Not too long ago, the world of videos being professionally translated was sufficient… but it wasn’t very fast, cost-effective, or automated.
Before the creation of our automated transcription, translation, subtitling and voice-over tool, the process to get videos translated was extremely difficult.
It would take three to four programs, four to five file transfers, and at least three specialists. This makes the process messy, take a long time, and cost more than it should. We developed a way to cut most of that mess out. But first, we must understand how to translate videos the regular way to understand that we didn’t cut any corners with our new breakthrough service. We get the same job done with a smarter, faster, and more cost-effective solution.
The Current Process to Translate Videos
There are a few steps involved in order to translate videos professionally:
Transcription, translation, subtitling, and voice-overs (if that’s a service you require).
Transcription would require a manual process of hiring an expert to listen and transcribe the video content. That would be time-consuming, expensive, and require reviewing for quality issues.
The next step in the transcription process, before someone translates it, is to properly time stamp the transcribed dialog. The spoken sentence in a video is split up in a logical manner. Time stamps are created to match the transcription to the voice of the on screen characters and to match the on screen events. Time stamps also need to conform to the client standards. This is done to avoid awkward endings and wraps that don’t flow naturally. A typical subtitle length is 35 characters per line, two lines max.
It’s crucial to do this step prior to translation because many languages expand as word sizes vary. The length of words in another language, other than English, during the transcription process, is a legitimate concern and has to be formatted properly. This process requires an expert.
This process takes time and effort which = $$
That being said, this is a best practice but it’s not always followed for a number of reasons but mostly to save money.
You are probably familiar with the main difference between subtitles and closed captions:
Subtitles are just the uttered words in dialog
Closed captions are the words but with context more aimed at hearing impaired folks (a requirement for ADA in broadcasting). It adds things that happen that surround the dialog such as [GUN FIRE] [SING] [SHOUTING].
A transcription by default doesn’t account for the CC attributes mentioned above. That is another job.
If you don’t time stamp and clean up the English transcription prior to translation then you have to complete the task after the translation phase and before engineering and also for each language. So here is the negative to that… if you have multiple languages, that will add up to a lot of time.
Time is $$$
Another solution to this costly process arose with the creation of software that can recognize speech and convert it into text. You probably are aware of this since every smartphone has this capability. But you’ve probably had troubles with your smartphone’s speech-to-text ability, and the same trouble occurs when weak software is used to transcribe video content. New technology has been developed by IBM that has incredible speech-to-text capabilities, but we’ll get into that later on.
After a proper and costly transcription and/or cc process, there is a translation of the transcription and then subtitling. Once the video’s dialogue is transcribed into text, there are several ways it can be translated:
- You can hire a translator specialized in the industry of the content for each language you want.
- You can hire a translation agency with a network of linguists and software to streamline the process.
- Or you can simply use google translate or other machine translation engines though the quality would not be good.
However you go about translating the text, once it is done, a video engineer will input the translated text into the video, with proper timestamps, through subtitling and closed captioning. Currently, it cost about $75 to transcribe one hour of content. As you can see, the video translation process keeps adding up with time and money spent.
A sound engineer has to synchronize the translated text with the video through time stamps that were created during the transcription phase, then input them. Then there has to be some kind of quality control built into this process to ensure there are no mistakes. Usually, two people play a role in the whole video translation process–one doing the task and a second to review it–and that’s the minimum effort to ensure quality.
If you want a voice-over done for a video, that is a whole other story and it’s an expensive one.
You probably guessed that in order to get a voice-over on a video, you need to hire a voice actor in the language that you want. That’s the expensive part.
On voice, there are two main processes. One is narration in the video–a voice is heard but the person is not on screen. Or dubbing–the voice is heard and the person uttering words is also on screen.
Narration is much more forgiving than dubbing. With dubbing, there may be a need to edit the translation to match the time duration of the person speaking. The foreign language dialog must match the source language to avoid a discordance on lip movement and words still being spoken and heard.
Current Video Translation Process
The bottom line is that this process, again, takes a lot of experts doing manual work to complete a professionally translated video.
Remember when I said that this is the usual process before we created our automated solution?
A Revolutionary Video Translation Service
Let me start by introducing the platform that allows for fast, accurate, and automated video translation.
UNIFY is a translation management system that streamlines the translation process by utilizing the most advanced technology and bringing all project collaborators together into one intuitive platform.
Utilizing UNIFY cuts time and costs down, as well as creating a streamlined translation process by offering things like:
- Never get charged for the same translation resulting in large long-term savings. So when repeat words, sentence, or phrases are being used you are not paying for those. “Boom, Boom, Boom” is only charging you for one word instead of three.
- Tracks every type of change that’s been made and provides a measurable overall quality score, which means users never have to guess what changes were made and why.
UNIFYed Project Collaboration
- Multilingual glossaries and in-context review allow translators, reviewers and project managers to communicate and collaborate together, resulting in unparalleled abilities that ensure the absolute best quality, consistency, formatting, and contextual accuracy.
- Superior Machine Assisted Rapid Translation, which is the fastest freaking translation tool ever created and cuts human translation time in half.
- Documents, websites, software, mobile applications, and more are translated and localized rapidly and cost-effectively with greater efficiency.
- UNIFY supports 50+ file types and features pre-built integrations and APIs, allowing for effortless connections to any other platform, content management system, or technology stack.
With UNIFY, you capture a lot of the functionality needed to complete a quality and professional video translation from beginning to end.
And now added to the UNIFY platform is an automated media translation tool.
Drum roll please for the thing that will change the global media world as we know it and make a dent in the history of video content…
Introducing UNIFY Media Workbench!!
Before you think, “wow big deal,” let me explain why it is a huge deal and why it’s super important and relevant to you if you need videos translated.
UNIFY Media Workbench uses technology developed by IBM to convert speech into text for automated transcription. But remember when I said that weak software had trouble producing accurate speech-to-text? IBM’s software works remarkably with a 97 percent accuracy level. That means only a quick clean up by a linguist is necessary to have a video transcribed. Then a professional human translator works within UNIFY, which has an advanced machine translation tool that allows for rapid human translation. Once the transcription is translated into the language of your choice, the tool has a built-in subtitling feature that is used by a linguist to input subtitles instantly and within the right time stamps that correspond with the video.
UNIFY also allows for on-the-fly editing and revising to meet the duration of a video.
So there you go, bada-bing bada-boom. A video transcribed, translated, and subtitled quickly and accurately.
Now as far as voice-overs go, this is where technology has really excelled.
IBM has developed a voice-over tool that sounds a little better than your average Siri or Google. And by a little better, I mean it’s practically human. This tool currently has 11 languages that convert text into amazing sounding speech.
Don’t believe me?
That’s okay, the other languages spoken in this video are a good example.
This powerful voice-over tool is also built into UNIFY Media Workbench and can be used to provide instant and great sounding voice-overs for video. Of course, if you don’t want an automated voice-over, you can take advantage of inWhatLanguage’s vast network of professional voice artists for your language of choice.
UNIFY also will have the ability to allow direct recording by a human into the system with the editing functionality all available.
Get all your media translated and localized quickly and simply with the most advanced tools and translation experts.