Automated screen recording and video replacement for localization

  • 5
  • Idea
  • Updated 5 years ago
  • Not Planned
It would be really useful if CS would allow the automation of a screen recording. This would make it possible to recapture the video recording for each localized version of a product.

It would be especially useful if the edited video retained a log of mouse movements and clicks. Replaying the log of the edited video would then be used to duplicate the existing video in each localized interface.

The re-recorded video would then replace the original video in a duplicated project and the localized video could be adjusted with much less effort.

So there are two requests:

*** Automate video captures based on a log of mouse movements
*** Make it possible to replace the video in a project while leaving all of the production elements in place for reuse
Photo of steve.hammill

steve.hammill

  • 32 Posts
  • 6 Reply Likes
  • hopeful

Posted 9 years ago

  • 5
Photo of Carlos Miguens

Carlos Miguens

  • 3 Posts
  • 2 Reply Likes
We would also find this very useful, I think that there is a pent up demand for this, currently people are using *very* costly re-work processes for localization. This would make it much easier and cheaper to localize: demos, screencasts and eLearning courses.
Photo of David M. Converse

David M. Converse

  • 440 Posts
  • 26 Reply Likes
Hi Steve, we do have an article about timed recording that you might find helpful for automating your recordings.

http://techsmith.custhelp.com/app/ans...

It sounds like your other request would actually involve Camtasia controlling the computer? Performing mouse clicks, menu actions, typing text, etc? Its an interesting idea certainly!
Photo of Carlos Miguens

Carlos Miguens

  • 3 Posts
  • 2 Reply Likes
Hi David, thanks for the response (to Steve not me). There is a package which our QA team uses to build scripts for testing user interfaces called Selenium (http://seleniumhq.org/). We investigated using this to provide repeatability for recording screencasts/demos in different languages. It would work but the playback does not show mouse movement or clicks and so can't be used for the purposes of re-recording a demo in a different language.

I think this is a significant problem for a large number of companies that will need to do this, but I suspect that they have been doing it by just throwing resource at it.

If Camtasia could do this it might be quite a big seller for you - I don't know of anything else that does this and I've been looking for a while.

Carlos
Photo of steve.hammill

steve.hammill

  • 32 Posts
  • 6 Reply Likes
Thanks for the link, David, but I do not understand how that helps.

My mission is record and edit one screen capture and then have the (edited) log of clicks and mouse moves drive subsequent recordings in each localized interface. All post-production work could then be moved into the localized interface screencap and adjusted.

Clearly, the call-out text and such requires manual intervention, but it would still save a considerable amount.

For instance, you localize to 25 languages. The decision is made to keep the native narration, but include sub-titles and localized call-outs. Such a project could be turned around in short-order compared to an entirely manual process.

Even localizing the narration for some of those programs becomes a far less daunting task.

In a global company, the inability to provide a global solution for such a project, takes the wind out of the sails of even the best ideas.
Photo of David M. Converse

David M. Converse

  • 440 Posts
  • 26 Reply Likes
Hi Steve I'm afraid that what you are looking for is not possible with Camtasia. Camtasia records what happens onscreen, it does not have any features to control or perform computer activity.

You would need some sort of macro program that actually records activity as events and controls the mouse and keyboard. This is a far different task than simply recording the screen.

I suspect something along these lines would do what you are seeking:

http://www.tethyssolutions.com/macro-...

http://startly.com/products/quickeys/...

http://freelabs.info/MacroRecorder.aspx
Photo of steve.hammill

steve.hammill

  • 32 Posts
  • 6 Reply Likes
I know it is not part of the feature set, but I'd sure like it to be.

I'll look into those solutions to see where they get me. Thanks.
Photo of Carlos Miguens

Carlos Miguens

  • 3 Posts
  • 2 Reply Likes
Hi Steve, we have the same problem, I'd be happy to share any research on this, perhaps together we can actually find a useful solution. My email address is carlos.miguens@alfresco.com.
Photo of Randy Schott

Randy Schott

  • 396 Posts
  • 47 Reply Likes
Camtasia does record information about the keyboard, mouse, windows on the screen, etc. It is used to drive our SmartFocus capability. It could potentially be used as input for manipulating the keyboard and mouse, and maybe positioning windows. However, David is correct in saying that such a feature would be a major undertaking. There are plenty of UI automation frameworks that are struggling to accomplish the task already.

Don't get me wrong, I absolutely love the idea. But, I think we would need a lot of community feedback supporting the idea to justify the technical resources needed. My only other concern would be possible patent infringement. I'm almost positive that several patents already exist around automating activity on the screen. If there's a potential legal issue, we probably aren't likely to touch it. There may also be issues with user permissions. On Vista and higher, I believe a program has to be run as Administrator to take control of the mouse and keyboard in certain situations. That's not always an option in corporate environments.

