R/E/P Community

R/E/P => Bruno Putzeys (Designer) - Dave Hecht (Master Tech) => Topic started by: Mango on June 02, 2013, 07:28:20 PM

Title: Audio Oscilloscope Comparison?
Post by: Mango on June 02, 2013, 07:28:20 PM

Hello. (Excuse the Noob, I hope I'm in the right place)

I'm very new to audio details & technicalities.  I'm trying to learn how to compare quality between 2 sounds. I'm not sure if this is relevant, but I compared and overlayed the Oscilloscope of 2 versions from the similar audio files.

I'm noticing the file that was degraded and compressed has higher peaks and more jagged edges compared to it's full quality counterpart.  Also, the full quality version has lower, but smooth waves and peaks.

Moving forward, can this be a simple method in determining a quality comparison if I have 2 versions of the same audio in 2 different files? Outside from looking at specs (i.e. bit rate, format), if I had 2 files that were 320 kbps .mp3 files, but one happened to have a history of extra compression and conversion, how would I determine the the right file?

Thanks in advance.

Title: Re: Audio Oscilloscope Comparison?
Post by: Bruno Putzeys on June 21, 2013, 08:09:02 AM

Not by looking at the waveform. With e.g. Spectrafoo you stand a slightly better chance in that you can see the subband coding at work especially at low bit rates when some subbands don't get coded at all (this is what causes "birdies" or "space monkeys").

Title: Re: Audio Oscilloscope Comparison?
Post by: John Roberts {JR} on June 22, 2013, 01:45:52 PM

One technique I like for helping "qualify" what differences are between two paths is to null or subtract the two audio stems from each other (ASSuming they have identical inputs). This will not tell you which one is wrong, but will help identify the sound that is different.

To get the best results files need to be time accurate and amplitude matched. Small time delay or phase shift will generally express as HF content in the difference. Also differences in the two stems frequency response will show up as signal equal to that response difference.

It is not trivial to evaluate differences from looking at scope waveforms, while experienced engineers spend decades building experience to that end. Note: some dual trace scopes provide a null or difference capability.

JR

PS: Hi bruno...