Start a new topic

Upcoming Webcast & QA Chat: Learn about (and participate in) Amara's Town Hall


A special thanks for all the wonderful Amara community volunteer captioners, translators, subtitlers, promoters and sharers who joined us in the live broadcast.

We'd love to get more feedback about the new Amara platform mockups we shared on the October 23rd, 2012 Townhall. 

The video:
With Communication Access Realtime Transcription (CART) provided by Jason Bradley from
 (Thank you so much Jason and Michael!)  
The slides with the mockups of the new Amara look are here: 

IMPORTANT INFO Before adding more ideas:

This forum is an "idea" forum. Every separate post can be voted up or down (and discussed), we can also track the progress of the ideas, to see which ones will be implemented and see how they are progressing. We suggest:

  • Post one idea per topic.
  • Add really descriptive titles.
  • Vote ideas up instead of duplicating, feel free to expand through replies.

Updated October 23rd.


The Amara Town Hall: happens today at 3pm EDT, (time zones here).

Participating is easy – just go to the Amara YouTube channel, where we'll be broadcasting a live Google Hangout:


Jules and I will be doing a presentation of some new Amara functionality, followed by a question and answer session (both live, from comment stream, and from this forum post). We're working on having live transcription during the event, and will also have a recorded video and transcript available, after the event.


Looking forward to meeting you soon!





And the Amara Team





We have set a date for the Town Hall: Tuesday, October 23rd, 2012 at 3pm ET, (time zones here).

We will be using a combination of Google Hangouts and broadcast the Meeting through a YouTube channel ( live, where the video will remain for those who can’t make it to the live broadcast. This is the first time we'll be doing something like this, we're very excited! We're doing our best to coordinate live transcription for the deaf and hard of hearing, more details to follow on that (but do get in touch if you'd be interested in helping with live translation or transcription!).

Thank you all for participating in this conversation and giving us your feedback on this forum, there have been many great ideas and suggestions. I’m sure the Town Hall will be a great opportunity to address them and also to share the plans for Amara, giving you a sneak peak on what is coming next.

Hope to see you in the Town Hall on Tuesday!


And the Amara Team



Hello community members!

We’re planning a town-hall style meeting that will happen in a few weeks. Ahead of the meeting, we’d like to compile community questions and feedback. We’ll announce the timing of the meeting, and in case you’re not able to make it live, we’ll post a recorded webcast and transcript.

Be sure you’re on the Amara mailing list to get updated once the time is finalized:

WHO: The backbone of Amara is the strength of it’s volunteer community, and we want to ask for your help in making it better. You are the ones who use it regularly, know it best, and we want to hear what you have to say.

WHY: Amara has been growing in the past few months, and trying to find ways to work with many different types of communities. We’ve experienced some growing pains but are incredibly excited about the enthusiasm so many people and organizations have shown towards captioning, subtitling, and translation. We have some exciting technology that we’ve started working on – we’ll use this meeting to share our progress and get your feedback.

WHAT: If you want your questions or issues addressed in this Town Hall, please write them as a reply to this forum post in the best detail you can as to what your issue is and why the issue is important for the wider Amara Community. This will mainly be a place for us to gather information for what the community wants or needs, so let’s please keep it on topic. We would like to use this thread for getting out all the potential questions and issues, but not for unpacking/discussing those (feel free to start additional threads to dive in deep).

HOW: We’re still figuring out the specifics, but we’ll pick a date/time and then log onto something like Google Plus, IRC, and a shared presentation – we’ll provide a real-time text transcript (it’ll be an Amara staff member typing as fast as they can in a chat channel!), but if anyone is interested in doing sign language interpretation (via webcam) or CART transcription, we’d be delighted!

WHEN: We’ll propose a few dates in the next few days and see which works best for the most people. Make sure you’re on the Amara mailing list to get updated once a date is finalized.

Lastly, we wanted to thank you for helping make the internet a more accessible place by using Amara and creating subtitles for the entire world. We want specifically acknowledge users like Claude and Verone-- people that took initiative, and stepped up to give other users a hand by being active in the forums, answering other users’ questions and bringing issues to our attention when unexpected changes to Amara took place. Thank you very much, we truly appreciate it.

1 person likes this idea

Thanks for the Town Hall meeting: it was great. Next time hopefully we can hear you too, Jules, but thanks for the replies you gave in the comments / chat.

Video recording: (nitty-gritty starts at ca 14:00+). The slides Dean refers to are at

Video language

(I forgot that before)
  • It should be possible to indicate several languages for a multilingual videos (check boxes instead of radio buttons)
  • there should be an "other" option with a text box for languages not thought of by the developers.