As a potential alternative, how would you feel about being able to automate this stuff on the editing side? For example, what if you could find a particular window in your screencast, and then provide a screenshot of the translated version and have CS automatically overlay the translated version on top of the original? Maybe that would end up being just as much work as re-recording the video, but given our current technical assets, it might be more doable than automating the recording.

Photo of David M. Converse

David M. Converse

  • 440 Posts
  • 26 Reply Likes
Hi Randy, thanks for the input! I always love it when the smart people come in to help answer questions!
Photo of ErikT

ErikT

  • 36 Posts
  • 2 Reply Likes
I like the idea of the OP as well and would certainly see use for it.

The potential alternative that Randy has, will not work, I'm afraid. Most applications have menu bars, popup menus etc, so simply overlaying an existing past in language A with a new part with language B would involve quite some work. As Randy already concludes himself: Maybe that would end up being just as much work as re-recording the video.

Since I want to capture screencasts of our own software, I see a possible workaround for me: if I had access to the mouse and keyboard information in the camrec file, I could create a special version of our software that can import that information to drive the mouse and keyboard, start a new recording in Camtasia , switch language and then execute the original mouse and keyboard actions.
But, as said, I'd need access to that info in the camrec file...

Point two of the OP is also very interesting, even without the combination of the first point.
Photo of Randy Schott

Randy Schott

  • 396 Posts
  • 47 Reply Likes
That's an interesting idea. You do have access to the data (just extract the camrec contents). But it's probably in a binary format, and we don't have any APIs or SDKs to help with reading that data.

I guess if you felt like reverse-engineering some things, nobody would stop you. Otherwise, the product team would have to discuss the possibility of putting a little time into the above-mentioned SDK, or maybe just a tool that can take in a camrec and spit out a few XML files?

Sounds like a fun spare-time project :-).
Photo of steve.hammill

steve.hammill

  • 32 Posts
  • 6 Reply Likes
Sorry it took so long for me to get back to this thread, Randy.

If I could slide a new interface background over the US_EN interface so that all the pixels and such matched and all the effect remained in place, that would be an excellent start.

Unfortunately, I bet if you delivered that to me today, I'd be crying about it not doing this or doing that tomorrow.

Automating mouse movement and button clicks is probably the key.

So if I could extract that info out of the camrec, modify timings to match the localized narration, import the modified information into an automation tool, perfectly align the new browser/app window with the original recording, open CS, start the recording, and launch the automation...

YUP! A big job. ...and what about the effects? OYE!

One of the things preventing wider adoption of videos in big corporations is that everything must be localized into X or XX number of languages. Until that requirement is met, product demos/trainings will remain one-offs created outside the customary scope of a project. :(
Photo of ErikT

ErikT

  • 36 Posts
  • 2 Reply Likes
Thanks Randy,

I hope one of you soon has some spare time ;-)

I did know that I have access to that data, but as you said, it's binary data so I do not really have access to the information, which is what I would need.

I guess if you felt like reverse-engineering some things, nobody would stop you.

Ehm, the TSLA does, in 2.1? Or does that only apply to the software and not to the saved files?

As Steve mentions, localization is very important in many organizations, including ours. So any feature that will help here is welcome.

I agree with Steve that being able to replace an image or video clip by another one without affecting all other stuff in the same time frame would be an excellent start

Having a way to extract mouse and keyboard information out of the camrec file to use this to drive the application that is recorded would be a welcome addition.

And while I type this I realize: the opposite would be very welcome too: extract the mouse data, edit it and then put it back in the camrec file. How often didn't I re-record a part of my screencast because my mouse movement was shaky or I made a strange curve etc, that didn't look well. With the mouse movement information editable available, one could correct those moves without having to record that part all over.
Photo of Randy Schott

Randy Schott

  • 396 Posts
  • 47 Reply Likes
Yeah, I should start thinking a little before I speak :-). You're probably right about the TSLA. I guess I was thinking of the fact that there is an open-source codec that (mostly) decodes TSCC, and that was definitely reverse-engineered. Maybe it depends on whether or not you plan to release the result commercially.

Either way, that's more work than you should have to go through as a customer! It's safe to conclude that the proper solution would be offering some way to "replay" a recording.

As for editing the cursor/keyboard data and putting it back in, I think that might be shaky territory. I remember hearing that somebody has a patent on editing mouse position in a screencast?

As an aside, I really appreciate all the discussion on this topic. It really helps us understand what users need, especially those with use cases we don't always consider.
Photo of Yves Barbion

Yves Barbion

  • 1 Post
  • 0 Reply Likes
Hi group

I just found this topic. I was wondering whether this is now possible in Camtasia?

Thanks

Yves
Photo of Henrik Jakobsson

Henrik Jakobsson

  • 1 Post
  • 0 Reply Likes
Hi all,

I believe that there is a company called www.lingoalta.com that uses editors to re-capture and localize software training videos including mouse movements/clicks etc. Don't know if they have found a workaround for this feature but I'm sure they would be interested.

This conversation is no longer open for comments or replies.