Suggestion: Add Producer Role

Comments about LibriVox? Suggestions to improve things? News?
annise
LibriVox Admin Team
Posts: 38542
Joined: April 3rd, 2008, 3:55 am
Location: Melbourne,Australia

Post by annise »

I'm not agreeing or disagreeing with you
However the main reason we have a volume range is so that we are able when cataloguing to equalise the volume level especially in group projects so that the final product can be listened to with earphones without having to change the volume settings each time when a very quiet recording is followed by one that damages your ear drums.
We also need to remember that many of our listeners are not listening on high quality speakers - they are jogging alongside a freeway, standing on a train station, excercising at the gym, milking the cows ........ etc.
So what it is called doesn't seem very important to me.

Anne
TriciaG
LibriVox Admin Team
Posts: 60512
Joined: June 15th, 2008, 10:30 pm
Location: Toronto, ON (but Minnesotan to age 32)

Post by TriciaG »

89 dB is about -16 to -18 LUFS.

Your scheme sounds like a lot of people managing/coordinating/engineering vs. recording. Our goal is to make audiobooks, not to make professional-quality audiobooks. You're adding a lot of additional work that adds barriers rather than forwarding our goal.

Executive Producer role is now fulfilled in the admins and occasionally Hugh (our founder).

Project Producers are our BCs, who coordinate the group projects.

As Anne said, we already have volume standards, albeit a bit flexible. And because the readers are all volunteers and because of our stated objective (to make every public domain text into an audiobook), we accept different sound quality recordings. Cheap mics are OK with us, as long as the noise level isn't too high, so the sound quality will differ - and we're not going to change our standard and exclude readers because of their mics or recording setup.

In a nutshell, you're trying to change us from a reader-focused enterprise into a listener-focused one. While we care about the quality of our recordings, our aim is not to make them all sound professional. Our aim is to have human readers make public domain texts into audiobooks, period.
Serial novel: The Wandering Jew
Medieval England meets Civil War Americans: Centuries Apart
Humor: My Lady Nicotine
ToddHW
LibriVox Admin Team
Posts: 19799
Joined: August 14th, 2011, 4:24 am
Contact:

Post by ToddHW »

Nicely said, Tricia.

Thanks, Todd
audiomike
Posts: 51
Joined: May 27th, 2016, 10:26 am
Location: Mid-Michigan

Post by audiomike »

I can envision some poor soul having loads of fun trying to edit/master a dramatic project where close proximity dialog is taking place between a person using a good mic in a quiet space and someone recording in their kitchen with the mic built into their laptop. Willing suspension of disbelief, indeed!

Seriously, though, it would require an incredible amount of work. The fact that you are not dealing with anything close to the recording quality and consistency of professional narrators would, in the end, not yield sufficient improvement to make it worthwhile, except maybe in volume leveling. Contrary to popular belief, two passes at 20 dB of noise reduction cannot fix everything.

I believe that Librivox would be better served by a structured, yet volunteer, effort to educate readers on even the most basic of production techniques. By this, I don't mean a bunch of 'How To' videos, but a live session, online course. It could utilize Zoom meetings with 4 or 5 people at a time. They would participate, say every few days for a week or two, in a live session with lecture and Q&A. At the end of each session they would have to download an audio file and apply the techniques discussed. Their results would be reviewed at the beginning of the next session. Something like this would be a big help to those that PL the one minute tests to more accurately advise new readers.
Creating the course would not be difficult for anyone with the necessary production knowledge and familiarity with Librivox. Though it would entail some effort to administer, and coordinate. And yes, I know it would take away from their reading time, but the knowledge will pay that time back in the long run.
But we can always dream.
TriciaG
LibriVox Admin Team
Posts: 60512
Joined: June 15th, 2008, 10:30 pm
Location: Toronto, ON (but Minnesotan to age 32)

Post by TriciaG »

I believe that Librivox would be better served by a structured, yet volunteer, effort to educate readers on even the most basic of production techniques. By this, I don't mean a bunch of 'How To' videos, but a live session, online course. It could utilize Zoom meetings with 4 or 5 people at a time. They would participate, say every few days for a week or two, in a live session with lecture and Q&A. At the end of each session they would have to download an audio file and apply the techniques discussed. Their results would be reviewed at the beginning of the next session. Something like this would be a big help to those that PL the one minute tests to more accurately advise new readers.
Creating the course would not be difficult for anyone with the necessary production knowledge and familiarity with Librivox. Though it would entail some effort to administer, and coordinate. And yes, I know it would take away from their reading time, but the knowledge will pay that time back in the long run.
But we can always dream.
All it takes is for someone to step up and do it.

