4/7/2024 0 Comments Decibel reference scale![]() This could be a simple volume control on a headphone amp, or on a mixing console, or dedicated monitor control, or for the "monitor" outputs on some type of external hardware computer audio interface (or even the volume knob on a stereo system if that's what you use to monitor). You should be monitoring through some type of monitor controller where you can control the volume on the Analog side After the signal has come out of your D/A converters. If you are plugging your D/A converters directly into powered speakers or an amp without any kind of analog volume control, then the only way you can control the volume is to turn the signal down digitally inside the software, which is NOT a good way to go because that's lowering your bit resolution of your audio and not fully utilizing all the dynamic range of your digital system. At some point, those digital values are converted back to analog signals via a D/A converter somewhere in your monitoring chain. While sound intensity (in W/m 2) is the SI unit, the sound intensity level in decibels (dB) is more relevant for how humans perceive sounds. The volume you hear is determined by how you are monitoring. You may have noticed that when people talk about the loudness of a sound, they describe it in units of decibels rather than watts per meter squared. What you see on your dBFS meters is only referenced to digital levels. ![]() That's where the floating point math comes in.īut, back to your question and how it relates. However, any time you do any kind of "math" inside the software (any kind of EQ or other processing, level adjustments, etc.), there is math being done and you'll get fractional results, or results that are bigger or smaller than can be represented by 24 bit math. If you stay within the normal 24 bit range of operation, the multiplier data is basically just fixed at 1, and it's the same as working on a 24 bit fixed point system. you can't go above 24 bits of data for your D/A converters or you'll clip them, even if the software you are using is 32 bit.ģ2 bit floating point is probably the most common software format, which you can basically think of as 24 bits of data, and the other 8 bits are used for the multiplier and sign. However, if your levels going into your A/D converter are too hot, you WILL clip the converters as all converters are Fixed point. However, in a lot of modern software, they use Floating Point math, where there are extra bits used for a multiplier, so you can go over "digital zero" without clipping.įor example, most A/D and D/A converters these days are 24 bits, although, due to electronic design, you only get about 20 to 22 bits of real information in the best converters, and the rest is noise from the electronics themselves. This is true for Fixed Point digital math. What that means is that when you hit 0 dBFS, all the bits are on and you can't go any higher in level without digital clipping. In digital software, the reference is "full scale", represented as dBFS.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |