What is Captions Quality and how to ensure it?

What is Captions Quality and how to ensure it?

Closed Captions are a must-have for deaf and hard to hear people comprehend and enjoy the audio-visual content.

According to DCMP (Described and Captioned Media Program), there are more than 30 million Americans with some type of hearing loss. If extended to the worldwide population, this number will easily grow to a few hundred million. This is a large population that needs effective and high-quality Closed Captions to comprehend the media content.

And it is often stated that Captions should be of “high quality” to be effective for this population.


But how does one define this “high quality” for Captions?

According to DCMP, the following are key elements of Captioning Quality:

  1. Errorless captions for each production.
  2. Uniformity in style and presentation of all captioning features.
  3. A complete textual representation of the audio, including speaker identification and non-speech information.
  4. Captions are displayed with enough time to be read completely, are in synchronization with the audio, and are not obscured by (nor do they obscure) the visual content.
  5. Equal access requires that the meaning and intention of the material are completely preserved.

Every caption service provider needs to ensure that they create “quality captions” meeting the above guidelines. And every content provider, whether a Broadcaster or a Streaming service provider, needs to ensure that they deliver high quality captions to their viewers. Failing these captioning quality standards can have a detrimental effect on their brand and people will be less enthused in signing up or continuing with their captioning/subtitling service.

Checking for all these parameters and correcting them can be resource-intensive, tedious, and cost-prohibitive. Nowadays, the same content is delivered through a variety of mediums, which may use different audio-visual content versions due to specific editing, frame rate, and other technical requirements. It is essential that the captions are also properly edited for these different content versions. Due to the sheer content volume and the cost involved in Caption/Subtitle QC & correction of each content version, captioning service providers may choose to not perform full QC on all the caption files, leading to low quality and erroneous captions. It is therefore important to bring automation in the QC process so that QC process itself becomes feasible and more manageable for everyone in the content production and delivery chain.

CapMate™, a native cloud captions verification & correction platform from Venera Technologies, allows users to automatically detect a variety of issues that affect quality. Some of these issues and their impact on quality include:

  1. Captions-Audio sync: Readability issue
  2. Detection of missing captions: Clarity issue
  3. Captions overlaid on burnt-in text: Readability issue
  4. Captions duration: Readability issue
  5. Characters per line: Readability issue
  6. Characters per second: Readability issue
  7. Words per minute: Readability issue
  8. The gap between captions: Readability issue
  9. Number of caption lines: Readability issue
  10. Spell Check: Accuracy
  11. Detection of Profane/foul words: Compliance issue
  12. Captions format compliance issues

CapMate™ not only allows Automated QC of Closed Caption and Subtitle files but it also provides for automated correction of a wide range of issues with an option for manual review. CapMate™ comes equipped with a browser-based, rich viewer tool, that allows users to review the results in detail along with an audio-video preview. This viewer application also allows users to edit the captions. Once all the edits are done (automatically or manually), corrected caption/subtitle files can be exported in order to be used in the workflow.

Usage of CapMate™ can save numerous hours which can lead to fast delivery times as well as reduced QC costs. Content providers who depend on closed captioning service providers can send detailed QC report to their vendors, reducing review iterations as well as turnaround times.

With its usage-based monthly or annual subscription plans, as well as a unique Ad-hoc pricing plan, CapMate™ can fit every budget for organizations of any size, proving to be an indispensable tool in improving the quality of captions.

Read more about CapMate™ at www.veneratech.com/capmate. You can also request a free trial on the same page.

Ref-Q™: Reference based Automated QC

Ref-Q™: Reference based Automated QC

A novel approach for automating the detection of differences between versions of media content.    

Did you know that a major studio may end up creating 300-400 versions of a premium title for distribution? Well, that certainly happens, and often!

This happens because of the following scenarios:

  1. Different content edits needed for different locations and customers
  2. Different texted versions of the content for different languages
  3. Multiple delivery formats
  4. Both HDR/SDR versions to be delivered for various distribution channels

Validating all the content versions for editorial and quality accuracy can be daunting and will require numerous, expensive person-days in order to complete. Even after investing all that time, the process still remains error-prone as it is very tedious and subject to human error. This may lead to operators missing important differences, resulting in heavy costs for the content providers in terms of content reproduction as well as impact on their reputation.

Ref-Q™, is a new reference based automated video comparison technology from Venera Technologies, available as a feature within Quasar® native-cloud QC service. It can significantly reduce the overall cost of the validation process by producing predictable and accurate results much faster and at much lower cost, while allowing your teams to focus on the core business activities. Your team needs to only review the differences reported by Ref-Q™ and decide whether to accept or reject those differences. Ref-Q™ works 24×7 for you in an automated manner, allowing you and your team to invest time on activities that requires human intervention.

Use cases

Following are the common use cases where detailed frame-by-frame comparison is needed:

1. Transcoding issues

On occasions, frame dropouts or significant degradation may be introduced due to the transcoding process. This may lead to introduction of black frames, colored frames or significant drop out artifacts being introduced in the frames. This may happen only for a few frames and so can be potentially missed during the manual QC process.

Transcoding issues - Ref-Q

Ref-Q™ can automatically compare the source and transcoded versions of the content and report the difference segments.

2. HDR/SDR versions

HDR capability is commonly available in television sets today and therefore it is common for content providers to deliver both the HDR and SDR versions of the content. HDR and SDR versions for a particular delivery must be exactly the same structurally. Even the minor differences in the structure of the HDR and SDR versions can lead to rejection of the file by the customer.  In other words, the same texted version must be delivered to a particular country/customer in both the HDR and SDR versions. This is very difficult to validate manually.



Ref-Q™ can perform automated comparison of HDR and SDR version of the content, and report if any structural differences between the two versions are found.

3. Texted/Textless version comparison

In addition to the comparison for HDR/SDR versions of the same texted delivery, there is also a requirement to spot all texted segments within a content. Doing this manually is tedious, error-prone and costly.

Texted/Textless version comparison - Ref-Q


Ref-Q™ can allow you to compare the textless and texted version of the content frame by frame, and report all the locations where the text has been overlaid on the video.

As an output of the QC process, you get a XML/PDF report with start/end timecodes of all difference segments. You may review these segments and make the appropriate decisions.

Ref-Q™ is offered as part of our Native Cloud QC service called Quasar®. All the advanced scaling and security capabilities of Quasar® are also available for Ref-Q™ analysis jobs. You can process hundreds of Ref-Q™ jobs at the same time, allowing immense scalability. To put things in perspective, a set of one-hour files that your manual resources could take a few hours to compare can be completed by Ref-Q™ in fraction of that time. This allows for significant business benefits with fast turn-around times and accurate automated analysis requiring only minimum manual review.

With its monthly or annual subscription based pricing, as well as its industry-first ad-hoc usage-based pricing, Quasar® and its Ref-Q™ feature is available to media companies of any size as well as media professionals.

We also offer a free trial for Ref-Q™, allowing you to test your files on our SaaS platform without any prior commitments.

Get in touch with us today at sales(at)veneratech.com to learn more and request a free trial.