27th November 2009

#

**361**Quote:

I am going to do my best to give you a thumbnail definition on what you clearly show a big confusion. If you don't get it, at least I made it clear in this thread to help those who might have gotten confused by your posts.

It doesn't matter how the FFT analysis is performed,

That all being said, I wouldn't trust HB's algorithm to measure response and would never use it as a teaching tool of anything. I wouldn't use it even as a linear phase equalizer. But that's just me. If you really want to teach your students how frequency distribution across the spectrum works and to understand about harmonic balance, then use a good quality spectrum analyzer. Mind you that those hardware units commonly known as swept spectrum analyzers, which are best suited for ultra-high frequencies, are expensive tools and don't have phase information or sufficient speed to do good transient analysis. Thus, any quality software FFT analyzer would be good to measure a basic resolution bandwidth of 20kHz.

I hope you understand now why your ban-RTA pic made absolutely no sense.

It doesn't matter how the FFT analysis is performed,

**real time or not, they are doing the same thing:**they are digitizing discrete sample values over time intervals in the time domain which can be stored and recalled from to generate spectrum. The Fourier transform process renders time domain waveforms to frequency domain. The more discrete sample values the more resolution you get and therefore, less spaces between FFT bins. So, an analog signal at a pre-determined frequency can be transformed to the frequency domain using a FFT algorithm. To achieve this function there are several ways to do so. I suspect Har-Bal does its FFT analysis directly from the audio file by measuring mathematically the frequency response {the Fourier transform of the output divided by the Fourier transform of the input} the information is transferred to a memory buffer and then rendered as a graph spectrum analysis. So, no AD conversion or filtering is needed prior to this process. As long as the analyzer is capable of resolving twice the frequency range of interest, the frequency domain information can be stored as discrete sample values. If the Fourier process is calculated before the next time-record is put in the buffer, then the analyzer is considered to be running in "real time".That all being said, I wouldn't trust HB's algorithm to measure response and would never use it as a teaching tool of anything. I wouldn't use it even as a linear phase equalizer. But that's just me. If you really want to teach your students how frequency distribution across the spectrum works and to understand about harmonic balance, then use a good quality spectrum analyzer. Mind you that those hardware units commonly known as swept spectrum analyzers, which are best suited for ultra-high frequencies, are expensive tools and don't have phase information or sufficient speed to do good transient analysis. Thus, any quality software FFT analyzer would be good to measure a basic resolution bandwidth of 20kHz.

I hope you understand now why your ban-RTA pic made absolutely no sense.

If you think that looking at an RTA makes you a better mastering engineer then use it. I have been saying that since the beginning of this thread.

**Do what ever pleases you.**You don't have to keep proving to any and all what a great tool an RTA is. I think by now we all get that you use an RTA and think that it is important in your work. If it makes your work easier. more accurate or more fun then use it.

As to teaching. I use many tools to help my students learn about audio and mastering. Since you seem to nit pick on everything I say I am not inclined to list them here but suffice it to say that I use the best tools I can to help my students in their quest for knowledge.

I had a GREAT Thanksgiving with plenty to eat and some time off.

I hope you had a GREAT turkey (or the vegetarian equivalent) day as well!