Why You Should Only Use US Closed Captioning Companies
Updated: January 24, 2019
When choosing a US English closed captioning company, cheaper pricing can be enticing. There are many mechanisms to drive costs down, such as offshoring and crowdsourcing, but often times these are used at the expense of caption quality.
With captioning services, you get what you pay for.
Poor caption quality harms your credibility, confuses your viewers and multiplies your work. Simply put, who your transcriptionists are usually correlates with how your final captions come out.
The 99% Standard
Perhaps you’ve heard of the 99%, an industry standard set to ensure the highest quality captions in a given media file. 99% accuracy, though close to perfection, means there is still a 1% chance of error. In a 10-minute file of 1,500 words, this leniency allows for 15 errors total.
The only way to achieve 99% accuracy is with a human editor who understands the intricacies of the language being spoken. For US English captioning, that requires that your transcriptionist is US-based.
Automatic captioning alone typically falls within the 60-70% range of accuracy. This means 1 out of every 3 words has an error, making the captions nearly impossible to comprehend.
Now, is this fair to your audience?
You want your caption quality to fall at the 99% accuracy rate, preferably higher.
Furthermore, the FCC has set rules (specific to network and broadcast video) so that the impact of the performance and overall message is not lost for the viewer. When deciding to use automatic captioning, non-US-based transcriptionists or crowdsourcing, be aware that these standards often get overlooked because many editors are not trained on these standards.
Lastly, you should always be cautious about vendor marketing. While a lot of vendors will state 99% accuracy, it’s important to test that accuracy rate yourself and learn more about their captioning process.
Who Captions For 3Play?
Upholding the highest standards in captioning quality is central to how we operate at 3Play.
After measuring our 1.5 million files, we found our actual measured accuracy rate is 99.6%. We are able to achieve this high quality because we use a multistep process involving automatic captioning and human intelligence.
All our editors are US-based, and they go through a rigorous certification process that runs through the necessary standards to certify our captions capture the full intent of the media.
Why a Non-US Based Transcriptionist are Bad for Caption Quality
Offshore transcription is when a vendor employs laborers outside of the native-language country. In captioning, this means vendors outsource their media files to employees in other countries.
Yes, it’s cheaper, but when you employ an offshore transcriptionist, you run the risk of missing the golden 99% accuracy standard.
With non-native English speakers, there is a concern that offshore transcriptionists don’t have the same handle of the English language as native speakers do. They can lack general knowledge of cultural and linguistic nuances that are essential to understanding the intent of the media. For example, an Australian might interpret a movie differently than an American because of the differences in slang, accents, and current events that influence the intent of the media.
Offshore transcripts are also frequently riddled with spelling and grammar mistakes, hurting the quality of the captions and failing to meet the FCC standard of accuracy.
How to Select the Right Closed Captioning Vendor: 10 Crucial Questions to Ask
Even if you’re using a US-based captioning company, make sure they don’t outsource their editing or transcription to offshore transcriptionists. The best transcriptionist is one native to the language of the media so that they capture the true intent as the media unfolds.
Crowdsourcing Will Lead to Inconsistency
Then there’s crowdsourcing. Crowdsourced transcription companies distribute your audio or video files among a large number of people. A single audio or video file is either segmented into smaller parts or assigned as one file.
This can be cheaper and at times faster, but it can also be very risky.
Often, a media file will have inconsistencies in punctuation, spelling, speaker identification, information and other non-dialogue information.
What you end up with is a transcript where in one section you have “donut” and in another section, it’s spelled “dougnut,” (though often these mistakes are more drastic).
When a single media file is split into multiple parts, these inconsistencies present a sloppy final product. But this can also happen with a media file that hasn’t been split.
If you upload three different videos that are assigned to three different transcriptionists, the quality and standards will likely be inconsistent across the files.
So, while one file will consistently use “ax,” another file may use “axe” instead.
These issues arise because consistent quality standards are not in place for crowdsourced transcriptionist to refer to. Often, these workers have not had proper training or been certified on existing standards. Furthermore, these transcriptionists can’t communicate with one another to discuss the work, so transcriptionists must make their own judgement calls for the file they are working on.
The only way to catch these mistakes is through editorial oversight, which not all companies offer, especially when the price to caption is cheap.
As for security, there are risks with leakages because there’s less control over the media file. If you have highly sensitive media, like a broadcast show, consider this element when choosing a captioning vendor.
Mechanical Turk is a web-based crowdsourcing platform where you can upload content and have freelancers work on your content.
You are in charge of uploading and splitting your content into tasks called Human Intelligence Tasks (HITs). The tasks are then published into a marketplace where self-employed workers can accept and begin to work on your content. After they have completed the task, they are submitted for you to review and either approve or reject them.
Though the Mechanical Turk allows you to dictate your own price (plus a fee to the Mechanical Turk), these costs can add up, especially if you employ multiple workers.
Furthermore, you run into the same quality issues of using a crowdsource-based vendor: inconsistencies and security risks. With Mechanical Turk laborers though, there’s a heavier emphasis on inconsistencies as they may lack proper training of the quality standards set by the FCC.
Lastly, a good captioning vendor will have an editorial process, where final transcripts are reviewed before they are handed back to the customer, to mitigate errors and ensure the content meets the standards. Through Mechanical Turk, this responsibility is placed on you.
Cheaper isn’t always better
With cheaper pricing, you forgo quality. When considering vendors, think about who will be touching your work and what risks you are willing to take. For optimal quality, choose vendors with US employed transcriptionists, who are trained on standards and promise consistency.
Artificial Intelligence Is Good, but Is It Good Enough for Captions?
“Artificial intelligence (AI) makes it possible for machines to learn from experience, adjust to new inputs, and perform human-like tasks.” In other words, rather than programming machines with a single rule for output, they are taught to recognize patterns and then utilize…
Automatic Captions: Are They Helping or Hindering Students?
In the 2019 State of Captioning report, 37% of respondents indicated that they use automatic captions for educational videos. At the same time, 50% said that they’re aware automatic captions don’t fulfill the legal requirements. These findings seem to conflict, so I…
2015 Harvard Accessibility Lawsuit Will Move Forward in Federal Court
On February 5, 2015, the National Association of the Deaf (NAD) filed a class-action lawsuit against Massachusetts Institute of Technology (MIT) and Harvard University. Harvard was accused of violating both the Americans with Disabilities Act (ADA) and Section 504 of the Rehabilitation…