If the above is not feasible, then developers should not give such importance to the video language indication. And if the above is feasible, anyway, users should be able to correct these indications themselves.

Update: Example of bilingual video

The Amara software identified of its own bat the video language for Catch the Ice Dude (added to the Captions Requested team) as English. However it is bilingual German - English, and actually mainly in German.

With a check box list, both languages could be indicated. And if users were empowered to modify this indication themselves, they could do it, thus making the video findable by German native speakers.

Up-Update (Oct. 21): Catch the Ice Dude has now been moved to the Volunteer team, where there are German native speakers.
Bump (event timing added to main post)
Hi Diomedes,

Also there must be a clear clarification on the matter of saving the work. For example when it is saved, when it can be considered final revision and when not etc.I would like to propose the following1) Use a subversion system. Like 1.1, 1.2, 1.3 etc. And when the checkbox is pressed as a final translation/transcription, then move to the next major revision (i.e 2.0). Maybe make it a seperate process (let's call it finalize translation) so that users don't do it by mistake.
I don't quite understand the difference between your subversion system and the present revision history accessible via the Revisions tab of each subtitle set: true, the revision history is purely chronological and does not differentiate between incomplete versions and revisions of complete versions, but is this so important?

And about finalized translation/transcription: for transcriptions, there is the check box for "these subtitles cover the entire video" when you save. If you don't check it, the subtitles are listed on the left with their number of lines - if you check it, they are listed as 100%.

Again, that's a bit simpler than what you propose. But Amara is meant in general to work like a wiki, without a final definitive version. However, Amara's enterprise services allow people to create teams, where the team owner and admins can fine-tune permissions and have workflows that have separate subtitling, revising and approval stages. If you want to see how that works, try joining the TED team, for instance.

2) Save the work of the user per line in a temporary revision(e.g something like -1.0) and offer this revision for continuation.  We had power problems in the last few days, and I was forced to rewrite lot of my work again.

I fully second saving per line - it was clearly so at one point, now sometimes if save doesn't work, you still find all your work when you reopen the widget, but sometimes not.

A tip: when saving does not work, click the save your subtitles link that appears in the dialog box: this allows you to copypaste your work into a text editor, and save it as a file you can reupload when your connection works again.

Problem which I hope will go away: presently, you cannot reupload your subs if someone else has started translating them. In this case, choose the .txt format: it's easier to copy-paste from sub by sub.
Also there must be a clear clarification on the matter of saving the work. For example when it is saved, when it can be considered final revision and when not etc.
I would like to propose the following
1) Use a subversion system. Like 1.1, 1.2, 1.3 etc. And when the checkbox is pressed as a final translation/transcription, then move to the next major revision (i.e 2.0). Maybe make it a seperate process (let's call it finalize translation) so that users don't do it by mistake.
2) Save the work of the user per line in a temporary revision(e.g something like -1.0) and offer this revision for continuation.  We had power problems in the last few days, and I was forced to rewrite lot of my work again.

Have an Amara player full screen mode that displays Amara subs.
See Claire Serbit's Subtitles desappear in full screen !! (and various other posts on this forum and on the old tenderapp forum about that).
@Jules : can we have back the possibility to download subs of video pages, please? It was removed a couple of weeks ago, and it's darned useful to overcome the effects of forced homogeneous sync. I.e. if someone starts subbing a video in another language than the original language, the only way to do the captions in the original language from the video is to do them elsewhere, then upload them. So the developers' platform was a convenient "elsewhere" solution.

Example: Achmed the Dead Terrorist Has a Son - Jeff Dunham - Controlled Chaos has incomplete Vietnamese subs made directly from the video. If I use the Start a new translation! link to start English subs, the only option offered is translating from Vietnamese. However, I could upload English subs made on the developers platform, if they were downloadable as they used to be.

So either re-enable the possibility to do any subs directly from the video, or please re-enable downloading from the developers' platform.

Update Oct 14

Actually, as the Vietnamese subs for Achmed the Dead Terrorist... were still unsync'd I was able to start English captions by uploading an irrelevant .txt file - see revision 0 - then replacing the resulting irrelevant subs by real ones.

Same with the English captions of Watch the Full 2012 Vice Presidential Debate, where Spanish subs had already been started, but not sync'd. Except that in this case, I uploaded a real plain transcript of the debate part, which had been published by NPR. However when I tried to add subs for the PBS intro, sync'ing along, the software refused to save. The reason apparently was that I had merged 2 subs of revision 0 that had already been translated into Chinese. At least, when I unmerged them, I was able to save.

And now we have a problem with  Melting Silver - Periodic Table of Videos, added to the Captions Requested team when the YT original already had full, sync'd Dutch subs that got transfered to Amara. The software does not even allow to start  English captioning from the video by uploading a file.

