1-minute Test [OK]

All languages: post your test recording here. Help check audio files, provide editing services, and advertise for proof-listeners.
Post Reply
HollyMasri
Posts: 4
Joined: May 18th, 2020, 7:03 pm

Post by HollyMasri » May 19th, 2020, 5:58 pm

Dear Librivoxians,
Here is the link to my 1-minute test: https://librivox.org/uploads/tests/test_hollymasri.mp3.
I'm afraid I uploaded it at least twice before I figured out what I was doing... :oops: I'm using Audacity on a MacBook Pro (2013) with a FiFine USB mic. I've ordered a somewhat better mic (Blue Yeti), which hopefully will come in the next few days.

Thank you,
Holly Masri

knotyouraveragejo
LibriVox Admin Team
Posts: 15913
Joined: November 18th, 2006, 4:37 pm

Post by knotyouraveragejo » May 20th, 2020, 5:44 pm

Hi Holly. Welcome to LibriVox.

Your test was exported at ~108 kbps variable bitrate. For LibriVox, you need to export at 128kbps constant bit rate.

In Audacity, you set the bitrate when you export to mp3. At the bottom of the export window under Format Options, choose the following settings.
Bit Rate Mode: Constant
Quality - select 128 kbps from the dropdown

I would recommend you apply a little noise reduction to tone down the background hiss in this recording. That said, the Blue Yeti will more than likely have a better noise floor so you might want to wait and see how it sounds compared to your current microphone first.

http://wiki.librivox.org/index.php/Noise_Cleaning

Please upload and post the link to a second test in this same thread to keep everything together.
Jo
My Librivox Solos
Books are the carriers of civilization. Without books, history is silent, literature dumb, science crippled, thought and speculation at a standstill. - Barbara Tuchman

HollyMasri
Posts: 4
Joined: May 18th, 2020, 7:03 pm

Post by HollyMasri » May 22nd, 2020, 11:52 am

Here is my second 1-minute test, done with my new Blue Yeti: https://librivox.org/uploads/tests/test_hollymasri.mp3. I do think it sounds a lot better... though whether good enough, I will wait to hear. I have not tried noise reduction on it yet.
I made sure it said 128 kbps constant bitrate when I exported it this time. Thought I did last time, too, but maybe not... this is all very new! :) Anyway, if those settings are still wrong, I trust you'll let me know.

Thank you,
Holly Masri

knotyouraveragejo
LibriVox Admin Team
Posts: 15913
Joined: November 18th, 2006, 4:37 pm

Post by knotyouraveragejo » May 22nd, 2020, 12:27 pm

Hi Holly. Your export settings are fixed on this test. The new mic sound great! Almost there

The volume on this recording is a little too low. (84.2 db). We aim for 89 db with an acceptable range of 86-92 db. Please try increasing the gain on the yeti. Alternately you can adjust the input volume in either Audacity or in the Sounds and Audio controls on your MAC. This adjustment will also help further increase the signal to noise with regard to the background.

To judge whether you have the volume set correctly, your waveform in Audacity should look similar to this example:

Image

Detailed instructions for adjusting your input levels can be found here:

http://wiki.librivox.org/index.php/Setting_Recording_Input_Level


You can download the checker program to check your volume after exporting to mp3 and before you upload your file.

https://wiki.librivox.org/index.php/Checker

If you prefer you can use the Replay Gain plugin in directly in Audacity to tell you how much your file needs to be amplified to get to 89 db before you export to mp3. Instructions here:

https://wiki.librivox.org/index.php/Measuring_Volume_within_Audacity

If you can't adjust your input volume to reach the target 89 db, then, you will need to use the Effect:Amplify to actually amplify the file. Important: be sure the box to allow clipping is not checked!

The background on this test is within acceptable levels. See how it sounds once you make the adjustment to the volume. Then you can experiment with the Noise Reduction and see what you think. You may not need to use it.

Please try increasing the volume and/or amplifying and upload another test. Just post the link to this same thread to keep all your tests together.
Jo
My Librivox Solos
Books are the carriers of civilization. Without books, history is silent, literature dumb, science crippled, thought and speculation at a standstill. - Barbara Tuchman

HollyMasri
Posts: 4
Joined: May 18th, 2020, 7:03 pm

Post by HollyMasri » May 23rd, 2020, 8:38 am

Dear Jo,
I very much appreciate your patience with me as I sort out how things work in this very new world! I did as you suggested last night, and played around with the gain on the Yeti (which I already love!) and the "mic" slider (recording volume) on Audacity. I can get plenty big waveforms... though I'm still confused about that. The first sample you posted below looks to me like it's "clipping". Aren't we supposed to avoid that?
:?

Anyway, I also downloaded the ReplayGain plug-in and used it on a recording I felt fairly good about... and finally exported that recording to my iCloud desktop. Then this morning, I downloaded the Librivox Checker and ran the recording through that. It took me awhile to figure out how to get the recording onto the Checker... but that's because I'm also new to Mac, AND to working in the cloud, which is where I've got the recording stored. I had to download the recording to my Mac before I could drag it on... THEN it worked. Sheesh! Talk about a steep learning curve... :lol:

The checker says that the track passes; now I'll be interested to hear what you think! :) I suspect that part of what has been confusing me is that I've been using Audacity to try to meet ACX's standards, so I can start auditioning for audiobook narration with them. When I ran the sample that passes the LibriVox checker through Audacity's ACX Checker plug-in, it said the noise floor and RMS were fine, but the peak (-2) was slightly higher than their limit of -3 db max. I'm guessing that LibriVox and ACX just do things differently...?

The sample I made last night has a slightly different name than the other two... my name, followed by "postreplaygain.mpg". If that causes problems in your system and you can't see it, or something, let me know and I'll do another. Here's the link: https://librivox.org/uploads/tests/test_hollymasri_postreplaygain.mp3

Thanks again for all your help!! :D

~Holly

knotyouraveragejo
LibriVox Admin Team
Posts: 15913
Joined: November 18th, 2006, 4:37 pm

Post by knotyouraveragejo » May 23rd, 2020, 11:14 am

Hi Holly,

The red lines in that graphic are just there to show the range. They are not the red that software adds if a peak is clipped. Clipping occurs when the waveform goes all the way over the top and bottom of the window (>1.0) and is therefore cut off . A clipped waveform is easy to spot as will be squared off flat across the top of the peak even if you deamplify the file. If it is badly clipped, you can also hear it. As for ACX standards, you are correct that they are different and probably stricter than LibriVox requirements. I haven't used the Audacity ACX Checker plug-in so I can't comment on it specifically, but I will say the RMS and Peak levels are different algorithms.

Meanwhile, this test is good to go to start recording for LibriVox. Have Fun!
Jo
My Librivox Solos
Books are the carriers of civilization. Without books, history is silent, literature dumb, science crippled, thought and speculation at a standstill. - Barbara Tuchman

Post Reply