Key Takeaways
- A project coordinator at a Montreal public university needed about 4 hours 12 minutes of bilingual (English/French) audio transcribed (not translated), clean read, with speakers designated from a list she supplied.
- The provider's consultant clarified scope (languages, purpose, style, timestamps, deadline, budget), the production team reviewed six files individually, and the quote split work into French (3–4 business days), English (5–6 business days) and integration (1–2 business days).
- Work paused while the university verified procurement rules restricting purchases from U.S. vendors; the vendor accommodated institutional invoicing (purchase order, NET 15) and supplied verification details, after which the order proceeded.
- Transcripts were delivered ahead of schedule in Word as a single integrated, clean-read document with speakers as provided and marked end sections omitted; bilingual transcriptionists handled mid-sentence language switches for consistency.
- Key guidance and services: transcription vs translation distinction; choice of clean read vs full verbatim; recommend sending a speaker list, language info, start/stop timestamps, specialist vocabulary and preferred style; institutional invoicing available and standard audio formats accepted.
A project coordinator at a public university in Montreal opened our live chat on a February afternoon with a reasonable question: “Are you actually a person?” She had roughly four hours of audio from an academic symposium, in English and French, and she needed it transcribed rather than translated. What followed was a careful exchange about durations, speakers and transcript style, then a full stop while her institution checked whether it was allowed to buy the service at all, and finally the finished transcripts, delivered two days ahead of the agreed date.
The Situation
Her first message was clear. She wanted about five hours of bilingual audio transcribed, she did not need translation, and she already had a list of speakers in order with the language each had used. The recording came from an academic symposium. Explore our research transcription services for details.
Our consultant asked the questions that decide how such a job is priced and staffed: did both languages need transcribing along with speaker information, what were the languages, was the purpose general or legal, and did the languages alternate or run in blocks. Her answers were precise. English and French, general rather than legal, speaker names only to mark who was talking. Most participants had kept to one language, at the organizers’ request, for the sake of the simultaneous interpreters.
She had to leave the chat, so our consultant emailed a secure upload link. The next morning our sales manager followed up on how we approach academic research transcription.
How We Handled It
The questions that made the quote accurate
That follow-up set out the approach: transcriptionists fluent in both languages, speaker designation from her list, careful handling of mid-sentence switches, and a quality review before delivery. It then asked five things. The minute split between English and French. Clean read or full verbatim. Whether timestamps were needed. The deadline. Any budget range.
Her reply reshaped the job. The audio came to four hours, eleven minutes and fifty-five seconds, not five hours, partly because she had overestimated the total and partly because sections at the end of some files were not to be transcribed. She attached a speaker list marking where transcription should stop in the files that needed it and giving the timestamp at which each speaker began. English audio came to 2:42:12, French to 1:29:43. Clean read, no speaker timestamps, six files of unequal lengths, one integrated transcript rather than six.
Six files, reviewed one at a time
She uploaded the recordings the next morning with the speaker list. Our production team reviewed each file separately: run times from under ten minutes to nearly an hour and a half, speaker counts from one to ten, two files entirely in English and one entirely in French with the remaining three mixing the two, overlapping voices in the final minutes of one recording, and two files where the conversation stopped before the audio did.
That review turned an estimate into a scope. The quote split the work into three stages: French transcription in three to four business days, the longer English portion in five to six, and integration into one transcript in one to two. Seven to eight business days, in Word.
When the order stopped
Then everything paused. She wrote that paying in advance is not usual in a university context, where work is normally invoiced to the institution directly. More significantly, she had to confirm whether the university was permitted to use the service at all, because of new restrictions on publicly funded Canadian and Quebec institutions buying goods and services from the United States. She added that the page she had found us through had suggested we were based in Montreal.
Our consultant replied the same evening with three things and no pressure: the project could start on a purchase order with an invoice under NET 15 terms or earlier and no advance payment; the company operates remotely and internationally, which he explained plainly; and we would supply whatever her institution’s verification process needed.
Ten days later she wrote back: “we have confirmed that we can use your company for this service and would like to go ahead with the transcription.” Two days after that, having waited on a co-director’s approval, she asked for the invoice upfront and told us to proceed. It went out the same afternoon, promising delivery before 5 PM on Saturday 21 March.
Delivered in two stages, ahead of the date
The English transcripts came first, on Tuesday 17 March. Her reply next morning: “Confirmed, thank you.” The other four arrived on Thursday 19 March, two days before the promised Saturday.
