
122     for (i=0;i<n*2;i+=8)
123     {
124       FIXP_DBL a00, a10, a20, a30;
125       a00 = (x[i + 0] + x[i + 2])>>1;  /* Re A + Re B */
126       a10 = (x[i + 4] + x[i + 6])>>1;  /* Re C + Re D */
127       a20 = (x[i + 1] + x[i + 3])>>1;  /* Im A + Im B */
128       a30 = (x[i + 5] + x[i + 7])>>1;  /* Im C + Im D */
129 
130       x[i + 0] = a00 + a10;       /* Re A' = Re A + Re B + Re C + Re D */
131       x[i + 4] = a00 - a10;       /* Re C' = Re A + Re B - Re C - Re D */
132       x[i + 1] = a20 + a30;       /* Im A' = Im A + Im B + Im C + Im D */
133       x[i + 5] = a20 - a30;       /* Im C' = Im A + Im B - Im C - Im D */
134 
135       a00 = a00 - x[i + 2];       /* Re A - Re B */
136       a10 = a10 - x[i + 6];       /* Re C - Re D */
137       a20 = a20 - x[i + 3];       /* Im A - Im B */
138       a30 = a30 - x[i + 7];       /* Im C - Im D */
139 
140       x[i + 2] = a00 + a30;       /* Re B' = Re A - Re B + Im C - Im D */
141       x[i + 6] = a00 - a30;       /* Re D' = Re A - Re B - Im C + Im D */
142       x[i + 3] = a20 - a10;       /* Im B' = Im A - Im B - Re C + Re D */
143       x[i + 7] = a20 + a10;       /* Im D' = Im A - Im B + Re C - Re D */
144     }
145 
146     INT mh = 1 << 1;
147     INT ldm = ldn - 2;
148     INT trigstep = trigDataSize;
149 
150     do
151     {
152         const FIXP_STP *pTrigData = trigdata;
153         INT j;
154 
155         mh <<= 1;
156         trigstep >>= 1;
157 
158         FDK_ASSERT(trigstep > 0);
159 
160         /* Do first iteration with c=1.0 and s=0.0 separately to avoid loosing to much precision.
161            Beware: The impact on the overal FFT precision is rather large. */
162         {
163             FIXP_DBL *xt1 = x;
164             int r = n;
165 
166             do {
167                 FIXP_DBL *xt2 = xt1 + (mh<<1);
168                 /*
169                 FIXP_DBL *xt1 = x+ ((r)<<1);
170                 FIXP_DBL *xt2 = xt1 + (mh<<1);
171                 */
172                 FIXP_DBL vr,vi,ur,ui;
173 
174                 //cplxMultDiv2(&vi, &vr, x[t2+1], x[t2], (FIXP_SGL)1.0, (FIXP_SGL)0.0);
175                 vi = xt2[1]>>1;
176                 vr = xt2[0]>>1;
177 
178                 ur = xt1[0]>>1;
179                 ui = xt1[1]>>1;
180 
181                 xt1[0] = ur+vr;
182                 xt1[1] = ui+vi;
183 
184                 xt2[0] = ur-vr;
185                 xt2[1] = ui-vi;
186 
187                 xt1 += mh;
188                 xt2 += mh;
189 
190                 //cplxMultDiv2(&vr, &vi, x[t2+1], x[t2], (FIXP_SGL)1.0, (FIXP_SGL)0.0);
191                 vr = xt2[1]>>1;
192                 vi = xt2[0]>>1;
193 
194                 ur = xt1[0]>>1;
195                 ui = xt1[1]>>1;
196 
197                 xt1[0] = ur+vr;
198                 xt1[1] = ui-vi;
199 
200                 xt2[0] = ur-vr;
201                 xt2[1] = ui+vi;
202 
203                 xt1 = xt2 + mh;
204             } while ((r=r-(mh<<1)) != 0);
205         }
206         for(j=4; j<mh; j+=4)
207         {
208             FIXP_DBL *xt1 = x + (j>>1);
209             FIXP_SPK cs;
