Abstract
Yüksek kaliteli ses işleme, ses üzerinde uygulanan efektin sebep olduğu işleme gecikmesini düşürmek için donanım hızlandırması gerektirir. Ses girişinden gelişmiş ses oluşurken hesaplanan gecikme, önemli bir gecikme bileşenidir ve özellikle sanatçıların canlı performansında veya yüksek kaliteli ses üretiminde performansı etkiler. Sanatçılar, müziklerine bir ses efekti eklemek ister ve bu efektler gerçek zamanlı sistemlerde kullanıldığında gecikme ana problem haline gelir. CPU tabanlı sistemler esneklik sunar, ama işleme esnasında oluşan büyük miktardaki gecikme sanatçıları olumsuz olarak etkiler. Bu çalışmada, esnekliğin yazılımla ve hızlandırmanın donanım özelleştirmesi ile elde edildiği, ses efektleri için yazılım/donanım ortak tasarım yöntemi ile bir sistem üzerinde çip (SoC) çözümü sunuyoruz. MATLAB üzerinden ardışık düzen uygulayarak gecikmeyi azaltıyor ve frekansı artırıyoruz. Sistem, çift çekirdekli ARM Cortex A9 işlemcisi ve ZC7020 Zynq çipi barındıran programlanabilir SoC platformu ZedBoard’da çalıştırılmış ve test edilmiştir. ARM işlemcisi, FPGA üzerinde çalıştırılan ses efekti donanım hızlandırıcısının yönetilmesine ve kullanıcı ile haberleşilmesine olanak sağlar. Ses efekti MATLAB & Simulink tarafından sağlanan blok modeller ile tasarlanmıştır. HDL Coder, bu blokları RTL seviyesinde donanım tasarımlarına dönüştürür. MATLAB & Simulink tarafından sağlanan bu tasarım yöntemi, yüksek hızlı donanım yazmaçlarının sağladığı hızdan yararlanmak için FPGA’ya RTL seviyesinde gömülebilen yüksek seviyeli blok tasarımı yapılmasına ve yazılım esnekliğinden yararlanmak için AXI ara bağlantısı ile arayüz oluşturulmasına olanak verir. Bu çalışma gecikmenin önemli bir ölçüde düşürüldüğünü gösterir.
References
- Y. E. ESEN and İ. SAN, “LowLAG: Low-latency hardware accelerator of a sound effect with system-on-chip design,” in ASYU 2020, Innovations in Intelligent Systems and Applications Conference, İstanbul, 2020.
- Nicolas Juillerat, Stefan Muller Arisona, Simon Schubiger-Banz, “REAL-TIME, LOW LATENCY AUDIO PROCESSING IN JAVA,” in Proceeding of the International Computer Music Conference, 2007.
- U. Meyer-Baese, Digital Signal Processing with Field Programmable Gate Arrays, 2nd ed., Springer, Berlin, Heidelberg, 2007.
- Yamaha, “Audio quailty on networked systems,” Yamaha, [Online]. Available: https://uk.yamaha.com/en/products/contents/proaudio/docs/audio_quality/05_audio_quality.html. [Accessed 25 06 2020].
- Markus Pfaff, David Malzner, Johannes Seifert, Johannes Traxler, Horst Weber, Gerhard Wiendl, “IMPLEMENTING DIGITAL AUDIO EFFECTS USING A HARDWARE/SOFTWARE,” in 10th Int. Conference on Digital Audio Effects (DAFx-07), Bordeaux, France, 2007.
- Kyungjin Byun, Young-Su Kwon, Seongmo Park, and Nak-Woong Eum, “Digital Audio Effect System-on-a-Chip Based on Embedded DSP Core,” in ETRI Journal, 2009.
- Crockett, Louise H.; Elliot, Ross A.; Enderwitz, Martin A.; Stewart, Robert W.;, The Zynq Book: Embedded Processing with the ARM Cortex-A9 on the Xilinx Zynq-7000 All Programmable SoC, First ed., Strathclyde Academic Media, 2014.
- I. The MathWorks, “HDL Coder™ Getting Started Guide,” March 2020. [Online]. Available: https://www.mathworks.com/help/pdf_doc/hdlcoder/hdlcoder_gs.pdf. [Accessed 25 06 2020].
- I. The MathWorks, “Limit input signal to the upper and lower saturation values,” [Online]. Available: https://www.mathworks.com/help/releases/R2018b/simulink/slref/saturation.html. [Accessed 25 06 2020].
- I. The MathWorks, “Fixed-Point Conversion,” in HDL Coder™ User's Guide, The MathWorks, Inc., 2020, pp. 117-162. The MathWorks, Inc, “Signal Processing Toolbox™ User's Guide,” 03 2021. [Online]. Available: https://www.mathworks.com/help/pdf_doc/signal/signal.pdf. [Accessed 12 03 2021].
- I. The MathWorks, “Multiply input by constant,” [Online]. Available: https://www.mathworks.com/help/releases/R2018b/simulink/slref/gain.html. [Accessed 25 06 2020].
