Re: conversion from low to high (Marty Huntzberry)
Travis McMurry <ttmcmurry-/[email protected]>
| Newsgroups | gmane.comp.audio.mp3.encoders |
|---|---|
| Message-ID | <[email protected]> |
It looks like you're talking about two things at once with your question & commandline. Bitrate = Quality of Compression & Bandwidth (--resample) = Audio Resolution. They are two completely different but things which directly affect what you are calling "quality." Assuming the 96kbit stream was < 44.1khz, in CBR mode, using same quality setting.. There would be no improvement in quality. In fact a slight reduction in quality would most likely occur due to recompressing with a lossy compression format such as MP3. Here's why: Basically it falls down to understanding the fundamentals of digital audio and how compression works. Regarding audio in general: Frequency: The number of samples taken per second (44,100 hz is CDquality) Bits: The amount of information taken per sample (16-bits is CD quality) Bandwidth (resolution): The amount of audio which can be stored in a given file, comprised of frequency & bits. This directly affects what you perceive as quality. For example a Stereo DVD-Audio File is 24-bit 192,000hz resolution - Stereo CD-Audio is 16-bit 44,100hz resolution. Just looking at the numbers alone shows more informaiton can be stored in the DVD-A file, hence better quality. Regarding compressed audio: Bitrate: Number of kilobits per second allocated to a compressed audio stream (the higher the number, the more accurately a compressed audio stream will resemble the original) All of these factors come into play when considering overall "quality" of an audio file. So that should explain why going from a lower frequency file to a higher frequency file does nothing but increases filesize because the source of the audio can only use up the bandwidth that has already been allocated. It also explains why you can go from a higher resolution file to a lower one and lose an equally proportional amount of audio information because it can no longer fit in the smaller space. This same ideaology also explains why if you go from an identical resolution MP3, which is 96kbit/s to 128kbit/s there is no improvement in quality. Yes, there is more space per second to store data but you cannot create quality by simply increasing the bitrate. The opposite is also true: if you go from 128kbit/s to 96kbit/s you're reducing the space per second to store data thus reducing the quality of the file being compressed. Also note every time you (re)compress with MP3 audio information (quality) is lost due to the nature of the format. There is more to it and I'm really trying to simplify it. Hope that explains! Travis McMurry [email protected] wrote: I'm wondering if by using the following: lame -b 128 --resample 44.1 low.wav high.mp3 if a 96 kbits stream audiodumped to low.wav by mplayer would actually be a higher quality and better sounding mp3 (high.mp3) that would have 128 kbits? Marty ------------------------------ _______________________________________________ mp3encoder mailing list [email protected] https://minnie.tuhs.org/mailman/listinfo/mp3encoder End of mp3encoder Digest, Vol 36, Issue 2 ***************************************** --------------------------------- Want to be your own boss? Learn how on Yahoo! Small Business.