11 Jun Transcription Trends: Software – The John Henry Paradox
Today I read an article about the newest update to Phillips’ Speech Processing Software SpeechExec Enterprise. The impressive dictation and transcription management software just got a full overhaul and is now bristling with new features. The following excerpt from the MarketWatch report sums it up pretty well.
“The new software version supports faster turnaround times through a variety of updated features, including an enhanced statistics module, which generates data on dictation and transcription workflow. This feature allows for quicker decision-making and facilitates continuous workflow optimization for ever-changing user patterns and preferences. The updated software also enhances user management with Active Directory synchronization allowing administrators to identify user groups and manage user rights with the click of a button. Additionally, administrators can assign group management rights to heads of department, making it easier to evenly distribute workloads among users. – Market Watch Press Release”
John Henry Vs. The Steam Engine
Are we in for a John Henry experience here at Word Wizards? Personally, I think not… Although SpeechExec and other dictation and automatic transcription software solutions continue to improve at a dramatic pace, the fact is that we offer a level of service and quality that wont be matched by these software competitors for many years to come. Sure, it looks cool when Morgan Freeman tells Siri to take the night off on TV, but try it in real life and she will simply tell you, “I don’t understand.”
Dictation software works great under ideal circumstances, but throw some background noise in the mix, multiple speakers, or even worse the dreaded accent, and your software solution breaks down. Only a human can understand the context of another human speaker, period. We also think that only a native English speaker will perfectly understand and transcribe that language, which is why we have never shipped our transcription work over seas.
Here are some things that people can do and machines just can’t (yet):
- Add accurate time codes to transcripts.
- Transcribe from poor audio sources, i.e. background noise, hum, ext…
- Understand context of information and proofread for continuity.
- Distinguish between important information and “off-topic” chatter.
- Understand strange accents.
- Feel accountable for the quality of their work.
- Learn and improve over time!
One day the technology to automatically convert speech to text might rival our highly trained speed typists, but until then were still ahead of the technology by leaps and bounds.
An Ace in the Hole…
Whatever the outcome of the battle between man (and / or woman) and machine, were ready to stay on top! Our new service, CloudScript is applicable to any transcript, either made by human hands or by automatic software. If you work with transcripts and media files on a regular basis, CloudScript offers a real advantage to make you life much easier.