agentleFS
Sign inStart free
LinuxSuRenLinuxSuRen/my-dsh-skills/ohos-media-dev/SKILL.md

ohos-media-dev

Skill2 starsChanged 35 days agoMITView on GitHub

鸿蒙音视频应用开发方法论:Camera Kit 多摄并发(并发类型打开)、HCODEC 多实例与 buffer 输入、码率 int64 键、NV12 32 对齐、OHAudio 采集/播放、长时任务与熄屏、后台相机限制、HAL 超时与自愈,均为真机实测陷阱。

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    buffer 输入模式陷阱: Start 后不可清空 need-input 队列——codec 启动即投递缓冲,清空会"泄漏"已投递缓冲(codec 视为未归还不再投递),该路编码器永久饿死(push-enter 有 have=0 特征);重建路径的队列清理放在销毁之后

    line 22

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

鸿蒙(HarmonyOS)音视频应用开发方法论

1. 相机(Camera Kit)

  • 多摄并发:每颗相机独立 CameraInput+Session;必须查询 getCameraConcurrentInfos 并对所有相机(含第一颗)用其返回的并发类型 input.open(type) 打开——首颗若用普通 open() 独占,后续相机报 7400102 Device conflict
  • 并发不足的设备单路失败自动跳过,全部失败才算失败(降级而非报错)
  • preview profile 按"最接近目标"挑选;实际尺寸可能≠请求值,C 侧缓冲按实际 profile 尺寸分配(用两阶段:先建 surface,session 配置前 setFormat(实际宽高))
  • 采集 buffer 平面布局实测差异:planeCount=3 且 plane[1].offset > plane[2].offset 时是 NV21(拷贝需交换 UV);rowStride/columnStride 语义以真机为准,首个 buffer dump 一次确认
  • 会话打开偶发 7400103(Session not config),重试一次即过;HAL 调用全部加超时保护(stop/release/close 个别永不回调,不设超时会卡死状态机;深度休眠下给到 8s)

2. 视频硬编(HCODEC)

  • 多实例可用(双路 720p 实测),但不支持进程内二次配置同一实例——重配置走"销毁→重建"或退出应用重启;方向/分辨率切换用"保存配置→terminateSelf→重开自动恢复"模式
  • 码率是 int64 键:必须 OH_AVFormat_SetLongValue(OH_MD_KEY_BITRATE, ...),SetIntValue 写入编码器读不到,实测码率失控到 14Mbps;BITRATE_MODE_CBR 配套
  • 输出宽 32 字节对齐:720 宽(非 32 对齐)硬编实测底部大面积色度损坏(绿块);旋转后裁剪到对齐(704x1280)即恢复
  • buffer 输入模式陷阱:Start 后不可清空 need-input 队列——codec 启动即投递缓冲,清空会"泄漏"已投递缓冲(codec 视为未归还不再投递),该路编码器永久饿死(push-enter 有 have=0 特征);重建路径的队列清理放在销毁之后
  • 编码帧回调经 threadsafe function 送 ArkTS,帧携带相机编号

3. 音频(OHAudio / OH_AudioEncoder)

  • 采集:OH_AudioStreamBuilder(capturer)回调 OnReadData 是实时线程,只做拷贝入队;AAC 编码器在该机型 CreateByMime("audio/aac") 可能失败,回退 CreateByName("avenc_aac")
  • 音频编码器仅提供 legacy AVMemory 回调(SetCallback + onNeedInputData/onNeedOutputData),写入用 OH_AVMemory_GetAddr,提交用新式 PushInputData(index+attr)——两者混用是官方示例做法;编码器输出裸 AAC 帧(无 ADTS)
  • G.711 旁路:采集 48kHz(48000/8000=6 整数倍)→ 混单声道 → FIR 低通(33 阶 sinc,截止 3.4kHz)再 6:1 抽取;裸均值抽取无抗混叠,高频折叠回语音频段听感沙哑
  • 播放(对讲):renderer 用回调拉取式(SetRendererWriteDataCallback),环形缓冲存收到的 PCMA、回调里 A-law 解码填充,无数据填静音;满则丢最旧保实时

4. 后台与熄屏

  • 长时任务必须双管齐下:backgroundTaskManager.startBackgroundRunning(DATA_TRANSFER) + module.json5 里 EntryAbility 声明 "backgroundModes": ["dataTransfer"]——缺声明注册失败(9800005 bgMode invalid),熄屏即被冻结(freezer cgroup 可查)
  • 后台相机是系统隐私限制:应用退后台后 CreateCameraInput 返回 null/会话被回收,长时任务也无法豁免——不做强行支持,动态适配:后台暂停重试(避免相机设备半开楔死 7400201,该状态仅进程退出可解除),回前台自动重建管线恢复
  • 帧停滞看门狗分级自愈:单次停滞重开会话 → 连续多次全量重建管线(采集面+编码器+会话,与停止/启动同路径可重复执行)

5. 工程杂项

  • 权限:相机/麦克风都是 user_grant,UIAbility 里 requestPermissionsFromUser 一次申请;麦克风不可用时功能降级而非失败
  • hilog 不支持 %f,浮点日志乘 100 取整用 %d,否则输出乱码
  • napi 垫片模式:C 层管采集/旋转/编码/音频全链,ArkTS 只做服务与 UI;按相机多实例(实例数组),全局单例改多实例时注意 listener context 指向各自实例
  • 系统返回键:onBackPress 返回 true 拦截 + moveAbilityToBackground 进后台;退出用显式关闭按钮(停止推流+长时任务+terminateSelf)

6. 实测陷阱速查

现象根因
码率失控 14MbpsBITRATE 用 SetIntValue(int64 键读不到)
竖屏底部绿块输出宽 720 非 32 对齐,硬编色度损坏
单路编码器零输出Start 后清了 need-input 队列,codec 缓冲泄漏饿死
第二颗相机打不开(7400102)首颗未用并发类型打开,独占设备
熄屏流断/进程冻结module.json5 缺 backgroundModes 声明(9800005)
后台回来相机全失效系统回收后台相机会话 → 看门狗重建
相机反复 7400201后台期间反复重开致设备半开,仅进程退出解除
AAC 编码器创建失败CreateByMime 不支持,回退 CreateByName(avenc_aac)
G.711 人声沙哑抽取无抗混叠滤波

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free