Skip to main content

Hi. I have audio and video lectures that I want to be able to easily navigate to find and listen/watch portions where specific topics are being talked about.



As an example, if it is a video about a chef talking about how to cook a steak, I would like to go in and manually tag the various things that are happening and different times during the video (e.g. “Talking about Meat Quality”, “Knife Skills”, “Temperature Controls”, “Chef Making a Joke”, etc.). Also, note that multiple tags can happen at the same time - i.e. Chef could be cutting meat while making a joke. It would be like “Chef Making a Joke” e1min 31sec to 2min 58 sec into the video], “Knife Skills” e1min 53sec to 2min 32sec], etc.



End goal is that I can have many audio and video files inside AirTable, and each tagged with topics being discussed within each. I should then be able to perform searches on tags and be able to directly go to the section of the audio(s)/video(s) file(s) where that tag is being talked about.



NOTE: In a way, YouTube’s “Chapters” feature come close to this, but what I am talking about is more advanced since multiple tags can happen in the same time period. Unlike “Chapters”, that do not overlap.



Any thoughts on how the end-goal can be achieved in AirTable?

In theory, the video snippets app should let you play video attachments based on a start time and end time, but I’ve never been able to get it to work properly. Maybe they’ve fixed it since the last time I played with it.



Also, Airtable has very low attachment limits (only 20 GB for your entire base on the Pro Plan), so if you’re planning on storing a lot of video attachments, I’m not sure if Airtable is the right tool for this sort of task.



You also mentioned audio files. I don’t know if those would work with the video snippets app or not.



Bad idea (as Scott said). You should consider a sustainable and vast data store for video because these files are not going to get smaller; only larger and vastly more of them. Also consider Airtable records as references to large binary files. The value of your service is (a) understanding and tagging the video, and (b) making it easy to connect them to that content. There is no value to storing the video a second time in Airtable.





You would be far better off leaving the tagging to the world of AI. If you want to scale, all tagging triage should be done through automation. Then, the second phase should involve humans to polish the tagging.

Also, to build upon what Bill said, there are probably highly-specific apps out there that are specifically designed to do what you want to do. Airtable is not well-suited for what you are looking to do.

Thanks, Scott and Bill.



RE: Saving Files in AirTable - I agree, saving files inside AirTable can quickly become a lack of space issue. That said, many of the videos are actually YouTube based, so I am sure with the right formula in the cell they could play at specific locations.



RE: AI Tagging - Bill, you’re absolutely right. The value of my time in the exercise would be the tagging element. AI would definitely be better, but for that I’ll either need to hire someone or learn it myself lol



Thank you, both, for you speedy feedback!



All said, I feel that if AirTable had a way to tag timeline of media (whether audio or video) it could be very powerful. Is there a way to request features?

Thanks, Scott and Bill.



RE: Saving Files in AirTable - I agree, saving files inside AirTable can quickly become a lack of space issue. That said, many of the videos are actually YouTube based, so I am sure with the right formula in the cell they could play at specific locations.



RE: AI Tagging - Bill, you’re absolutely right. The value of my time in the exercise would be the tagging element. AI would definitely be better, but for that I’ll either need to hire someone or learn it myself lol



Thank you, both, for you speedy feedback!



All said, I feel that if AirTable had a way to tag timeline of media (whether audio or video) it could be very powerful. Is there a way to request features?




What you’re describing is currently possible with Airtable, not in Airtable. And since this is a very specific use case that’s not generally considered a mainstream business case, it’s unlikely that even if requested, the features would probably not rise to the importance you feel it should.



There are many specific-purpose systems available today that are designed for media tagging, storage, and playback. Typically, these come with media server capabilities because without them, consuming the media is not pleasant for end users. This is yet an additional reason to not store the video in Airtable. Aside from the storage ceiling, it is not a streaming server and it lacks the ability to dynamically control bitrate when rendering the audio or bitrate over different connection speeds.





This is probably not the case. Timelines in video require precision time protocol to make the tagging truly useful in a number of use cases. The best you can do is seconds and this is generally considered a poor substitution for true video platforms. It might be good enough for your envisioned use case.



I recommend you take a look at the Ant Media Server; it’s an open source platform for building very powerful media services and it supports tagging. I would guess mapping this into Airtable would be possible.



You might also want to check out LionBridge.ai. If they’re too expensive, learning the approach will give you ideas for chasing this idea. Lastly, Cloudinary offers a free tier for building new services like this.



What you’re describing is currently possible with Airtable, not in Airtable. And since this is a very specific use case that’s not generally considered a mainstream business case, it’s unlikely that even if requested, the features would probably not rise to the importance you feel it should.



There are many specific-purpose systems available today that are designed for media tagging, storage, and playback. Typically, these come with media server capabilities because without them, consuming the media is not pleasant for end users. This is yet an additional reason to not store the video in Airtable. Aside from the storage ceiling, it is not a streaming server and it lacks the ability to dynamically control bitrate when rendering the audio or bitrate over different connection speeds.





This is probably not the case. Timelines in video require precision time protocol to make the tagging truly useful in a number of use cases. The best you can do is seconds and this is generally considered a poor substitution for true video platforms. It might be good enough for your envisioned use case.



I recommend you take a look at the Ant Media Server; it’s an open source platform for building very powerful media services and it supports tagging. I would guess mapping this into Airtable would be possible.



You might also want to check out LionBridge.ai. If they’re too expensive, learning the approach will give you ideas for chasing this idea. Lastly, Cloudinary offers a free tier for building new services like this.


Thanks, Bill. I’ll explore the services you’ve identified. thumbs up

Reply


Terms & Conditions