Forensic Transcription

IT’S HIDDEN IN THE DETAILS

Forensic transcription doesn’t refer to scientific work in a crime lab, but instead describes our analytical, detail-oriented approach to each client’s project. It is our approach, and our clients’ understanding, that we will pay attention to the minutiae of each and every project we receive: audio, video, handwritten, or any other type of content that we’re given by every client. 

At the Audio Transcription Center, we’ve spent the better part of our 56 years in business investigating changing technologies to always improve upon how best to transcribe our clients’ recordings and materials. But our methods have always stuck true to our beginnings: harnessing the power of human ingenuity and critical thinking first and foremost. 

It is through this forensic approach by our all-human team that we’re able to understand that each project has its own set of exciting challenges, and requires a level of attention to detail that artificial intelligence and other services just don’t provide. 

At the Audio Transcription Center, nothing about our intelligence is artificial.

Accents, Artificial Intelligence and Humans.

Accents. There are an estimated 30 accents that span the landscape of the United States. Tell me, if we as humans have a hard enough time parsing out the dropped “Rs” in words from a Bostonian (please note we’re a bunch of Bostonians here at ATC), how is Artificial Intelligence (AI) ready and able to do so? It isn’t!

There’s a reason we continue to be used as a human test-case against AI.

Is the adaptability of artificial intelligence’s deep learning modules able to discern all of these accents, colloquialisms, and dialects the same as the adaptability of a human team of transcriptionists? We think not. Who better to transcribe that Bostonian than fellow Bostonians? Who better to comprehend the words and colloquialisms from recordings of oral histories from folks in New Orleans (for instance) than people from New Orleans? We’ve been custom-matching client content to every human transcriptionist for 55 years, and we’ll keep doing so. We guarantee it!

Lastly, I know when I talk to the AI of my phone asking it one question or another, inevitably, it gets something wrong every time. And mind you, I’m somehow one of those Bostonians who no one ever believes is actually a Bostonian. Yet, it still has a hard time understanding me. Go figure.

At the Audio Transcription Center, nothing about our intelligence is artificial!

SUPERIOR TRANSCRIPTS REQUIRE MORE THAN ARTIFICIAL INTELLIGENCE!

Time and again we’ve tested, measured, and evaluated voice recognition software to determine their highest possible accuracy level. We have determined that with broadcast-quality audio of two well-spoken people, AI can presently reach 96% accuracy, at best. Translated that means approximately 10 errors per page versus our greater than 99% accuracy rate resulting in an average 2.5 errors per page.

So what does AI continue to have difficulty with?

  • less than broadcast-quality audio
  • multiple voices (interviews with 2 or more people, multi-person focus groups, etc.)
  • accents
  • ambient noise
  • special vocabulary
  • grammar
  • punctuation
  • spelling

Our production team custom-matches our clients’ subject matter to each transcriptionist’s particular strengths, knowledge, and interests. Our goal is to always make sure that what is said, is what is heard, is what is transcribed. Capturing each recording in a transcript that follows the client’s directions in intricate finite detail. 

We have the most selective hiring standards in our industry. Aside from a minimum typing speed of 80 wpm, we only select people who are well-educated, culturally diverse, intellectually curious, and possess excellent grammar skills.

We are able to quickly mobilize a dedicated team for your time-sensitive projects as well as highly confidential work, and we also have a nationwide team available for projects of any size and subject matter. 

AT THE AUDIO TRANSCRIPTION CENTER
NOTHING ABOUT OUR INTELLIGENCE IS ARTIFICIAL! 

Client Spotlight – July 2020

Since our founding in 1966, oral history interviews and archival recordings have been a main focus of our transcription service. Each time we partner with an organization it is always amazing to learn the ins and outs of each project. It is with great pleasure that we share these two fascinating projects that we proudly transcribed.

Lynda Kachurek, Head of Rare Books & Special Collections at the Boatwright Memorial Library at the University of Richmond shared these comments with us about two of her projects.

This two-part oral history with Dr. Wyatt Tee Walker and his wife, Theresa Ann Edwards Walker, was conducted by the University of Richmond in 2016 and is believed to be the last interview Dr. Walker recorded before his passing in 2018. The excellent transcription completed by the Audio Transcription Center assists with accessibility online, as well as allowing greater subject analysis and research access for the material.

The Birmingham Tapes are a set of 10 mass meetings that span the duration of Project C, the SCLC’s (Southern Christian Leadership Conference) work during the 1963 Birmingham Campaign. They are believed to be the only publicly available recordings of these meetings, presenting a significant and unique perspective on the detailed planning and management of the Birmingham Campaign. The transcriptions done by the Audio Transcription Center not only assist with accessibility but they allow for a deeper research potential, greater subject analysis, and wider availability to these materials.

Fast Isn’t Meaningful if a Transcript isn’t Accurate

There’s a reason the top academic institutions, government agencies, financial organizations, and market research firms continue to call on the Audio Transcription Center for their most challenging projects.

Our clients want their transcripts fast, but understand the importance of balancing speed with the delivery of a 99% accurate transcript.

Fast isn’t meaningful if a transcript isn’t accurate.

