Craig, Thanks, Gordon Date: Tue, 03 May 2016 02:22:37 +0000 From: Feng Yu Hong <fengyuhong@vimicro.com> To: "cdooley@apple.com" <cdooley@apple.com> Cc: "bluetooth-dev@lists.apple.com" <bluetooth-dev@lists.apple.com> Subject: Re: A2DP AAC codec bandwidth Message-ID: <38007BD70F89B94F9FDE5515BBC44E5A12A92E@exch-b01.vimicro.com> Content-Type: text/plain; charset="gb2312" Thank for your reply. I know AAC codec is perceptual codec, and subject listing test is best method of measuring perceptual quality. But lots customers take the codec bandwidth as an important objective evaluation index. Is it possible that measuring codec bandwidth reaches 20kHz on IOS 9.2 if setting MTU value is more than 895 byte? Yuhong ________________________________ 发件人: cdooley@apple.com [mailto:cdooley@apple.com] 发送时间: 2016年4月30日 8:34 收件人: Feng Yu Hong 抄送: bluetooth-dev@lists.apple.com 主题: Re: A2DP AAC codec bandwidth How are you measuring quality? I’m not a codec engineer, but I would imagine you need a psychoacoustic test to measure “quality”. Bitrate or sample rate alone can tell you the data pattern, but it’s very possible that the AAC encoder has improved over time and can now encode the same data with similar or better perceived quality at the same or lower bitrates. - Craig -- J. Gordon Rankin Owner and Chief Scientist ====== Wavelength Audio, ltd ====== _______________________________________________ Do not post admin requests to the list. They will be ignored. Bluetooth-dev mailing list (Bluetooth-dev@lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/bluetooth-dev/site_archiver%40lists.... This email sent to site_archiver@lists.apple.com I have to agree with Yuhon, it's an easy test to verify bandwidth and the greater bandwidth has been proven in many a AES articles to give better perceptual reproduction. But it's also easy to take an App like Faber Acoustical SignalSuite and send various fixed waveforms (sine at periodic or sweep) to an AAC (or SBC) endpoint and check quality with Ap or Prism tests sets for distortion and SNR values relative to OS's versions. A few months ago we had our MTU set low and testing proved poor response on our endpoint. We upped the value to an acceptable level and saw a huge improvement.