[grc] Weird new audio editing software
Frieda Werden
wings at wings.org
Fri Dec 9 16:21:21 PST 2022
Interesting to learn what workarounds folks are coming up with to do
complex editing.
For now, I just use Audacity in mono with label track and make divisions
and notes as I go along, noting what might be moved or featured. I do most
of my level adjustments by hand. I do spend hours and hours on fiddling
with shows (one of my friends said audio editing is like sculpture). As I
age I am getting a bit more interested in likely shortcuts, but also not
enthusiastic about investing in learning new tricks that might not be as
good as what I've got.
Frieda Werden, Series Producer
WINGS: Women's International News Gathering Service www.wings.org
On Fri, Dec 9, 2022 at 11:06 AM Carl Etnier <carl at wgdr.org> wrote:
> Thanks for sharing these programs and the workflow!
>
> I use AI-generated audio transcripts for finding needles in haystacks,
> like short (say, 30 seconds to a couple minutes) clips from an hour-long
> news conference.
>
> For editing my own interviews, I can work quickly enough within Audacity
> that I'm satisfied with that method. After saving a raw recording of the
> interview, I speed-listen to the interview in Audacity, with a Label Track
> below the audio track(s). Where there's something that needs to be cleaned
> up or potentially edited out, I make a note in the Label Track. Then I go
> back and do the editing.
>
> That's for interviews that just need to be tightened and cleaned up a bit.
> If I'm producing a segment with multiple voices, from multiple interviews,
> I'll still use the Label Track and speed listening, but I'll jot down quick
> notes about who is saying what as I listen. Then I'll save the label tracks
> into text files (which include time stamps), open up the text files and
> highlight the bits of the interviews that seem to be most important, and
> cobble together the segment from there.
>
> Your mileage may vary! It's always interesting to hear what others are
> doing.
>
> --Carl
>
> Carl Etnier (he/they)
> CVCR Programming
> 802.441.3337 (mobile)
> wgdr.org
>
>
> On Wed, Dec 7, 2022 at 3:54 PM Frieda Werden via grc <
> grc at maillist.peak.org> wrote:
>
>> I just attended a Zoom demonstration session with the programmer who
>> created Trebble, hosted by the NCRA. I think there will be a recording
>> available and that I will be allowed to share the link. It looks
>> potentially quite useful. If not for finalizing a show, at least for
>> roughing it out. There are quite a few bells and whistles, all optional
>> to
>> use, including noise reduction/removal that may be better than what I've
>> been able to find on Audacity. The demo went very quickly so I didn't get
>> to savour the sound changes.
>>
>> This website includes a link to try it for free. Would be interested to
>> hear from anyone who does:
>> https://www.trebble.fm/#make-audio-editing-a-breeze
>>
>> Trebble lets you store drafts, undo, etc. It can record from your mic or
>> you can upload audio. It automatically crossfades ambience when you make a
>> delete edit (don't know if that is optional). Seems fairly inexpensive.
>> You
>> pay for blocks of time.
>>
>> Arbie Fru, a producer at the Nanaimo BC station, has started using it.
>> When
>> the snow melts, I may go down there (about an hour from here) and watch
>> him
>> work. Interestingly, the creator of the app sounds as if he is likely
>> from
>> Africa, though maybe living in Canada (I am not experienced enough to pick
>> out language of origin from his accent). He said he developed the app to
>> help a group of women who wanted to start a podcast but had no familiarity
>> with audio editing.
>>
>>
>> Frieda Werden, Series Producer
>> WINGS: Women's International News Gathering Service www.wings.org
>>
>>
>> On Wed, Dec 7, 2022 at 2:17 AM Spencer Graves via grc <
>> grc at maillist.peak.org>
>> wrote:
>>
>> > Hi, Frieda et al.:
>> >
>> >
>> > I've not used Trebble. I'll be interested in the reports.
>> >
>> >
>> > I've been using I've been using computer transcription for a
>> few
>> > years now. From that I get time stamps to the nearest second. I record
>> > where I want to cut in spreadsheet software like MS Excel, LibreOffice
>> > Calc or Google Sheet and compute the times to keep and to cut. Then I
>> > switch to Audacity to get cut points to the nearest hundredths of a
>> > second and modify the spreadsheet as I edit.
