Determining which noise source defines the audible noise-floor of a recording can be challenging. Grawk's proposed test is a good one for discerning the increased noise floor of recorder vs preamp at various settings, but may not help much in determining what defines the lowest noise floor when using more optimal settings. Is it the environment, microphone self-noise, preamp, recorder's analog input stage?
Of course system noise matters, but we can manage it pretty well with proper attention. There is no question that the dynamic range of inexpensive gear has improved significantly over the years. That's good. Yet I wonder if loss of awareness about the fundamentals of good gain structure may end up being an unfortunate byproduct of the widespread adoption of 32-bit float recording.
Years ago I put a good amount of effort into optimizing the gain structure of my 24-bit stealth rig to achieve the maximum possible dynamic range. I never measured, but I think I got it to around 100dB or so which is reflects the dynamic range capability of the mics.. and probably the real-world dynamic range of the 24bit recorder I was using as well. That process started with choosing the appropriate mic sensitivity, then determining the optimal input setting on the recorder, then carefully tweaking the fine gain trim of the preamp such that I could choose between three gross gain settings when recording any music performance. I mostly used just two. The most sensitive setting was appropriate for the most demanding recording environments in terms of noise-floor and overall dynamic range, and was determined by my desire to keep the impulse peaks of the loudest applause from a potentially over exuberant person sitting immediately adjacent just below clipping. That was the practical constraint, and at that point it was still not a simple task to determine if either microphone self-noise or HVAC defined the noise-floor of the recording. I later determined that the noise-floor was indeed dominated by the ambient noise-floor of the performance space, which included a couple modern purpose built, highly acoustically isolated classical music halls which are the quietest places I've recorded music with an audience present.
I used the second, lower gross gain setting for most PA amplified material, where the peak SPL of the music regularly exceeds that of adjacent applause impulse peaks - the specific setting determined somewhat more arbitrarily via experience. In the third lowest gain position, the highest SPL the recording chain was able handle was determined by the SPL limitations of the microphones. That one useful for especially loud material, but one I rarely used.
Three gain settings- highest determined by the max SPL of the environment (with the noise-floor falling where it may upon optimal gain staging, fortunately also determined by the environment), the lowest being determined by the SPL limits of the microphones (the noise floor in those decidedly non-quiet situations pretty much always being ambient audience noise), and one somewhere in between that suited most amplified music situations. It required an external preamp to do that.
With a goal of eliminating the external preamp among other things, I'm still trying to rebuild an equivalent performing rig using newer recorders, yet can't quite do so yet while achieving the same specs. Passing off all gain-staging to the manufacturer of the recorder would be convenient, but is not quite there yet for me.
Topic: Is system noise a thing of the past, which can usually just be ignored? (Read 625 times)