We're all volunteers, so any initiative requires volunteers to do it.

Oh, and BTW: "Contrary to popular belief, two passes at 20 dB of noise reduction cannot fix everything." Umm, that's not the popular belief that I see reflected in the forum, in test recording responses or elsewhere. :hmm:
Serial novel: The Wandering Jew
Medieval England meets Civil War Americans: Centuries Apart
Humor: My Lady Nicotine
philchenevert
LibriVox Admin Team
Posts: 24587
Joined: October 17th, 2010, 9:23 pm
Location: Basking by the Bayou
Contact:

Post by philchenevert »

It could utilize Zoom meetings with 4 or 5 people at a time. They would participate, say every few days for a week or two, in a live session with lecture and Q&A. At the end of each session they would have to download an audio file and apply the techniques discussed. Their results would be reviewed at the beginning of the next session. Something like this would be a big help to those that PL the one minute tests to more accurately advise new readers.
Hmmm. You have my interest. Thanks to Bill Guillaume, I now have a Pro version of Zoom which is a big step up from the free version. This is something that could be done. I immediately thought of basic editing with Audacity as a start but you seem to have something different in mind, more along the lines of setting up technical specs or a recording area? Please expand on this, it opens up a lot of very interesting doors. There just might be a use for an ongoing zoom 'class' that would be open to a few people on a recurring basis. :hmm:
Peruvian owls always hunt in pairs because they are inca hoots.
89 Decibels? Easy Peasy ! https://youtu.be/aSKR55RDVpk
mightyfelix
LibriVox Admin Team
Posts: 11082
Joined: August 7th, 2016, 6:39 pm

Post by mightyfelix »

I like the overall vision of your idea. But in this one aspect, at least, I think it is asking too much of most of our volunteers:
audiomike wrote: September 23rd, 2020, 8:14 am They would participate, say every few days for a week or two, in a live session with lecture and Q&A. At the end of each session they would have to download an audio file and apply the techniques discussed. Their results would be reviewed at the beginning of the next session.
I know that I, for one, would not be able to commit to a schedule like this. And I don't even have kids. For someone with a full-time job and a family, this kind of ask would be well-nigh impossible. I wouldn't be opposed to offering this kind of crash course as an option for those who are interested, but I don't see any way we could implement it on a large scale. If this time frame were stretched out to be maybe once a week instead of two or three times a week, it might be better. But just know that you still might get a low turnout.

(Or you might not. There's no way of knowing until you try it. That's my impression of it, though.)
philchenevert
LibriVox Admin Team
Posts: 24587
Joined: October 17th, 2010, 9:23 pm
Location: Basking by the Bayou
Contact:

Post by philchenevert »

I agree with Devorah, asking people to attend regularly to a series, downloading files and doing homework seems doomed to failure. Not because of lack of interest or desire or intelligence, but because of real life stuff. But as always, if someone wants to try it, that would be great.

I am going to do this but in this way: A recurring zoom session called "Checking the 1 minute test" as a test. I would repeat the same stuff each week at different times, showing how we check for the tech specs and volume using Audacity. I'd show 'em using share screen and then let them experiment with their own programs and ask questions. 30 minutes. The idea here is that, instead of an ongoing class, this would be a one shot class with questions, repeated at intervals. I have more flexibility than most people and anyone interested in that topic could drop in. Perhaps a session early in the day (8 am) and late in the day (9 pm) for me would allow more people to join. Obviously many other areas could be explored. Let me know what you think.

As Tricia said, we are all volunteers and someone has to do it at least to try it out. Someone once asked me how I got to be the "LibriVox Video Guy" and I said that I just assumed that august mantle with the approval of 2 of our cats. The hubris has not caught up with me yet.
Peruvian owls always hunt in pairs because they are inca hoots.
89 Decibels? Easy Peasy ! https://youtu.be/aSKR55RDVpk
audiomike
Posts: 51
Joined: May 27th, 2016, 10:26 am
Location: Mid-Michigan

Post by audiomike »

Geeeez, I only meant to read the label on the can of worms, not open it.
But now that it's open, let's see what's inside.

