[COMPLETE] Book of Thousand Nights & a Night (Arabian Nights), Vol 12-mtf

Solo or group recordings that are finished and fully available for listeners
Post Reply
madhatteralice
Posts: 4
Joined: June 6th, 2021, 2:10 pm

Post by madhatteralice »

Hi I would like to claim sections 20 through 22 and this is my first recording so you can refer to me as Ally.
SkyRider
Posts: 700
Joined: May 3rd, 2013, 10:27 am
Location: Cambridge, UK

Post by SkyRider »

BrizeCrize wrote: June 7th, 2021, 3:29 pm Attached is section 43,
Super - section 43 is Spot PL OK. Thank you, Brize!
SkyRider
Posts: 700
Joined: May 3rd, 2013, 10:27 am
Location: Cambridge, UK

Post by SkyRider »

madhatteralice wrote: June 7th, 2021, 6:37 pm Hi I would like to claim sections 20 through 22 and this is my first recording so you can refer to me as Ally.
Welcome to LibriVox, Ally! Before you record your first piece we like to you record a test piece first. This isn't an audition but serves to make sure that you can produce files in a format that LibriVox requires. You can find details on submitting your 1-minute test (a misnomer since most are 20-30 seconds!) here: https://wiki.librivox.org/index.php?title=1-Minute_Test

I'll allocate the passages to you pending you passing the 1-minute test; once you've done that, you might decide that three is too many until you've had a bit of editing experience (generally when I'm reading, 90% of my time is spent editing and only 10% reading!) so let me know if you want to relinquish any.

Welcome again!

Cheers,
Paul
SkyRider
Posts: 700
Joined: May 3rd, 2013, 10:27 am
Location: Cambridge, UK

Post by SkyRider »

SkyRider wrote: June 8th, 2021, 4:08 am Welcome to LibriVox, Ally!
By the way, now you have some allocations against your name, you have your own reader page: https://librivox.org/reader/16320
BrizeCrize
Posts: 674
Joined: June 20th, 2020, 2:36 pm
Location: USA

Post by BrizeCrize »

Attached is uploaded Section 23, Tale of the Damsel Tohfah... with a running time of 22:15.

https://librivox.org/uploads/mightyfelix/thousandnightsandanight12_23_anonymous_128kb.mp3

That appears to be it for this volume, but I'll check back in case any orphans appear...
Thanks, Brize
madhatteralice
Posts: 4
Joined: June 6th, 2021, 2:10 pm

Post by madhatteralice »

i just uploaded my test bur I will also post it here so you know it's up. https://librivox.org/uploads/tests/test_madhatteralice.mp3
Trutter00
Posts: 69
Joined: May 18th, 2021, 5:10 pm

Post by Trutter00 »

simba
Posts: 41
Joined: October 7th, 2015, 9:22 am
Location: Hastings, England

Post by simba »

My photography https://sim.photos
SkyRider
Posts: 700
Joined: May 3rd, 2013, 10:27 am
Location: Cambridge, UK

Post by SkyRider »

BrizeCrize wrote: June 11th, 2021, 12:30 pm Attached is uploaded Section 23, Tale of the Damsel Tohfah... with a running time of 22:15.
Section 23 is PL OK! Sorry for the slow turnaround.

Cheers,
Paul
SkyRider
Posts: 700
Joined: May 3rd, 2013, 10:27 am
Location: Cambridge, UK

Post by SkyRider »

Thank you, Trish - that was an enjoyable and captivating read. May I ask you to correct one technical issue though? We try to ensure that LibriVox audios are a similar loudness so the listener is not constantly having to change volume between tracks and your reading is a little on the quiet side. If you could amplify it by 3.8bB that would be perfect; because your reading is so expressive you don't have the headroom for this you'll need to lightly compress it first. If that sounds like gibberish, let me know and I'll put that in more step-by-step terms.

Cheers,
Paul
SkyRider
Posts: 700
Joined: May 3rd, 2013, 10:27 am
Location: Cambridge, UK

Post by SkyRider »

Trutter00
Posts: 69
Joined: May 18th, 2021, 5:10 pm

Post by Trutter00 »

SkyRider wrote: June 21st, 2021, 7:13 am
Thank you, Trish - that was an enjoyable and captivating read. May I ask you to correct one technical issue though? We try to ensure that LibriVox audios are a similar loudness so the listener is not constantly having to change volume between tracks and your reading is a little on the quiet side. If you could amplify it by 3.8bB that would be perfect; because your reading is so expressive you don't have the headroom for this you'll need to lightly compress it first. If that sounds like gibberish, let me know and I'll put that in more step-by-step terms.

Cheers,
Paul
Ummmm, I’ll take that step by step. I use GarageBand to record.
SkyRider
Posts: 700
Joined: May 3rd, 2013, 10:27 am
Location: Cambridge, UK

Post by SkyRider »

Trutter00 wrote: June 22nd, 2021, 8:32 pm Ummmm, I’ll take that step by step. I use GarageBand to record.
OK. I've never used GarageBand so I'll have to give you a fairly generic explanation and leave you to find the appropriate options in the menus.

All the issues here stem from us wanting to have all the files at a comparable loudness. I'll come back to this in a couple of paragraphs time.

Sometimes you can get away with simply amplifying (or deamplifying) the file to suit. This isn't always desirable though as there's a maximum level you can amplify without causing a form of distortion called clipping - in short, this means that waveform for everything above the maximum volume is flattened. This is more problematic with more expressive voices as there's more of a variation between loud bits and quiet bits.