- Louise H. Crockett, Ross A. Elliot, Martin A. Enderwitz, David Northcote, “Zynq Book Tutorials,” Aug 2015. [Online]. Available: http://www.zynqbook.com/download-tuts.html. [Accessed 25 06 2020].
Abstract
High-quality sound processing requires hardware acceleration in order to reduce the processing latency of the applied effect over the sound. Computational latency of creating enhanced sound from the audio input is an important delay component and affects the performance especially in artists’ live performance or high-quality sound generation. Artists want to apply a sound effect on their music and latency is the main problem when these systems running in real-time. CPU-based systems present flexibility, but introduce a high amount of latency while processing, which in fact affects the artist negatively. In this study, to get the flexibility through software and the acceleration via hardware specialization, we present a system-on-chip (SoC) solution with HW/SW co-design methodology for sound-effects. We reduce the latency and increase the frequency by applying pipelining through MATLAB. The system is implemented and tested on programmable SoC platform, ZedBoard, which contains ZC7020 Zynq chip with a dual-core ARM-Cortex-A9 processor. The ARM processor enables the management of sound-effect hardware accelerator running on FPGA and communication with user. Sound effect is designed with block models provided by MATLAB & Simulink. HDL Coder converts these blocks into RTL-level hardware designs. The followed design methodology provided by MATLAB & Simulink enables high-level block design that can be embedded into FPGA at RTL-level to benefit from the speed provided by high-speed hardware registers and to have an AXI interconnect interfacing with software in order to utilize the software flexibility. The study shows that latency is reduced significantly.
References
- Y. E. ESEN and İ. SAN, “LowLAG: Low-latency hardware accelerator of a sound effect with system-on-chip design,” in ASYU 2020, Innovations in Intelligent Systems and Applications Conference, İstanbul, 2020.
- Nicolas Juillerat, Stefan Muller Arisona, Simon Schubiger-Banz, “REAL-TIME, LOW LATENCY AUDIO PROCESSING IN JAVA,” in Proceeding of the International Computer Music Conference, 2007.
- U. Meyer-Baese, Digital Signal Processing with Field Programmable Gate Arrays, 2nd ed., Springer, Berlin, Heidelberg, 2007.
- Yamaha, “Audio quailty on networked systems,” Yamaha, [Online]. Available: https://uk.yamaha.com/en/products/contents/proaudio/docs/audio_quality/05_audio_quality.html. [Accessed 25 06 2020].
- Markus Pfaff, David Malzner, Johannes Seifert, Johannes Traxler, Horst Weber, Gerhard Wiendl, “IMPLEMENTING DIGITAL AUDIO EFFECTS USING A HARDWARE/SOFTWARE,” in 10th Int. Conference on Digital Audio Effects (DAFx-07), Bordeaux, France, 2007.
- Kyungjin Byun, Young-Su Kwon, Seongmo Park, and Nak-Woong Eum, “Digital Audio Effect System-on-a-Chip Based on Embedded DSP Core,” in ETRI Journal, 2009.
- Crockett, Louise H.; Elliot, Ross A.; Enderwitz, Martin A.; Stewart, Robert W.;, The Zynq Book: Embedded Processing with the ARM Cortex-A9 on the Xilinx Zynq-7000 All Programmable SoC, First ed., Strathclyde Academic Media, 2014.
- I. The MathWorks, “HDL Coder™ Getting Started Guide,” March 2020. [Online]. Available: https://www.mathworks.com/help/pdf_doc/hdlcoder/hdlcoder_gs.pdf. [Accessed 25 06 2020].
- I. The MathWorks, “Limit input signal to the upper and lower saturation values,” [Online]. Available: https://www.mathworks.com/help/releases/R2018b/simulink/slref/saturation.html. [Accessed 25 06 2020].
- I. The MathWorks, “Fixed-Point Conversion,” in HDL Coder™ User's Guide, The MathWorks, Inc., 2020, pp. 117-162. The MathWorks, Inc, “Signal Processing Toolbox™ User's Guide,” 03 2021. [Online]. Available: https://www.mathworks.com/help/pdf_doc/signal/signal.pdf. [Accessed 12 03 2021].
- I. The MathWorks, “Multiply input by constant,” [Online]. Available: https://www.mathworks.com/help/releases/R2018b/simulink/slref/gain.html. [Accessed 25 06 2020].
- Louise H. Crockett, Ross A. Elliot, Martin A. Enderwitz, David Northcote, “Zynq Book Tutorials,” Aug 2015. [Online]. Available: http://www.zynqbook.com/download-tuts.html. [Accessed 25 06 2020].
Details
| Primary Language | English |
|---|---|
| Subjects | Engineering |
| Journal Section | Research Articles |
| Authors | |
| Publication Date | December 30, 2021 |
| Published in Issue | Year 2021 Volume: 33 - ASYU 2020 Özel Sayısı |