I'm glad someone mentioned the Noise Reduction remark.
Yes, it was a stretch, but not by much. I have seen PLs of the one minute test repeatedly tell people to use as much as 16 to 20 db of unnecessary noise reduction and if they still see a 'squiggly line', to do it again. Although this is done with good intentions, they do it because they don't understand how it is supposed to work. The goal of Noise Reduction is not to make the 'squiggly line' flat. It is to properly manage your noise floor without introducing artifacts into your voice while still being able to pass the technical specs. There is a right way to set up Noise Reduction. It's something that a reader should only have to go through once and then leave it alone. But a PL needs a greater understanding of the process since they are listening to many different recordings. This is just one of the primary reasons I suggested a learning course in the first place.

Now, the part about a few days a week was just throwing it out there. We would definitely need to find out how many would be interested and what kind of schedule they would be comfortable with. The only time anyone would have to actually dedicate is for the session itself. Downloading and working on a sample file can easily be done at their leisure, especially if it was once a week. And there is no commitment involved. People could sign up for a session they want and not for others. The whole format is open to suggestions.

To answer Phil's question, it could cover as much as it needs to. But it would include things like:

An Audacity setup that focuses on getting consistent results for Librivox.
The use and misuse of Noise Reduction, Compression, and other effects.
Setup and proper use of recording equipment.
Recording environment.
And so on....

The list would only be limited by what people want.
If someone has the desire to learn, and improve not only their recordings, but those of others, they may want to participate. On the other hand, if they are content with the way they do it now, that's fine too. No one is twisting arms.

I didn't realize there were so many worms in this can. Maybe next time I'll get a smaller one.
KevinS
Posts: 15465
Joined: April 7th, 2019, 8:32 am
Contact:

Post by KevinS »

If anyone wants to teach me anything, I'll be happy to watch a live or recorded zoom presentation. One just has to give the topic title and post somewhere what time I should tune in. (Oh, I guess I would need a code, too. You all know how it works.)

I would love to hear about noise reduction, or DC bias, or where that blasted tinny sound comes from!
philchenevert
LibriVox Admin Team
Posts: 24587
Joined: October 17th, 2010, 9:23 pm
Location: Basking by the Bayou
Contact:

Post by philchenevert »

Kevin, this is off topic, but please show up Sunday to talk about poetry.
Peruvian owls always hunt in pairs because they are inca hoots.
89 Decibels? Easy Peasy ! https://youtu.be/aSKR55RDVpk
ToddHW
LibriVox Admin Team
Posts: 19799
Joined: August 14th, 2011, 4:24 am
Contact:

Post by ToddHW »

audiomike wrote: September 23rd, 2020, 8:14 am I can envision some poor soul having loads of fun trying to edit/master a dramatic project where close proximity dialog is taking place between a person using a good mic in a quiet space and someone recording in their kitchen with the mic built into their laptop. Willing suspension of disbelief, indeed!

Seriously, though, it would require an incredible amount of work. The fact that you are not dealing with anything close to the recording quality and consistency of professional narrators would, in the end, not yield sufficient improvement to make it worthwhile, except maybe in volume leveling. Contrary to popular belief, two passes at 20 dB of noise reduction cannot fix everything.

I believe that Librivox would be better served by a structured, yet volunteer, effort to educate readers on even the most basic of production techniques. By this, I don't mean a bunch of 'How To' videos, but a live session, online course. It could utilize Zoom meetings with 4 or 5 people at a time. They would participate, say every few days for a week or two, in a live session with lecture and Q&A. At the end of each session they would have to download an audio file and apply the techniques discussed. Their results would be reviewed at the beginning of the next session. Something like this would be a big help to those that PL the one minute tests to more accurately advise new readers.
Creating the course would not be difficult for anyone with the necessary production knowledge and familiarity with Librivox. Though it would entail some effort to administer, and coordinate. And yes, I know it would take away from their reading time, but the knowledge will pay that time back in the long run.
But we can always dream.
I run and edit a lot of plays here. I cast my plays first come, first serve regardless of past experience - the Librivox standard. I don't "fix" any of the received parts. There are differences in accent and pronunciations between the readers, which I embrace; and these plays are also gender neutral. I am more interested in a reader's enthusiasm and communication of the character they are playing than in their audio quality. A raspy voice for a character due to reader choice (like me deciding to record before dawn) may not be distinguishable from a later-in-the-day recording with a raspy microphone. If I have to choose between playing more roles to get better at emoting rather than taking a class in Audacity audio processing, I'd got for the former every time.