>> >
>> >
>> > An example transcript with the accompanying audio file is
>> > available at:
>> >
>> >
>> > https://otter.ai/u/7eRY0BzhugBlyu5FPXteeWdfKoc
>> >
>> >
>> > The spreadsheet edit plan I created for that transcript is at:
>> >
>> >
>> >
>> >
>> https://docs.google.com/spreadsheets/d/19h4QyJMHK2VqmD_FmAG55HcQ3e8q4Fu5zMjot9XlRf0/edit?usp=sharing
>> >
>> >
>> > The podcast I created from that is available at:
>> >
>> >
>> >
>> >
>> https://kkfi.org/program-episodes/kc-homeless-union-and-hope-for-the-homeless-project-2/
>> >
>> >
>> > I started using Trint but switched to Otter.ai. I need to
>> > identify
>> > the speakers, but I often only have to identify the first few times each
>> > speaker talks, because Otter.ai does moderately well at identifying
>> > subsequent orations by the same speaker -- and makes it easy for me to
>> > correct the speaker. The transcript is not good enough to publish but
>> > is good enough to use for timing, usually. Occasionally, Otter.ai gets
>> > the times out of order. It allows me to download the transcript with
>> > speakers and time in, e.g., docx format. Sadly, it sometimes
>> > incorrectly marks the times, then sorts the segments in the wrong order
>> > when I download the transcript. Then I have to manually correct that
>> > problem.
>> >
>> >
>> > Doing this may be more labor intensive than it should be.
>> > Yesterday,
>> > I recorded an interview of just over an hour and needed to cut it to 28
>> > minutes. I ran out of time to complete my editing plan before broadcast
>> > time at 6 PM Central, so the result was 28:31 mm:ss; I had two or three
>> > more cuts I wanted to make but gave up.
>> >
>> >
>> > Otter.ai offers a free service for files that are no more than
>> > 30
>> > minutes each up to a total of 300 minutes per month. I have a paid
>> > account for $8.33 per month, up to 90 minutes per recording and a total
>> > of 1200 minutes per month.
>> >
>> >
>> > And I sometimes use this to create a complete transcript with
>> > time
>> > stamps, e.g.:
>> >
>> >
>> >
>> >
>> https://en.wikiversity.org/wiki/Social_construction_of_crime_and_what_we_can_do_about_it
>> >
>> >
>> > However, cleaning up a transcript and adding links and
>> > references
>> > like this can take another day or three. And without serious
>> > references, it's not suitable for Wikiversity.
>> >
>> >
>> > I'm willing to believe that there are better options both in
>> > terms of
>> > the transcription service and easier editing like Trebble. I've heard
>> > of a service that supports editing videos like this, but I can't
>> > remember it's name. I haven't tried it but probably should.
>> >
>> >
>> > Comments?
>> > Spencer Graves
>> >
>> >
>> >
>> > On 12/7/22 12:33 AM, Frieda Werden via grc wrote:
>> > > https://youtu.be/5ah0IoZCIxI
>> > > It's called Trebble. It lets you edit audio from its transcription.
>> > Also
>> > > has a suite of effects. Canada's national campus and community radio
>> > > association has partnered with Trebble and is offering discounts for
>> its
>> > > members. They are recommending it for beginner editors. I think it
>> > likely
>> > > there will be a zombie-like quality to audio edited this way - but it
>> > might
>> > > be useful as a way to start cutting down longer pieces. I haven't
>> tried
>> > > it, but I might attend their demonstration workshop. It also says it
>> > lets
>> > > you capture audio from your website....
>> > >
>> > _______________________________________________
>> > grc mailing list
>> > grc at maillist.peak.org
>> > http://maillist.peak.org/mailman/listinfo/grc
>> >
>> _______________________________________________
>> grc mailing list
>> grc at maillist.peak.org
>> http://maillist.peak.org/mailman/listinfo/grc
>>
>
More information about the grc
mailing list