Above image for dCS 972 spectrum, DC-22.05kHz, 1kHz at -0.1dBFS, 24-bit word length, downsampled from 96kHz to 44.1kHz (linear frequency scale, 22.5dB/vertical div.).
Infinitewave measures down-sampling performance from 96kHz to 44.1kHz. It’s spectral analysis has an excellent dynamic range of 180db!
@fmak,
Your comments:
“PC systems are problematic as you and I have found (so have the Promoters)and nobody quite knows what is going out. The fact that some actually promote 96k from 44.1 as being superior in audio terms just shows the underlying problem of using an src which itself is problematic. http://src.infinitewave.ca/.”
And later to audioengr: “asrc not good; should use integer multiples. Maths difficult especially real time.
src has plenty of images compared to good programs such as Audition. Thus
intermod.”
refers.
Secret Rabbit Code (version 1.03) has following specifications (http://www.mega-nerd.com/SRC/index.html): The current best converter provides a signal-to-noise ratio of 97dB with -3dB passband extending from DC to 96% of the theoretical best bandwidth for a given pair of input and output sample rates. Measurements from infinitewave concur with these specs albeit using version 0.1.2 (Best Sinc). Both sweep and 1kHz test tone spectral measurements show no content over -97db. The purple lines/scatter shown correspond to below -97db (colors are mapped to db levels). The so called ‘plenty of images’ and inter-modulation distortion you’re referring to would be noise below this level.
Now your choice of dCS 972 is no different to these measurements. After all it’s a PC dressed differently but nonetheless software driven with its own flavor of SRC. The spectral measurement as in image above (1kHz at -0.1dBFS, 24-bit word length, downsampled from 96kHz to 44.1kHz) suggests a SNR performance of 118db (see link below). Likewise, had Inifinitewave done similar measurements on this SRC, one would likewise see purple/blue lines and scatter as that for other SRCs.
So, can you explain how PC based systems are ‘problematic’ whilst your own choice of dCS 972 is in fact a PC with similar resulting measurements. It also outputs noise around -118db which would be very visible when graphed in infinitewave’s very impressive 180db dynamic range (either color coding frequency sweep or standard spectral graphs).
