Record Level using the record example

Status
Not open for further replies.

DaQue

Well-known member
The recorded sample sounds great in Audacity but when I play it back the level bar maxes out in the red at 0 on voice peaks. Is this the proper level or should I attenuate it a bit? I am using the line level input on the audio shield. I found the "Find clipping" tool under analyze and it does show somc clipping. Also the wave form goes to 1 on the positive peaks but only to about -.75 on the negative peaks. I thought the line level was isolated by a cap on the input. Should I try to rebalance the center bias? Fullscreen000.png.
 
PS you can play both raw and wave files. Which would have have the fewest dropouts playing from an SD?
 
The recorded sample sounds great in Audacity but when I play it back the level bar maxes out in the red at 0 on voice peaks. Is this the proper level or should I attenuate it a bit?
Yes, attenuate it somewhat - aim for about -6dB peaks and you'll have reasonable amount of headroom
Also the wave form goes to 1 on the positive peaks but only to about -.75 on the negative peaks. I thought the line level was isolated by a cap on the input. Should I try to rebalance the center bias?
This i sperfectly normal with audio signals, many sound sources have even harmonic content leading to asymmetrical waveforms.
 
Thanks. I wasn’t aware this offset was normal. I will do a voltage divider to cut the peak to peak voltage in half voltage. If that is not enough I will adjust as needed. I am working with announcements stored in an ancient “analog voice storage” chip and once I have good copies of all 7 announcements I can restore the corrupted ones on other chips. The levels only have to not clip on one copy of each.
 
Last edited:
The 6db attenuator did the trick and the waveform is now centered around 0v in Audacity too. My samples are all stored and ready. Now to wait on the next revision of my pcb to arrive.
 
The 6db attenuator did the trick and the waveform is now centered around 0v in Audacity too.
Glad it worked out - note that the peak asymmetry is not a DC offset - it can happen whenever a signal has even harmonics, and
bears no relation to the mean value of the signal. The simplest demonstration is a rectangular wave with a non-50% mark-space
ratio.
 
Status
Not open for further replies.
Back
Top