What's needed is a way of turning down just the louder bits in your file. This can be done either with something called a compressor or a limiter. Google tells me that GarageBand has a compressor that can be found under Plug-Ins | Dynamics ( https://thegaragebandguide.com/compressor ). What a compressor does is to look at any part of your file that is louder than a particular level and to reduce the volume a little if it goes above that threshold. This flattens out the loudness over your file giving you more headroom to amplify by. Note that you'll need to finally amplify by more than the original shortfall as the compression process itself reduces the volume.

So what numbers make sense for your compressor?

The first is the Threshold. Any sound below this volume will not be affected at all and 0dB represents the maximum possible value. Let's use a value of -18dB for now.

The Ratio tells the compressor how much to turn the volume down. A Ratio of 2:1 will decrease the volume by 1dB for every 2dB it is above the Threshold. So with a Threshold of -18dB, a volume of -18dB remains at -18dB, a volume of -16db is reduced to -17dB, a volume of -14dB reduced to -16dB and so on. Often people will compress harder than this, but we only need a light compression and want to avoid the artefacts that come from overdoing it so let's set the Ratio to 2:1.

Ignore the Attack parameter - that's more something that affects music rather than speech, so we'll hope the default is sensible.

Gain then tells the compressor how much to amplify by after doing the compression. I personally prefer to do this as a separate process but you can do it here if you prefer. Trying the figures above on your file, the compression changes the ideal amplification from 3.8dB to 4.1, but as we give you 3dB either way there's plenty of latitude to experiment. :-)

Assuming you read in a similar way, a similar distance from the microphone, the 3dB margin of error should mean that this will always work for you. If you vary your volume more in future you may have to look at measuring the loudness of the file. Most software does this using a measure called RMS. LibriVox uses a more sophisticated measure called ReplayGain that gives extra weighting to frequencies that the human ear is more sensitive to. That said, the mix of frequencies in your voice won't change so if you look at the RMS measure of something you know is the right loudness, you can be reasonably confident if you just hit the same measure again.

There's a GarageBand page on the wiki. It's marked as out-of-date, but gives the ID of a GarageBand user here that might be worth messaging for GarageBand specific advice. https://wiki.librivox.org/index.php?title=GarageBand

Hope that's helpful,
Paul
Trutter00
Posts: 69
Joined: May 18th, 2021, 5:10 pm

Post by Trutter00 »

SkyRider wrote: June 23rd, 2021, 10:06 am
Trutter00 wrote: June 22nd, 2021, 8:32 pm Ummmm, I’ll take that step by step. I use GarageBand to record.
OK. I've never used GarageBand so I'll have to give you a fairly generic explanation and leave you to find the appropriate options in the menus.

All the issues here stem from us wanting to have all the files at a comparable loudness. I'll come back to this in a couple of paragraphs time.

Sometimes you can get away with simply amplifying (or deamplifying) the file to suit. This isn't always desirable though as there's a maximum level you can amplify without causing a form of distortion called clipping - in short, this means that waveform for everything above the maximum volume is flattened. This is more problematic with more expressive voices as there's more of a variation between loud bits and quiet bits.

What's needed is a way of turning down just the louder bits in your file. This can be done either with something called a compressor or a limiter. Google tells me that GarageBand has a compressor that can be found under Plug-Ins | Dynamics ( https://thegaragebandguide.com/compressor ). What a compressor does is to look at any part of your file that is louder than a particular level and to reduce the volume a little if it goes above that threshold. This flattens out the loudness over your file giving you more headroom to amplify by. Note that you'll need to finally amplify by more than the original shortfall as the compression process itself reduces the volume.

So what numbers make sense for your compressor?

The first is the Threshold. Any sound below this volume will not be affected at all and 0dB represents the maximum possible value. Let's use a value of -18dB for now.

The Ratio tells the compressor how much to turn the volume down. A Ratio of 2:1 will decrease the volume by 1dB for every 2dB it is above the Threshold. So with a Threshold of -18dB, a volume of -18dB remains at -18dB, a volume of -16db is reduced to -17dB, a volume of -14dB reduced to -16dB and so on. Often people will compress harder than this, but we only need a light compression and want to avoid the artefacts that come from overdoing it so let's set the Ratio to 2:1.

Ignore the Attack parameter - that's more something that affects music rather than speech, so we'll hope the default is sensible.

Gain then tells the compressor how much to amplify by after doing the compression. I personally prefer to do this as a separate process but you can do it here if you prefer. Trying the figures above on your file, the compression changes the ideal amplification from 3.8dB to 4.1, but as we give you 3dB either way there's plenty of latitude to experiment. :-)

Assuming you read in a similar way, a similar distance from the microphone, the 3dB margin of error should mean that this will always work for you. If you vary your volume more in future you may have to look at measuring the loudness of the file. Most software does this using a measure called RMS. LibriVox uses a more sophisticated measure called ReplayGain that gives extra weighting to frequencies that the human ear is more sensitive to. That said, the mix of frequencies in your voice won't change so if you look at the RMS measure of something you know is the right loudness, you can be reasonably confident if you just hit the same measure again.

There's a GarageBand page on the wiki. It's marked as out-of-date, but gives the ID of a GarageBand user here that might be worth messaging for GarageBand specific advice. https://wiki.librivox.org/index.php?title=GarageBand

Hope that's helpful,
Paul
Yes, thanks. That seems to be what I was reading as well. I learn much better by doing, so I’ve been playing around but it’s good to have perimeters.
Trutter00
Posts: 69
Joined: May 18th, 2021, 5:10 pm

Post by Trutter00 »

https://librivox.org/uploads/mightyfelix/thousandnightandnight12_19_anonymous_128kb_2_-_6:23:21,_8_06_PM.mp3

Here's a compressed version, hope it's not too much. I appreciate the feedback and the explanation. I'm happy to make any other needed tweaks.
Trish
Post Reply