The sync'ing and the number of subs for each subtitle set must really be dissociated from other existing sets. And meanwhile, downloading subs from the developers platform must be reenabled: in the Melting Silver case, I could have uploaded the Dutch subs as English ones  there, then replaced them by real English subs, but keeping the Dutch subs number and time codes, then reuploaded the thus produced English subs to the main video.
Thanks for the clarification, Diomedes,

About having to be owner - or rather the uploader - of a YT video to create subtitles there: in some cases, I've downloaded the original video and reuploaded it privately so as not to compete with the original, then deleted it after I got the sync'd subs. In one instance, however - a video uploaded by Google itself - this got me an immediate whacking copyright violation notice, with downgrading of my status and capacity to upload longer videos, in spite of the "private" setting, before I could download the sync'd subs. So I deleted the video and regained my good status etc. But I'm a bit wary of doing that again now.

Re COGI's generation of complete automatic subs, see the history of revisions of  the English subs of 8. Case Study: The Rotavirus Vaccine: Revisions 0 and 4 are the COGI version.
@ Claude Almansi,

In my point 8 I meant to create the timecodes for the subtitles and not the subtitles themselves.
I have used the automatic transcription service from youtube, and also the CMUSphinx ( I have never used Cogi in order to give feedack about it. The ones that I used were pretty bad even in videos that there is no background noise (only speaker).
The procedure that I propose to be automated to substitute step 2, is as followed
1) Create the transcription of the video in Amara(step 1)
2) Export the transcription in .txt file without any timecodes(the default format for txt files in Amara)
3) Import the file into the youtube and let the youtube to create the timecodes
4) Export the file from youtube in .srt format (youtube export is only in .sbv so a conversion tool is needed)
5) Import it in the Amara platform and perform the step 3.

From my experience these are the major problems
1) You must be the owner of the video in order to create any subtitles.
2) When subtitle timecodes are created from a script without breaklines(i.e. Line change with Enter) ,the split is done randomly without some obvious rules.
3) When done with breaklines, some subtitles were split even if there were small enough.
4) When there are subtitles that refer to a text in the screen and there is no sound in order to match, timecodes are a bit messy.

Otherwise the service performs really well. Just very small discrepancies in some subtitles that switch off before the actual speaker finishes the sentence.
You can find all the above testings at the following video, which I uploaded just for the testings.

The English refers to the synchronization done with the transcript without any enters.
The English - with enter split refers to the synchronization done after splitting the original transcript with enters.
The English (transcribed) refers to the auto transcription done by the youtube.
Hi, I'm Verone, Webmaster of, a website I created 2 years ago. We got around 6000 amara's videos and around 1 milion page views a month (100 000 visitors), we made the subtitles of each video. We got around 10 active translators today, (grand total : 100)

From my point of view, priorities should be :

 1) When there are more than 1 source URL for an Amara work page, decide which URL is primary
We got around 200 censored videos on and we can't add a new video url to edit the dead one. To create a new video page and then make a copy of the translation is taking so much time and there is so many inconvenients doing this. Everything was perfectly working 3 month ago

2) Copypaste a plain transcript.
Now members have to encode their own text in UTF-8, which is a nightmare for many people. Everything was perfectly working 5 month ago

3) Have different sync'ing for the various subtitle sets in various languages of the same video
I'm on Amara for 2 years now and we have a french guide with around 20 rules about "How to make a good subtitling with Amara". We can't have the same synch for different languages, so sometimes we have to remove the english subtitles to make our own synch in French. And this is not good, but we have no choice with the actual system. 

 This is all we need, thank you.
Thanks to all for the further points added. I second in particular Diomides Skalistis':
3) Are teams available in the non-payable services?
(more clarity about that for each team would facilitate volunteers' decision to participate or not)
and also
8) Use the youtube service to create automatic timing for the subtitles (instead of manually create times). This should be tricky and must be examined very thoroughly if to be implemented. I have some feedback if you want.
Can you post this feedback, please? I've occasionally used the YT autotiming to create draft subs, but then I edited them, because the arbitrary YT timecoding creates subs that are difficult to translate.
But in the same line, could all users avail themselves of Cogi voice recognition? Its forced implementation in the Coursera team seems to have caused mixed reactions - see
.. A speech recognition system has left a big mess that can confuse and mislead students, sometimes to the point of telling them the exact opposite of what’s being said. Your mission is to slay the dragon and return peace to the land of Coursera :-) ...
But it's a bit better than YouTube's, apparently. So if those Cogi revision 0's were taken for what they are, they could be handy to elaborate on, especially for long videos, provided it's the subtitlers who decide to use Cogi or not.
Good evening everyone!