Good storytelling can be done in adverse circumstances with less-than-stellar quality and still be immensely enjoyed by the audience. A great example being released right now is "Table Top Shakespeare At Home" by Forced Entertainment: one hour videos of all 36 of Shakespeare's plays being produced solo in various kitchens using bottles and jars and flower vases for the characters. I almost cried when the actor knocked over the gluestick playing MacDuff's son as he said "and he died". As you say, "Willing suspension of disbelief, indeed!"

Thanks, Todd
knotyouraveragejo
LibriVox Admin Team
Posts: 22067
Joined: November 18th, 2006, 4:37 pm

Post by knotyouraveragejo »

@audiomike -

Recommending 1 or 2 passes with noise reduction for 1-min tests rather than a multistep process is intended to quickly get people who know nothing about audio work started recording without overwhelming them with technical jargon. If they stick around and eventually want to further improve their post-production routine, that is up to them. I do recommend this simple one step clean-up of background noise often, but at the minimum settings that clean up the worst of the background noise, and only after checking the results first myself. Making a brand new volunteer jump through too many post-production hoops will only discourage those who are just at the stage of just figuring out how to plug in the mic and record the 1-minute test.

Having said that, I agree that what you are suggesting would be a great resource for more experienced readers after they have recorded and edited some on their own with a few simple instructions. Considering the wide range of equipment and recording environments of our readers, however, there is never going to be a one-size fits all solution.
Jo
JordanN
Posts: 110
Joined: April 13th, 2014, 2:18 pm

Post by JordanN »

TriciaG wrote: September 16th, 2020, 5:12 am 89 dB is about -16 to -18 LUFS.
From Wikipedia (https://en.m.wikipedia.org/wiki/ReplayGain):
The target loudness of ReplayGain utilities is 89 dB sound pressure level.[note 1] The SPL reference comes from a SMPTE recommendation used to calibrate playback levels in movie theaters.[note 2]

A more common means of specifying a reference level is relative to a full-scale signal. ReplayGain nominally plays at -14 dB relative to full-scale leaving 14 dB of headroom for reproduction of dynamic material. In contrast, the SMPTE RP 200:2002, on which the ReplayGain reference was originally based, recommends 20 dB of headroom. The more recent EBU R 128 suggests 23 dB.[6]
So, to meet the ReplayGain “89 dB” standard, I should aim for -14 LUFS. I recommend putting this equivalence into the wiki for reference for users trying to comply with the modern LUFS mastering loudness standards.

Loudness normalization is a standard now, and I would expect to find at least basic loudness analysis tools in any DAW, including Librivox’s beloved Audacity.

(Update) Audacity Loudness Normalization: https://manual.audacityteam.org/man/loudness_normalization.html

https://manual.audacityteam.org/man/perceived_loudness_ebu_r_128.html
Last edited by JordanN on November 4th, 2020, 4:53 pm, edited 1 time in total.
TriciaG
LibriVox Admin Team
Posts: 60512
Joined: June 15th, 2008, 10:30 pm
Location: Toronto, ON (but Minnesotan to age 32)

Post by TriciaG »

I've added the LUFS level in a couple key spots in the wiki.
(Update) Audacity Loudness Normalization: https://manual.audacityteam.org/man/loudness_normalization.html
Is this a built-in effect in the new version(s) of Audacity? If so - cool! 8-) That would be very helpful!

P.S. I wonder if it allows clipping, or if it will only normalize up to that level. :hmm:
Serial novel: The Wandering Jew
Medieval England meets Civil War Americans: Centuries Apart
Humor: My Lady Nicotine
Post Reply