This is a short demo/tutor for our usual CD Mastering Final QC, or a null test.
CD format is 16bit/44.1kHz sampling rate. Every second there are total 88,200 sample data. 44,100 samples x 2 channels
CD with exactly 1 hour long carries 317,520,000 samples of data. Each sample is 16bit = 1 out of (2^16) 63353 value.
CD ripping means we move the data from the RedBook audio CD to HDD. For audiophile who think perfect CD ripping task is still meant for Ethan Hunt, should learn about null test.
#1 – First step for the Final QC – CD Ripping
#2 – Loaded original CD Master Files. The one we sent to the plant for CD pressing.
For audiophile, you can rip the CD second time to get second sets of data, and place them on Track 1, Track 2 like above.
Track 1 – Top blue colour are the ripped CD data.
Track 2 – Bottom yellow colour are the original CD Master Files. (or your second CD ripping data)
#3 – Zoom in and line up Track 1 & Track 2, so that they are exactly at the same timeline.
#4 – On Either track, invert both channels Phase (Polarity)
#5 – We sum Track 1 and Track 2 together, adding them together. If Track 1 & 2 carry exact same data, after either track is inverted phase, they will cancel each other perfectly during add together. Perfect phase cancellation is very easy to be done in digital domain because digital is precise and accurate.
#6 – Here you can see the result of ripped CD data add together with the inverted CD Master Files.
The result peak audio level is -∞ dB
This CD is longer than 1 hour in length.
It carries more than 317,520,000 samples of digital audio data. Each sample is 16bit = 1 out of (2^16) 63353 value.
regardless how many different samples or value of each sample, they cancel each other with the original CD Master Files.
This proves they are the exact same data. The Final QC is done~!
If you ripped CD twice and compare them using this same method, you can also determine if your CD ripping procedure is perfect, on not.
If files carry the same data, they are a clone, identical.
If files carry the same data, they will sound the same.
If you feel familiar, indeed I wrote the exact same article 10 years ago and I am repeating it once. 🙂