On streaming from a computer, I didn’t think it would matter much to me. But I’m trialing the Phonak Lumitys and it’s easy to hook them up to a computer. I’ve been serving as the secretary of our neighborhood association board for the past year. My modus operandi has been to record the audio of the Zoom meeting, to feed the .MP3 audio to the online MS 365 Word for Web audio transcription, then use the text from the audio transcription to do the meeting minutes. The A.I. of the audio transcription identifies speakers, eg., Speaker 1, 2, 3, 4, etc., and you can globally replace the name of a speaker in one go throughout the whole transcript, i.e., tell the transcription program that Speaker 1 = Jim Lewis.
The only problem is that the A.I. does a very good but imperfect job. There is some wrong speaker identification, particularly when speakers are interrupting each other. And some incredible transcription bloopers.
So, what I do, is I play back the audio transcript (which is in text snippets with effective play buttons for each snippet linked to its timestamp). When I find a speaker misidentification from the global replacement or an incorrect transcription, I correct it using voice recognition software. But listening to the audio behind a transcript segment streamed directly from the cloud to my trial Lumitys is fantastic. I don’t have to drive the wife crazy (my monster computer setup is on one end of our dining room-sized kitchen table) and as other Lumity users have been praising the device, the audio quality is better than my computer speakers, better than over-the-ear BT headphones, etc. The only downside is there is a definite lag of half a dozen words or so before the audio reaches the Lumitys when I click the “play button” for a particular transcript section. Whenever BT LE Audio with its low latency gains a foothold in computers and HA’s, that would make what I’m doing A LOT better. Right now, if I want to go back to a text snippet and hear the full audio, I have to start replay a snippet or two before the point where there’s a problem, and if the preceding snippet was particularly long, I waste my time just waiting to get to the audio I really want to hear again.
I’ve done lots of Minutes long ago into the past using tape cassette recorders and Sony digital recorders where’d I’d just replay the audio using skip forward, skip back buttons either on the recording/playback device itself or in a Windows media player. Playing back the exact audio snippet directly out of the audio transcription applet for MS 365 Word online into my HA’s is great and much more efficient at finding exactly what I want to replay in what would otherwise be a linear stream. The computer and the Word transcription GUI convert the linear stream into an audio random access device and the streaming quality of the Lumitys over anything else I’ve tried so far helps me get what was said right (almost all the time) when the audio quality sometimes isn’t that good for a variety of reasons including the ability of a particular speaker to speak clearly, loudly, directly into their computer microphone, not be interrupted by anyone else attempting to talk over them, interject, or affirm what the original speaker is saying, etc.