210             int r = n;
211 
212             pTrigData += trigstep;
213             cs = *pTrigData;
214 
215             do
216             {
217                 FIXP_DBL *xt2 = xt1 + (mh<<1);
218                 FIXP_DBL vr,vi,ur,ui;
219 
220                 cplxMultDiv2(&vi, &vr, xt2[1], xt2[0], cs);
221 
222                 ur = xt1[0]>>1;
223                 ui = xt1[1]>>1;
224 
225                 xt1[0] = ur+vr;
226                 xt1[1] = ui+vi;
227 
228                 xt2[0] = ur-vr;
229                 xt2[1] = ui-vi;
230 
231                 xt1 += mh;
232                 xt2 += mh;
233 
234                 cplxMultDiv2(&vr, &vi, xt2[1], xt2[0], cs);
235 
236                 ur = xt1[0]>>1;
237                 ui = xt1[1]>>1;
238 
239                 xt1[0] = ur+vr;
240                 xt1[1] = ui-vi;
241 
242                 xt2[0] = ur-vr;
243                 xt2[1] = ui+vi;
244 
245                 /* Same as above but for t1,t2 with j>mh/4 and thus cs swapped */
246                 xt1 = xt1 - (j);
247                 xt2 = xt1 + (mh<<1);
248 
249                 cplxMultDiv2(&vi, &vr, xt2[0], xt2[1], cs);
250 
251                 ur = xt1[0]>>1;
252                 ui = xt1[1]>>1;
253 
254                 xt1[0] = ur+vr;
255                 xt1[1] = ui-vi;
256 
257                 xt2[0] = ur-vr;
258                 xt2[1] = ui+vi;
259 
260                 xt1 += mh;
261                 xt2 += mh;
262 
263                 cplxMultDiv2(&vr, &vi, xt2[0], xt2[1], cs);
264 
265                 ur = xt1[0]>>1;
266                 ui = xt1[1]>>1;
267 
268                 xt1[0] = ur-vr;
269                 xt1[1] = ui-vi;
270 
271                 xt2[0] = ur+vr;
272                 xt2[1] = ui+vi;
273 
274                 xt1 = xt2 + (j);
275             }  while ((r=r-(mh<<1)) != 0);
276         }
277         {
278             FIXP_DBL *xt1 = x + (mh>>1);
279             int r = n;
280 
281             do
282             {
283                 FIXP_DBL *xt2 = xt1 + (mh<<1);
284                 FIXP_DBL vr,vi,ur,ui;
285 
286                 cplxMultDiv2(&vi, &vr, xt2[1], xt2[0], STC(0x5a82799a), STC(0x5a82799a));
287 
288                 ur = xt1[0]>>1;
289                 ui = xt1[1]>>1;
290 
291                 xt1[0] = ur+vr;
292                 xt1[1] = ui+vi;
293 
294                 xt2[0] = ur-vr;
295                 xt2[1] = ui-vi;
296 
297                 xt1 += mh;
298                 xt2 += mh;
299 
300                 cplxMultDiv2(&vr, &vi, xt2[1], xt2[0], STC(0x5a82799a), STC(0x5a82799a));
301 
302                 ur = xt1[0]>>1;
303                 ui = xt1[1]>>1;
304 
305                 xt1[0] = ur+vr;
306                 xt1[1] = ui-vi;
307 
308                 xt2[0] = ur-vr;
309                 xt2[1] = ui+vi;
310 
311                 xt1 = xt2 + mh;
312             }  while ((r=r-(mh<<1)) != 0);
313         }
314     } while (--ldm != 0);
315 }
316 
317 #endif /* if defined(FUNCTION_dit_fft)  */
318 
319 #endif /* if defined(SINETABLE_16BIT) */
320 
321 #endif /* ifndef FUNCTION_dit_fft */
322