follow up on last one
- Thread starter RAO
- Start date
AlChuck
Well-known member
AlChuck
The data cannot be more acccurate than the sample rate that it was recorded at.
When you convert from 16-bit files to 24-bit files, all you can do is interpolate the sample values to spread them out over the extra available bits. Think about it for a moment. At 16-bit you have 65,536 possible different values for an individual sample in the waveform. When you made the recording the analog-to-digital converters converted the input signal to a whole blizzard of such samples. Take one sample, maybe it's value came out to be 2,222. Maybe the mext is 2,223. The converters creating those samples decide this based on what the voltages at those instants were. The voltage values are actually anywhere between these "steps" or bins that they can be chopped up into, so there's always a little inaccuracy in this conversion. But once you have done this step, it's data.
Now if you had been recording at 24-bit resolution, the converters would have 256 smaller divisions to use. So the value they record for the same waveform's same voltage value at the same instant could be much more precisely decided -- smaller steps between possible values. Now if you scale the range from 0-max value, between these two -- every 256 24-bit samples being equivalent to one 16-bit sample -- our more-precise converters might decide that the first sample was actually 568,840 and our second was 569,120.
Now if we take our 16-bit recording and change the sample rate to 24-bit, 2,222 -> 568,832 and 2,223 -> 569,088. There is no way for the software, after the fact, to know that a 24-bit converter would have read these instead as 568,840 and 569,120. It could only make wild guesses based on no information.
So the resolution of the data captured at 16 bit cannot be improved simple by saving it in 24-bit file form. It's still the same data and sounds identical.
Sorry!
The data cannot be more acccurate than the sample rate that it was recorded at.
When you convert from 16-bit files to 24-bit files, all you can do is interpolate the sample values to spread them out over the extra available bits. Think about it for a moment. At 16-bit you have 65,536 possible different values for an individual sample in the waveform. When you made the recording the analog-to-digital converters converted the input signal to a whole blizzard of such samples. Take one sample, maybe it's value came out to be 2,222. Maybe the mext is 2,223. The converters creating those samples decide this based on what the voltages at those instants were. The voltage values are actually anywhere between these "steps" or bins that they can be chopped up into, so there's always a little inaccuracy in this conversion. But once you have done this step, it's data.
Now if you had been recording at 24-bit resolution, the converters would have 256 smaller divisions to use. So the value they record for the same waveform's same voltage value at the same instant could be much more precisely decided -- smaller steps between possible values. Now if you scale the range from 0-max value, between these two -- every 256 24-bit samples being equivalent to one 16-bit sample -- our more-precise converters might decide that the first sample was actually 568,840 and our second was 569,120.
Now if we take our 16-bit recording and change the sample rate to 24-bit, 2,222 -> 568,832 and 2,223 -> 569,088. There is no way for the software, after the fact, to know that a 24-bit converter would have read these instead as 568,840 and 569,120. It could only make wild guesses based on no information.
So the resolution of the data captured at 16 bit cannot be improved simple by saving it in 24-bit file form. It's still the same data and sounds identical.
Sorry!
dachay2tnr
One Hit Wonder
What AlChuck sez!!
Think of it like a photograph that you try to blow up to poster size. All of sudden you see the dots and grainy-ness - because the needed resolution was not there to begin with.
Same with sound. If the resolution was not there to start, you can't create it.
Think of it like a photograph that you try to blow up to poster size. All of sudden you see the dots and grainy-ness - because the needed resolution was not there to begin with.
Same with sound. If the resolution was not there to start, you can't create it.