I am working on my very first project and have the following comments:

1. I would love a pause button so that when there is music, I am not leaving the old text where it should not be.
2. In a few instances, I have had a really hard time hearing the individual speaking and have left the following (???) in the subtitles as a reminder to go back to that particular subtitle. I would love to bold that area so that I do not overlook it. Ideally, I would love for others to help out, if possible. Somebody might not want to translate the entire video but could help with a short clip.
3. The TEDx community has created a TEDxLicensee Google Group for us to ask questions or lessons learned from each other. Does the Amara community have something like this?  

Thank you so much! This has been a great experience so far!


It has been GREAT to read your comments and suggestions, please keep them coming, we will certainly be addressing these in the Town Hall, if not earlier!

Through email, Vivienne Tran sent me these suggestions:
I'm not sure where to submit my comments for the townhall meeting, so let me send them by this reply. Here are my thoughts:

1) After each volunteer finishes captioning a video (that he or she uploads on his own to Amara), that person should be able to download the entire video with captioning. Right now, only the subtitle can be downloaded, not the video file itself.

The reason I am asking for this feature is because some deaf or hard of hearing friends may not be inclined to go to Amara website to view the closed captioned video. If I can down load a closed captioned video (that I did the captioning for a deaf friend, etc), I can send such video to her directly.

2) There should be instructions on how to type captioning for other languages (other than English). Sometimes, I want to do captioning in Vietnamese as well, but find it impossible to know how to type Vietnamese on the QWERTY keyboard, or make the Vietnamese typing into the videos on Amara web sites.

3) There seem to be the choice of subtitling to ASL (American Sign Language), but there is no instruction how to do this. My understanding is, it is NOT possible to subtitle into ASL because such language is visual-pictoral and not word-based. It would help if there are more instructions or some explanation of how subtitling to ASL would work.

4) Not many Deaf, deaf, or hard of hearing people know about Amara's captioning work or web site. Volunteers like us can tell our deaf friends, but we are only limited number of people. To reach more deaf people, I suggest Amara to "advertise" its work and mission purposes directly to more Deaf communities like Deaf Institutions, Universities, Colleges, Deaf Schools, etc. and even to regular Universities, College, and Schools where deaf students may study mainstream with other students.

Small practical things, I think:

1. Space for text boxes when working: When translating, I cannot see much of what I'm doing. I can only read one line. I also can access only a couple of textboxes. With all that space available, I think it would make things a lot better to leave more room for text boxes (taller) and have more text boxes.

2. Spell check in the destination language: Spell check in english is great. It would really help to have spell check in the language you are translating to.

3. Bookmark, or resuming work: I spend a lot of time going back to where I was the day before. There should be a way to resume work that involved a little les "research" :)

4. Latinamerican Spanish v/s European Spanish: These two variations of Spanish are very different. I see an attempt to differentiate, with Spanish(Argentina), Spanish(Mexico), etc. This goes in the right direction, but it does not work, since these countries are too specific (and too many to translate for them all).
A good point in the middle would have only two Spanish langs: Spanish(Europe), Spanish(Latinamerica)... it would make life a lot better for us Latams and for the people up there in Spain :).
Obviously, it should be possible to translate from one to another.

Finally, just to state how great Amara projects are. It's pretty cool to be able to help in a collaborative way, and learn so much from others in the process.

Thanks for listening too!
Hi Jules,

Generally I think this subtitling stuff is awesome! But you're right, things can always be better :-)

Apart from some really great improvements as mentioned above, I would like to add a few of my own:
- when timing, you can only click the down arrow to start the next caption. When there is a lot of time in between, it is really awkward to move the end of the previous caption back. So now I have to press tab again to stop the video, place the end correctly and then press tab again to wait for the next caption. I was wondering if you could either add a button for stop caption or extend the timeline, so it wouldn't be so awkward. 
- The layout of the screens are taken up by a lot of stuff that is handy when you start, but when you get more experienced, you don't need the explanation anymore. Would it be possible to change the layout so that you can see more of the text that is coming up? (I am working from a laptop and the screen is too small to see everything.)
- Could you add a button for an eight note (♪) in there? Again I'm working from a laptop and I have no idea how I can do the ALT-# combination (but that may just be me).
- I gave up on translating video's because the system is just too inflexible. When you translate you need to transform the grammatical order of sentences. If you translate literally it really doesn't make any sense and you want different breaks in Dutch as opposed to the English version. There should be an option that allows you to translate the text properly and then time it again appropriately. When I subtitle in the original language, I try to take into account that it may have to be translated one day, so I try to make the breaks in a natural spot, but unfortunately that doesn't happen everywhere.

Those are my two cents,
Login or Signup to post a comment