To our initial question, about how Daniel would explain what QScan is, he answered: “QScan is a web-based AQC for on-premise, in-cloud, or Hybrid Quality Control of media content.” Or, to put it in our own, more amateurish vernacular: QScan, together with similar offerings, is what every professional media production company needs, once content is ready for publication – and must pass the requirements of broadcasters, streaming providers and the like. LWK’s homepage can be consulted directly for information about QScan 2023. See here: is.gd/QScan2023.

For a comparison with another competitor on the Quality Control market, coming from Switzerland, read our interview with Dan Tatut Vice President for Business Strategy and Development at Marquise Technologies. The article is titled “Ice, Mist & Tornado: Three elements of Quality Control”, can be found on page 32 of our issue DP 06:22.

DP: Could you please explain what an eyeball report is?
Daniel Roques: Eyeball reports are created through human supervision. An eyeball report usually contains problems detected by a QC operator while viewing the file. This could be as simple as “wrong colour correction”, “microphone on screen”, or more complex ones that an AQC could never detect. AQC usually detects problems that can be identified either by objective technical measurements (such as video levels, audio levels, and loudness levels) or complex algorithms that would identify bad quality by analysing blocks and macroblocks in a frame and consecutive frames (freeze, black, drops, etc.).

DP: Another new feature is the ability to combine eyeball events into a customizable AQC report. How does this work?
Daniel Roques: Most post-production facilities have to provide both, an eyeball report and an AQC report. In some cases, they need to provide one unique report combining all of this. QScan can generate one unique report by importing external Eyeball QC reports or allowing the manual insertion of events in the timeline view. This Eyeball QC event will automatically be added to the AQC report, thus saving time for report generation.

DP: New templates for Disney Plus have been introduced, right?
Daniel Roques: Yes. Disney Plus templates have been a very popular demand from our customers since we relaunched QScan in LWKS. It assures post houses delivering content to Disney Plus that the content delivered will comply with the basic QC specifications – defined by Disney Plus.
Some parameters had to be modified, and even some new parameters needed to be added and new macros had to be created. Most of the basic requirements were covered (IMF, J2K, HDR, UHD), but we wanted to cover them all.

DP: LWK’s website says: “Other important additions include enhanced PDF report customization and generation.” Where lies the advantage in generating PDF reports in parallel – as opposed
to sequentially?
Daniel Roques: In previous QScan versions, if you had four files being analysed, once a file was finished, the PDF was generated. Now, if the second file was through being analysed, while the first file was still in the process of generating a PDF, the second file would withhold starting the PDF generation until the first file had finished. Thus, generating a bottle neck in the PDF generation, as well as some performance problems that would lead to PDF generation failing. Doing this in parallel processes minimizes the risk, speeds up the creation of PDF reports.

DP: In tandem with QScan 2023, LWKS releases React Share & Review. How does React Share & Review differ from QScan?
Daniel Roques: React is a different platform to QScan. QScan is purely AQC, while React allows collaborative work in the cloud, with sharing, uploading, downloading, commenting, transcription (speech to text), publishing and many more tools in a secure environment. QScan integrates with React in such a way that any file that is managed within React can also be sent to AQC. By contrast, QScan is the tool used for AQC, returning a PDF report to React.

DP: What are some of the most complicated Quality Control Challenges you experienced up to this point?
Daniel Roques: The most complicated Quality Control challenge is trying to satisfy all users worldwide, with many different requirements, specifications, and parameters they need to detect. Heavy high-resolution formats are usually the most challenging ones; because of the processing power, and higher resolution (and dynamic range), those come with new problems that must be detected.

DP: How do the demands of streaming services, broadcasters, YouTubers, etc. differ from one another?
Daniel Roques: There are big differences. Most of the streaming services have similar requirements concerning file format packaging, resolution, HDR requirements or compliance to standards. At the moment, UHD and HDR are rare for broadcasters; they are usually more tied to lower resolutions and SDR (Software Defined Radio) content, although they are also very strict regarding compliance to standards. YouTubers would be the lower tier in the segmentation. Requirements are flexible in terms of quality and standards, although now more resolutions and dynamic ranges are supported.

DP: Thinking the other way around: What feature of QScan 2023 would you cut out for the next version?
Daniel Roques: Nothing! We’re too busy listening to our customers on what to add next.

DP: What will QScan be capable of in five years?
Daniel Roques: QScan will support more AI features, as well as additional formats, resolutions and dynamic ranges — some of which are not even invented yet! Video technology evolves quickly and we at LWKS are flexible enough to adapt our roadmap.
To be more concrete: There will be mobile versions of QScan, to run on tablets, phones, and other mobile devices. Enabling the confirmation of AQC in the cloud would be another step for the future, since much of media content may be located in the cloud by then.›

QScan 2023: New Features

Upon our request, about what the most outstanding features of QScan 2023 are, Daniel provided us with a concise list of bullet-points, highlighting what Lightworks considers noteworthy.

• Template creation from reference files
• Comparative analysis
• Video/audio timeline view: Here, the user can see the behavior of all measured values throughout the whole file
• Eyeball event insertion
• Customisable PDF report
• Macros creation for complex detections (combination of different parameters for advanced detections)
• Single pass analysis (analyze once, change your thresholds many times)
• Chunk-based analysis for maximum
• CPU efficiency
• Comprehensive HDR reports for MaxFALL/MaxCLL (Maximum Frame Average Light Leve/Maximum Content Light Level) measurements (for more information about MaxFALL/MAXCLL, see here: is.gd/DolbyMaxFALLMaxCLL)
• CSV (Comma Seperated Value) download of all measured values per frame
• API (Application Programming Interface) and integrations with third-party tools

The post Trust is good, control is… QScan? first appeared on DIGITAL PRODUCTION and was written by Patrick Poti.