Accuracy is a necessity, and they know they’ll always get the most out of the diverse and knowledgeable transcriptionists who are custom-matched to each project. These awesome people bring their varied and unique experiences to each recording they’re transcribing. This allows for better understanding of the content, a more accurate transcript, and less time reviewing and fixing errors, saving time and dollars.

“I am so pleased with the transcriptions your people have done. This is a very large project, and I’ve only this week gotten to review the transcriptions. They are amazingly accurate, seemingly even looking up names to get them right.”

-An ATC client

At the Audio Transcription Center, our transcripts may be 99% accurate, but they’ll also always be unconditionally 100% guaranteed or no charge – no ifs, ands or buts.

Michael Sesling, Vice President

ATC Client Spotlight

May 31, 2020
ATC’s Client Spotlight is on the Hatfield Historical Society


As we wrap up the month of May we wanted to share a fascinating oral history project that our client, the Hatfield Historical Society, has just published on their Hatfield Vietnam Stories website. These oral histories capture the untold stories of Hatfield veterans from the Vietnam War. These captivating oral histories are full of memories with raw emotions, and they intertwine their amazing recollections from their lives and experiences during the war.

It was an honor to be a part of this project, and our team was so moved by the veterans’ responses.

These interviews help us better understand not only the experiences of these veterans from the Vietnam War, but the oral histories also give a greater sense of our nation’s history from that era.

What did they just say?

One of the great things about studying abroad in St. Petersburg is that the world-famous Hermitage is free for anyone with a student ID.

Kris’ student ID in Russia

All of the guides are fully bilingual in English and Russian, and know every detail about every floor, window, article of clothing, and gift from the Grand Duchy of Wherever in 17XX.

A friend of mine went on an English-language tour and told me this story. The guide was giving the full rundown on a room, when they got to one piece of furniture, which they said “is made of, um, some sh*t.” Without missing a beat, as though nothing had happened, they then moved the tour along.

The group had no idea how to react. What’s going on? The guide’s English is immaculate, they must know what they just said. Was it a joke?

My friend had a hunch, so he looked in his pocket dictionary for самшит – S-A-M-Sh-I-T – which is the Russian word for “boxwood.”

I think about this story a lot at ATC. If I hear a word or name I’m pretty sure I know how to spell, I always Google it anyway just to be safe. If a speaker says something that doesn’t quite add up, I think about the context. Are these database engineers really saying “sequel” a hundred times, or is there some industry jargon/acronym they’re using? (In this case, SQL.)

As the old Russian proverb goes, trust but verify, otherwise you may end up stuck in самшит.

-Kris

The Hidden Truths of Voice Recognition Software

  • Q: Why the Audio Transcription Center cannot use Voice Recognition Software?
    • A: Because Voice Recognition Software is not yet capable of producing to our strict standards.
  • Q: What strict standards?
    • A: Let us count the ways:
      • VRS has difficulty in recognizing, simultaneously or not, two or more voices. Of course, two or more voices are intrinsic to oral histories.
      • VRS has difficulty with accents.
      • VRS has difficulty in dealing with less than broadcast quality sound.
      • VRS has difficulty with overlapping dialogue, idioms, collaquialisms, and especially ambient sound.
      • VRS – Formatting? Fuggedaboutit!
      • VRS developer IMB reached a 94.5% accuracy milestone which they are very proud of in its evaluation by “using the SWITCHBOARD corpus, a collection of telephone conversations that’s been used for decades.” “SWITCHBOARD is not the industry standard for measuring human purity, however, which makes breakthroughs harder to achieve.”
      • Finally, an important factor of VRS accuracy is the need for “training” the software to recognize the speech patterns and idiosyncrasies of the speakers. Imagine asking your narrators train the software that will be transcribing the session before each of your interviews. Oy!

BUT IT AIN’T ALL BAD

There are many projects where a very rough transcript is used as a quick reference source, and an actual verbatim transcript isn’t even required. In those cases, perfect transcripts are not needed, and VRS fits the bill… As well as lowers your initial budget.

In summary, if you don’t need a near-perfect transcript, VRS is a wonderful tool at a reduced cost. If you’re looking for an accurate transcript that is also 100% guaranteed, then the only option is to call your transcription vendor of choice. You might want to try us. Call us at (617) 423-2151, or click on the GET A QUOTE link in red.

 

What Is Verbatim Transcription and When Do You Need It?

What Is Verbatim Transcription and When Do You Need It

When requesting a quote from a transcription service, it’s important to know the level of detail you require. Do you need every single word including filler (“like,” “you know,” etc.)? Do you need every utterance and sound noted? This will help you get an accurate estimate in the short term, and will prevent you from paying twice for the same project in the future.

Continue reading “What Is Verbatim Transcription and When Do You Need It?”

The Hidden Costs of “Cheap” Transcription Services

The Hidden Costs of Cheap Transcription Services - Audio Transcription Center Blog

Before you select a transcription service, ask yourself “what are my priorities?” Cheap transcription services may be tempting as a way to save money but beware: low rates do not automatically mean low total cost. These are some of the many traps and hidden costs you may find:

Continue reading “The Hidden Costs of “Cheap” Transcription Services”