PrecisEQ Repository Generator

English | 中文

---

Simply upload the frequency response measurement data (CSV) of headphones to automatically generate a dedicated online headphone calibration repository for PrecisEQ. Calibration IRs are generated using AutoEq.

Repository Address

After you upload the CSV files for the first time and successfully generate the headphone calibration files, the following URL will be automatically replaced with your own repository address. Enter it in Headphone Calibration -> Headphone model -> Repo -> Import:

https://raw.githubusercontent.com/yokodev-pro/PrecisEQ-Repository-Generator/main/RepositoryFiles/

How to Use

Step 1: Fork and customize repository

1. Click Use this template at the top right to create your own Public repository.
2. Modify the repository information in RepositoryFiles/repo_info.json.

Step 2: Upload measurement data

1. Collect headphone frequency response measurement files (.csv format), and name them according to the rule: Brand Model.csv, for example Sony IER-Z1R.csv, Sennheiser IE 900.csv.
2. Enter the corresponding folder in the repository based on the headphone type:
   • In-Ear: measurements/0_in-ear/
   • Over-Ear (Open-Back): measurements/1_open-back/
   • Over-Ear (Closed-Back): measurements/2_closed-back/
3. Click Add file -> Upload files, drag and drop the .csv, and submit (Commit changes).
4. If you want to update measurement data for a headphone, just upload a new CSV file with the same name to overwrite the old file.

Step 3: Wait for processing

• After submitting the files, GitHub will process them automatically. You can click the Actions tab at the top of the repository to check progress.
• After processing is complete (usually takes a few minutes), the generated files will be stored in the RepositoryFiles folder, which PrecisEQ will download for headphone calibration. The images folder will be used to store the generated frequency response graphs; these images are for manual inspection only and PrecisEQ will not download them.
• Restart PrecisEQ to sync the repository content.

Optional Steps

• Target Curve Compensation: If the ear simulator used as the data source has measurement results that differ significantly from the B&K 4195 (In-Ear) and GRAS RA0045+KB0065/66 (Open-Back Over-Ear) + GRAS 45CA (Closed-Back Over-Ear), it is recommended to replace (In-Ear) targets/0_zero.csv/, (Open-Back Over-Ear) targets/1_zero.csv/, and (Closed-Back Over-Ear) targets/2_zero.csv/ with compensation curves from the data source's ear simulator to the aforementioned ones. This ensures compatibility with the user's saved target curves. You can obtain the compensation curve by measuring multiple headphones under the official repository's Flat Target preset, averaging and then inverting the curves.
• Chinese Localization: If you want to add Chinese localization for brand and model names, add the Chinese name to the second item in the list in RepositoryFiles/headphone_list.json:

  "brandName": ["Brand", "品牌名"],
  "modelName": ["Model", "型号名"]

• Subjective Volume Correction: If there is a significant difference in subjective volume perception before and after calibrating a headphone, modifying the noDspOffsetDb field in RepositoryFiles/headphone_list.json can increase/decrease the volume before calibration, in dB.

Directory Description

• measurements/ - Stores headphone frequency response measurement data (CSV).
• RepositoryFiles/ - Automatically generated PrecisEQ repository folder, containing repository description JSON and IR files.
• images/ - Automatically generated headphone and calibration frequency response graphs.
• scripts/ - Python backend operational scripts for automated processing.

---

只需要简单地上传耳机的频率响应测量数据（CSV），就能自动化生成 PrecisEQ 专用的在线耳机校准仓库。校准 IR 使用 AutoEq 生成。

仓库地址

首次上传 CSV 文件并成功生成耳机校准文件后，以下链接会自动替换为你的仓库地址。在 耳机校准 -> 耳机型号 -> 仓库 -> 导入 填入：

https://raw.githubusercontent.com/yokodev-pro/PrecisEQ-Repository-Generator/main/RepositoryFiles/

如何使用

第一步：获取和自订仓库

1. 点击右上角的 Use this template 创建一个你自己的 公开（Public） 仓库。
2. 修改 RepositoryFiles/repo_info.json 中的仓库信息。

第二步：上传测量数据

1. 收集耳机频响测量文件（.csv 格式），并按以下规则命名：品牌 型号.csv，例如 Sony IER-Z1R.csv、Sennheiser IE 900.csv。
2. 根据耳机的类型，进入仓库里的对应文件夹：
   • 入耳式：measurements/0_in-ear/
   • 头戴式（开放）：measurements/1_open-back/
   • 头戴式（封闭）：measurements/2_closed-back/
3. 点击 Add file -> Upload files，把 .csv 拖入并提交（Commit changes）。
4. 如果你要更新某款耳机的测量数据，上传同名的新 CSV 文件覆盖旧文件即可。

第三步：等待处理

• 提交文件后，GitHub 会自动处理。你可以点击仓库顶部的 Actions 标签查看进度。
• 处理完成后（通常需要几分钟），生成的文件会存放在 RepositoryFiles 文件夹，PrecisEQ 将下载这些文件用于耳机校准；images 文件夹用于存放生成的频响曲线图，这些图片仅用于人工核查，PrecisEQ 不会下载这些图片。
• 重新启动 PrecisEQ 即可同步仓库内容。

可选步骤

• 目标曲线补偿：如果作为数据来源的人工耳，其测量结果与 B&K 4195（入耳式）和 GRAS RA0045+KB0065/66（开放头戴式）+GRAS 45CA（封闭头戴式）有较大出入，建议将（入耳式）targets/0_zero.csv/、（开放头戴式）targets/1_zero.csv/、（封闭头戴式）targets/2_zero.csv/ 替换为数据源的人工耳到上述人工耳的补偿曲线，以兼容用户已经保存的目标曲线。可通过测量多个耳机在官方仓库预设的 Flat Target 下的频响曲线，平均并取反后得到补偿曲线。
• 中文本地化：如果你想为品牌名、型号名添加中文本地化，在 RepositoryFiles/headphone_list.json 把中文添加到列表第二项：

  "brandName": ["Brand", "品牌名"],
  "modelName": ["Model", "型号名"]

• 主观音量校正：如果某个耳机校准前后的主观音量感受有明显差异，修改 RepositoryFiles/headphone_list.json 中的 noDspOffsetDb 字段可以增大/减小校准前的音量，单位为dB。

目录说明

• measurements/ - 存放耳机频率响应测量数据（CSV）。
• RepositoryFiles/ - 自动生成的 PrecisEQ 仓库文件夹，包含仓库描述 JSON 和 IR 文件。
• images/ - 自动生成的耳机和校准频响图。
• scripts/ - 后台运行的 Python 自动化处理脚本。