After watching the video, yes the video maker is indeed misinformed about noise shaping but his heart is in the right place. He is explaining how the goal of the process is to have “quantization noise” (in his case he is confusing quantization noise with added noise) be softer (less power) at higher frequencies and more power at lower ones.
As for the “quantization error noise” shaping (as a math major ) I understand that in a dithering process, quantization error from dithering of the previous sample can be added (after processing the error using a finite impulse filter) back to the next sample like a feedback loop. And result is the same, the error gets shaped .
With noise shaping in the dithering process, we lower the noise floor in sensitive areas of the band. Certain plugins allow you to choose whether or not you wish to shape the noise caused by dithering, and they allow different methods (filters) for shaping that noise as well.
and yes this is hugely off-topic from the original topic lol
Oh—haha—didn’t notice. Well, it was a little nerdy detail, and for that matter he may have well understood the point I made, but chose to demonstrate it (the shaped noise part) as being close enough to the idea. In the sense that the noise is pushed up out of the way, it is (though the results are different). But I did like the first part about showing the truncation error, and showing how dither not only fixed it, but allowed the signal to drop below a level that it would otherwise remain viable.
Just adding nerdy details: Broadband static is white noise. The difference is that you’d have (ideally) equal energy across 100-200Hz and 2000-2100Hz, for instance. Pink noise is equal energy per octave. So, you’d have equal energy across 100-200Hz and 2000-4000Hz. Since the ear hears in log frequency, pink noise sounds more even, white noise overly bright. You can make pink noise from white with a 3 dB/Octave filter slope.
yeah thats what I meant and was referring to, its still a type of broadband static.
id rather call it red noise
red, white pink noises are all referred to as static at a generic level, there are many other different types of static for different engineering purposes (us European engineers refer to them as static really, nothing too fancy, but static, noise , error all interchangeable terms)
shouldnt that be a flat curve? brightness of the white should get even by the bass heavy pink
or something like that, not sure if phase reversal will do much
you guys are way too excited about noises … then again you are also excited about adding ice to tea calling it ice tea…and ‘revolutionizing’ tea (sorry European origin pun intended )
without a daw as well…but yep looks like it is flat. Didn’t need a daw to answer it was kind of a pretty straightforward question , got any harder ones???
What happens if you inverse brown noise, combine it with white noise, run it through a series of vacuum cleaner attachments into a Fender Champ, out into an SM-57 directly into 64-bit 192kHz in your DAW, then render it during a full moon on a leap day in a leap year?
The Captain and I are considered the local nuts, so is @Wicked.
will tell you in 30 years. Next full moon on feb 29 will occur in 2048 (almanac rules)
hopefully by then dolphins and bats will sing for humans so we can justify a sample rate of 192khz at a 64 bit recording. SM-57 will probably still be the mic everyone jerks off to and fender champ will cost a fortune in pristine condition.
No argument from my end. My video was intended for a non-technical audience, so I made no attempt to show the best way to create dither noise, only to give a simple audible example of how dithering will increase the apparent dynamic range when going to a lower bit level.
The noise shaping section was technically wrong (you can even hear the tone level decrease when the filter is turned on). I should have just left that out. But I don’t think anybody is going to go writing any white papers on dither after watching that video. The hope is that it would cause fewer “why and when do I dither” questions.
id rather call it red noise