This content has been marked as final.
Show 4 replies
-
1. Re: Tin Can Voice
Craig Jackman Mar 23, 2009 7:15 AM (in response to Rich146)MP3 is a destructive, compression algorhythm. Put simplistically, it "throws out" audio it thinks you can't hear giving the bit rate settings in the MP3 settings you tell it. So if you are recording narration at 96kHz ... which is total and utter overkill for the human voice ... and then compress that down to a 24kbs MP3, you are "throwing away" 86.4% of the data in your original audio file. You will also be going through a stage of sample rate conversion as they don't, to my knowledge, support 96kHz MP3 files.
If you want to get rid of the tin can effect, you have to use a MUCH high MP3 bit rate. The smallest I ever use is 48kbs mono for a streaming podcast. If I'm sending a voice narration from one studio to another I use 128kbs mono. If I'm sending a stereo file I use 192kbs minimum.
Basically if you want to pack more audio quality into your MP3 file, you have to make it a bigger file.
Since you are going to MP3 format, and since it's just voice narration, you are fooling yourself to be recording at 96kHz. All you are doing is wasting disk space. If I was going to end up with a small MP3 file in the end, I'd be recording at CD quality or 44.1kHz. For video, 48kHz. -
2. Re: Tin Can Voice
Rich146 Mar 23, 2009 7:45 AM (in response to Rich146)Craig
Thank you for the feedback. I will start recording at 44.1kHz and increase the MP3 to 48kbps mono. I am sure it will sound better, I was worried about the file size but I guess now days with most users equipped with highspeed internet it does not make that much difference.
I am glad to know what your settings are for streaming voice.
Thank you again for your suggestions.
Rich Gilman -
3. Re: Tin Can Voice
Wild_Duck Mar 23, 2009 9:02 AM (in response to Rich146)Rich, I don't know about the Quicktime part of the question (and don't understand the reference to 11kHz mono on the upper of the 2 boxes on your site - surely this is too low whatever it means?), but, like Craig, I have done a lot of mp3 encoding of mono speech files.
We never used to go below 56kbps as the mp3 data rate, and would usually use 64k.
We also got better quality by using a starting sampling rate on the original file of 32kHz, which gives adequate audio bandwidth for most uses and leads to less compression. -
4. Re: Tin Can Voice
_durin_ Mar 24, 2009 3:33 PM (in response to Rich146)Rich,
I sent a response to your post in the Soundbooth forum. I think the problem is your RE-Compressing the mp3 files in Flash. Flash gives you the option of recompressing the audio in your library on export, or leaving it at the default settings. the screenshots you posted indicated it was being compressed a second time.
Durin



