Spurred on by Podcasting 2.0 and reflecting on my previous attempt at transcriptions, I thought it was time to have another crack at this. The initial attempts were basic TXT files that weren’t time-synced nor proofed and used a very old version of Dragon Dictate I had laying around.

This time around my focus is on making Causality as good as it possibly can be. From the PC2.0 guidelines:

SRT: The SRT format was designed for video captions but provides a suitable solution for podcast transcripts. The SRT format contains medium-fidelity timestamps and are a popular export option from transcription services. SRT transcripts used for podcasts should adhere to the following specifications.

Properties:

• Max number of lines: 2
• Max characters per line: 32
• Speaker names (optional): Start a new card when the speaker changes. Include the speaker’s name, followed by a colon.

This is closely related to defaults I found using Otter.ai but that’s not free if you want time-sync’d SRT files. So my workflow uses YouTube (for something useful)…

STEPS:

1. Upload episode directly converted from the original public audio file to YouTube as a Video (I use Ferrite to create a video export). Previously I was using LibSyn as part of their YouTube destination which also works.
2. Wait a while. It can take anywhere from a few minutes to a few hours, then go to your YouTube Studio, pick an episode, Video Details, under the section: “Language, subtitles, and closed captions”, select “English by YouTube (automatic)” three vertical dots, “Download” (NOTE BELOW). Alternatively select Subtitles, and next to DUPLICATE AND EDIT, select the three dots and Download, then .srt
3. If you can only get the SBV File: Open this file, untitled.sbv in a raw text editor, then select all, copy and paste it into: DCMP’s website, click Convert, select all, then create a new blank file: untitled.srt and paste in the converted format.
4. If you have the SRT now, and don’t have the source video (eg if it was created by LibSyn automatically, I didn’t have a copy locally) download the converted YouTube video using the embed link for the episode to: SaveFrom or use a YouTube downloader if you prefer.
5. Download the Video in low-res and put all into a single directory.
6. I’m using Subtitle Studio and it’s not free but it was the easiest for me to get my head around and it works for me. Open the SRT file just created/downloaded then drag the video for the episode in question onto the new window.
7. Visually skim and fix obvious errors before you press play (Title Case, ends of Sentences, words for numbers, MY NAME!)
8. Export the SRT file and add to the website and RSS Feed!

NOTE: In 1 case out of 46 uploads it thought I was speaking in Russian for some reason? The auto-translation in Russian was funny but not useful, but for all others it correctly translated automatically into English and the quality of the conversion is quite good.

I’ve also flattened the SRT into a fixed Text file, which is useful for full text search. The process for that takes me two steps:

1. Upload the file to Happy Scribe and select “Text File” as the output format.
2. Open the downloaded file in a text editor, select all the text and then go to Tool Slick’s line merge tool, pasting the text into the Input Text box, then “Join Lines” and select all of the Output Joined Lines box and paste over what you had in your local text file.
3. Rename the file and add to the website and RSS Feed!

As of publishing I’ve only done the sub-titles in SRT and TXT formats of two episodes, but I will continue to churn my way through them as time permits until they’re all done.

Of course you could save yourself a bit of effort and use Otter, and save yourself even more effort and don’t proof-read the automatically converted text. If I wasn’t so much of a stickler for detail, I’d probably do that myself but it’s that refusal to just accept that, that makes me the Engineer I am I suppose.

Enjoy!