You can check your own Technical Specifications!

All languages: post your test recording here. Help check audio files.
Post Reply
WoollyBee
Posts: 4754
Joined: January 9th, 2013, 9:08 am
Location: NY

Post by WoollyBee »

Just joined; going to be a DPL... 8-)
This is going to be SO awesome! Thanks! :mrgreen:
ArielX
Posts: 5
Joined: September 27th, 2012, 2:27 pm

Post by ArielX »

I have an idea, and it's a shame that I don't know anything about programming or I'd try this myself: what about a program like the Checker or maybe just writing a plugin for Audacity, that can take any file people have recorded and automatically converts it to the specifications required? Checker does a nice job checking the specs of files, but imagine if you'd just open a program that takes any file someone has recorded, and converts it to the 128 kbps mono -0dBs spec automatically? This would facilitate the process for everybody, and there'd be another dedicated program that all LibriVox users would be making use of. No more fussing, trials and errors trying to get the audio file to standard--the program would do it automatically. Who should I run this by?
TriciaG
LibriVox Admin Team
Posts: 60576
Joined: June 15th, 2008, 10:30 pm
Location: Toronto, ON (but Minnesotan to age 32)

Post by TriciaG »

There are some problems with that.

1. Plosives. If you have plosives, they cannot be cleaned up automatically. You must readjust your mic, etc., to prevent them.

2. Background noise. Some doesn't automatically clean up well, so it's best to prevent it.

3. Quality. If you export at 32 kbps rather than 128 and have something adjust it to 128, your quality is going to stink because you're going from poor quality to start with. ;)

There are other issues like that as well.

That said, there is Auphonic.com, which does a good job at some of those corrections that CAN be done automatically.
Serial novel: The Wandering Jew
Medieval England meets Civil War Americans: Centuries Apart
Humor: My Lady Nicotine
ppcunningham
Posts: 1142
Joined: July 21st, 2009, 3:26 pm
Location: Northern California, USA
Contact:

Post by ppcunningham »

You can also use "chains" in Audacity to automate several jobs. ( go to File, Edit Chain) It's basically an automatic que, where you set up a list of things you want it to run, with the correct specs. Then when you are ready, you apply the chain to the file.

My chain for instance, consists of:
Normalize at -3
Compress
Equalize
Normalize again at -3 (from advice from sound editors)
Bass Boost at 8 ( since I am female and have a cheap mic )


I usually noise clean first and check for clipping, but I suppose I could add that to the chain as well.
Afterwards, I run it through Checker (which is AWESOME, by the way) to check the sound level, and Amplify + or - as necessary. This has given me very good results, and I don't have to try and remember how I did it "last time". This is very helpful when you have to go back and edit a finished file due to PL notes. You can run the chain on just the bits you've fixed, so they match the rest of the file.

Patti
The trouble with life isn't that there is no answer, it's that there are so many answers. Ruth Benedict

Non curo. Si metrum non habet, non est poema

My Page
RuthieG
Posts: 21957
Joined: April 17th, 2008, 8:41 am
Location: Kent, England
Contact:

Post by RuthieG »

I must admit that I am a bit of a dinosaur, but my regime - which comprises:
  • Analyze | Plot Spectrum to get the precise frequencies for the Single Band Parametric
    2 x Single Band Parametric
    Noise Clean
    Compress
    Noise Clean
takes about 2 minutes to do. Why would I automate it, even if I could? (I can't because those precise frequencies can vary between 134-137 and between 264-269.) I'd have to Effect | Noise Removal |Get Noise Profile separately anyway, as the profile isn't saved (and it's a good thing it isn't, because I don't know many people whose noise is always identical).

Ruth
My LV catalogue page | RuthieG's CataBlog of recordings | Tweet: @RuthGolding
ppcunningham
Posts: 1142
Joined: July 21st, 2009, 3:26 pm
Location: Northern California, USA
Contact:

Post by ppcunningham »

RuthieG wrote:
Analyze | Plot Spectrum to get the precise frequencies for the Single Band Parametric
I had to look that one up -
I couldn't figure out why you would use it.

The article talked about mains hum - which I guess I don't have, because my section of town has no overhead wires.

Good to know about though - might be useful later.

P.
The trouble with life isn't that there is no answer, it's that there are so many answers. Ruth Benedict

Non curo. Si metrum non habet, non est poema

My Page
RuthieG
Posts: 21957
Joined: April 17th, 2008, 8:41 am
Location: Kent, England
Contact:

Post by RuthieG »

The article talked about mains hum - which I guess I don't have, because my section of town has no overhead wires.
A goodly percentage of readers here have mains hum to a greater or lesser extent - I can tell if someone is in the USA just by looking at their waveform, because the USA uses an electrical mains frequency of 60 Hz, while in Europe and many other places it is 50 Hz. It doesn't just come from overhead wires, but from anything electrical in the vicinity. Often readers who record on a laptop can get rid of it entirely just by recording on battery power.

I don't have mains hum, but my computer (quiet though it is) does produce noise from the fan at two specific frequencies (one is the harmonic of the other) both of them at frequencies used by the voice. Single Band Parametric whips out those very precise frequencies without affecting my voice at all, which is why I use it. Notch Filter does the same thing, but has only limited settings.

Ruth
My LV catalogue page | RuthieG's CataBlog of recordings | Tweet: @RuthGolding
ppcunningham
Posts: 1142
Joined: July 21st, 2009, 3:26 pm
Location: Northern California, USA
Contact:

Post by ppcunningham »

Is there a video?
The trouble with life isn't that there is no answer, it's that there are so many answers. Ruth Benedict

Non curo. Si metrum non habet, non est poema

My Page
RuthieG
Posts: 21957
Joined: April 17th, 2008, 8:41 am
Location: Kent, England
Contact:

Post by RuthieG »

I don't know, to be honest. Phil (philchenevert) will be able to tell you if he has made one.

SBP is also brilliant for taking out particularly sibilant esses. You can see in Plot | Analyze Spectrum the very obvious spike in the 3000s Hz and take it out, leaving a perfect s.

Ruth
My LV catalogue page | RuthieG's CataBlog of recordings | Tweet: @RuthGolding
ArielX
Posts: 5
Joined: September 27th, 2012, 2:27 pm

Post by ArielX »

Many thanks for the replies, everyone; great points made. ppcunningham: the chain idea is a great one! will try that. I just figured there's got to be a way to automate some of the processes in the path from our mouths to the resulting audio files, for everyone's sake--and yes, my own! Even though I'm learning about audio production on my own, there's a lot I've still no idea about, and the technical aspects are always a challenge for me. RuthieG, I think I'm going to camp at your feet, my dear--there's alot I can learn from you!
It took me quite a few tries to get my first introductory file polished to standard--I believe RuthieG will attest to this since, if memory serves, she was the kind person who advised me if it was up to snuff, and did so a few times (til I got it right; for your patience, I thank you). I then realized, if I had such a tough time with it, and I've been learning audio production, I can imagine how much more frustrating the process must be for folks just coming into it, handling the processes on both ends of the mike (the performance, and the production). Hence the suggestion for some automation facilitating the process, where possible. Besides, a little automation never hurt anyone. Wait--I don't know that for sure....
smijen
Posts: 6674
Joined: May 14th, 2007, 7:34 am
Location: Ottawa, Canada

Post by smijen »

Just an FYI for those who haven't seen it - an updated version of checker was posted about a month ago:
-now gives a warning if volume is too loud (rather than just too quiet)
-changed DC bias threshold so it won't give a warning if DC bias is tiny
-background noise checker is disabled by default
-fixed/change wording on several help files
-other stuff too; see version history if interested

And as always, it will remember your tab preference and open accordingly. So if you like the "Information" tab rather than the "Validation" tab, just stay there and it will open there next time.
Android users - try Orthografiend, a free word game from the maker of Checker.
TriciaG
LibriVox Admin Team
Posts: 60576
Joined: June 15th, 2008, 10:30 pm
Location: Toronto, ON (but Minnesotan to age 32)

Post by TriciaG »

YAY! Thanks for letting us know!

I've updated the Wiki page.
Serial novel: The Wandering Jew
Medieval England meets Civil War Americans: Centuries Apart
Humor: My Lady Nicotine
smijen
Posts: 6674
Joined: May 14th, 2007, 7:34 am
Location: Ottawa, Canada

Post by smijen »

smijen wrote:Just an FYI for those who haven't seen it - an updated version of checker was posted about a month ago:
-background noise checker is disabled by default
For anyone who is using an old version, please do download the new one, and leave the background noise thing-y disabled if you are a newbie. It's there for the truly geeky who want some numbers to guide their post-processing, but checking background noise absolutely requires a human ear. You can't just look at a number and know if the background noise is OK.
Android users - try Orthografiend, a free word game from the maker of Checker.
Archeanassa
Posts: 113
Joined: May 19th, 2016, 6:31 am

Post by Archeanassa »

Hello. I'm trying to install the Checker but everytime I get this message:

http://prntscr.com/b7xqu3

can anyone help me please? why is this happening.
Archeanassa
Posts: 113
Joined: May 19th, 2016, 6:31 am

Post by Archeanassa »

I tried to install again the Checker and now I got this message:

http://prntscr.com/b7y0y9

I'm not lucky with this application....
Post Reply