Preparing data for bulk-uploading

If you have many talks to upload, Dharmaseed can upload the talks for you. For every talk to be uploaded, we ask that you provide:

1. An mp3 audio file of the recording

We prefer 64kbps to 96kbps mono .mp3 files, but can handle other bit-rates and stereo recordings if necessary. In general, we prefer to use the highest quality recordings you have available (we will downsample them if necessary). If all you have is .wav files, we can accept them as well.

File names can be arbitrary, they will be renamed to a standardized template by our system. MP3 tags are also not necessary, they will be added by our system based on the metadata (below).

2. The metadata for each talk

In order to enter the talk into our database, we will need a text file for each talk with the following pieces of information ("fields"), each placed on a single line. The format of each line is:

<field name> : <field contents>

Here are the list of fields that should be provided:

  • date: YYYY-MM-DD (the recording date)
  • time: HH:MM (the time of day when the recording started, if available, in 24 hour HH:MM format)
  • teacher: Teacher name
  • title: The title of the talk
  • description: Talk description (if any)
  • venue: Venue name or code (e.g. Forest Refuge, IMSRC, SR, etc. If you don't know the code, just spell out the venue and we'll assign a code for it)
  • publishability: "retreatants-only" or "public" (if not given, "public" is assumed)
  • language: Spoken language the talk is given in (if not given, English is assumed. This field can also be a description, such as "English with Italian translation")

The information for each talk is placed in its own file using the same file name stem as the mp3 file, but with a ".txt" extension. So, if your audio file is named: opening_talk_july_3_08.mp3 your text metadata file should be named: opening_talk_july_3_08.txt

Here's an example of what one of these metadata text files look like:

date: 2008-07-07 time: 17:00 teacher: Larry Rosenberg title: The world exists to set us free III venue: CIMC publishability: retreatants-only description: {Life is to be lived. Learning how to live skillfully in a world that is challenging us by changing in uncertain ways.}

3. The metadata for each retreat

All the .mp3 and .txt files for each retreat should be placed in their own retreat folder. The retreat folder name is not significant, but it is usually helpful to include the retreat name and/or date to reduce confusion. The retreat folder must contain a text file named "retreat.txt".

The retreat.txt file will contain the name and date of the retreat. Here's an example:

date: 2008-07-07 name: Stillness and Insight

4. The metadata for each collection

The Dharmaseed website allows talks to be grouped into collections, such as a CD full of chants that should be considered as an "album" of material. (Do not use this to group a collection of talks in a retreat, that is done by using the "retreat.txt" file, specified above).

Collections are created by placing the collection's title as well as its other applicable metadata such as the teacher, description, publishability, etc. in a standard talk metadata file (described above), but with a .collection rather than .txt extension. The filename's stem (the part before the extension) is ignored.

Within the .collection file, the collection field specifies a comma-separated list of the component pieces, in the order they are expected to appear. Here is an example:

collection: 3_06a_TB_Meditation.mp3, 3_06b_TB_Karma_Intro.mp3, 3_06c_TB_The_Way_Out-Doing.mp3, ...


  • You may provide titles and descriptions in other languages. In this case, please ensure that text file is in Unicode UTF-8 format in order to preserve special accent characters. To do this using e.g. Microsoft Word, select 'Save As' then, in the 'Save as type' field select "Plain Text (*.txt)" and click "Save". This will pop up a further window where you should select "Other encoding:" and "Unicode (UTF-8)". Also select the checkbox "Allow character substitution".
  • To provide titles or descriptions in more than one language (if, for example, the talk was given in English with an Italian translation), please place the English translation of the title in square brackets []. So a Spanish and English title would be written: title: Meditación De La Respiración [Meditation on the breath].
  • Extra spaces and lines may be used freely without harm.
  • Field names are case-insensitive.
  • The field contents should be placed on a single line. If you need to make the contents extend across line endings, use curly braces {} to indicate the start and end of the field, as in the description field in the example above.
  • The "time", and "publishability" fields are optional. If the time is unset, we'll use 6pm. If publishability is unset, we'll use "public".
  • If a retreat with the given date and name does not yet exist in the database it will be added during the upload.
  • If a teacher's name is not in the Dharmaseed database it will be added during the upload (in the case of a new teacher, please supply their contact email address in separate email).
  • Venue should generally indicate the sponsoring organization. E.g. Spirit Rock conducted talks given at Yucca Valley should still be listed as Spirit Rock. If the venue does not exist, it will be added to the Dharmaseed database (in the case of new venues, please also provide their location and website URL if available).
  • It is easiest for us to work with filenames that have no spaces in them (use, e.g. "_" instead of space). However, if this requires you to do a great deal of manual work, we can change the spaces to "_" for you.

Sending us your files for uploading

1. Via mail

Please place your files on CD or DVD media and contact us at for our postal address.

2. Via filesharing

a) Upload to a file-sharing service like WeTransfer, Dropbox, Box, etc, and send us a link

b) Making your files available to us through an FTP server: If you have the capability of placing your finished .mp3 and .txt files on an FTP or HTTP server, we can fetch them directly. In this case please email us the URLs and any account name and password necessary for access.

Thank you for contributing your talks to !