The Result
The university received its bilingual transcripts in Word, clean read, with speakers designated from its own list and the marked end sections left out. Her reply: “Thank you very much for your quick work on this file. We have a graduate student preparing a summary of the event, and who will be using these transcriptions.”
Transcription Is Not Translation
- Transcription converts speech into written text in the same language it was spoken.
- Translation converts text from one language into another.
- Bilingual transcription means a recording containing two languages is transcribed in both, each passage in the language actually used.
In this project the coordinator was explicit that she did not need translation, so French passages stayed in French and English stayed in English. Wanting the whole recording in one language means transcription followed by translation, a larger job worth flagging early. Our French transcription service covers French audio on its own.
Clean Read or Full Verbatim
- Clean read is lightly edited for readability. False starts and fillers go, and the meaning is preserved. It suits symposium recordings, interviews and lectures people will actually read.
- Full verbatim keeps everything, including fillers and self-corrections. It suits linguistic analysis and any case where how something was said matters as much as what was said.
Timestamps are a separate choice: fixed intervals, speaker changes, or none. This project used clean read with no speaker timestamps. Our verbatim transcription service covers the other end.
What to Send With Your Audio
- A speaker list in order, with each name spelled as it should appear
- The language each speaker used, if the recording is bilingual
- Timestamps marking where transcription should start and stop in each file
- Specialist vocabulary, acronyms or place names likely to come up
- Your preferred style, and whether timestamps are wanted
- Whether you want one document or a file per recording
Here the speaker list was the most useful attachment. It removed guesswork about who was speaking and marked the point at which transcription should stop in the files that needed it. The transcribable duration came down from an estimated five hours to four hours and twelve minutes, partly because of those stop points and partly because she had recounted the audio herself before sending it. A duration you have checked is worth more than a rounded estimate, on your side and on ours.
Research Transcription for New York’s Universities and Institutes
New York City generates an enormous amount of recorded academic material: conference panels, oral history interviews, dissertation fieldwork and multilingual community research across Manhattan, Brooklyn, Queens and the Bronx, much of it in more than one language.
Everything can be handled remotely. Our office is at 224 W 35th St #1200-8, New York, NY 10001, between Seventh and Eighth Avenues, a short walk from Penn Station and Herald Square. See our New York transcription services, or read how a French transcript was handled for a Midtown walk-in.
Frequently Asked Questions
Can you transcribe a recording that switches between two languages?
Yes. Bilingual audio is transcribed by people fluent in both languages rather than split between two monolingual teams, which keeps terminology consistent across a switch. In this story a few speakers changed language mid-comment, and she told us she had recorded each of those points.
Do I need translation as well as transcription?
Only if you want the whole recording in a single language. Transcription writes down what was said in the language it was said. In this story the coordinator stated at the outset that translation was not needed, so each passage stayed in its original language.
How long does four hours of bilingual audio take to transcribe?
It depends on the number of speakers, the audio quality and the transcript style. For this project the scope was three to four business days for the French portion, five to six for the longer English portion, and one to two for integration. Delivery came two days early.
Can you leave parts of a recording out?
Yes, if you tell us precisely where. Mark the timestamp at which transcription should stop or start in each file and send that alongside the audio. In this story the speaker list marked where transcription should stop in the files that needed it, and with the coordinator’s own recount of the recordings the transcribable audio came to four hours and twelve minutes.
Do you invoice institutions instead of taking payment upfront?
Yes. Orders can run on a purchase order with an invoice under standard terms and payment made after delivery, which is how many universities and public bodies are required to buy services. In this story the invoice was issued before work began, the coordinator asked us for a voided check so her university could process the payment, and the transfer followed several weeks after delivery.
What file formats do you accept, and what do you deliver?
Standard audio and video formats are accepted, including large uncompressed files like the WAV recordings in this project, uploaded through a secure link. Transcripts are delivered in Word. Caption files such as SRT, VTT and SCC can be quoted separately once a transcript exists.
Have Recorded Research That Needs Transcribing?
Send your audio with a speaker list and your preferred transcript style, and we will confirm the scope and delivery date in writing. Institutional invoicing is available. Get a free quote or call 1-800-230-7918. Vanan Services has served customers since 2009 and is rated 4.9 stars from 970+ Google reviews.
This story is based on a real Vanan Services order. Names, dates and other identifying details have been removed to protect our customer’s privacy.
