I did a simple 3KHz mono tone using audacity 1.26 for a duraion of 30sec, I export this as a WAV file (16bit PCM) using 44.1kHz sampling rate. When I playback this wav file using a audio analyzer “Wavosaur” I can see that this audio file is not table. See attached picture for more detail.
Is this normal? Can this be removed?
We need a bit more information. We need the rest of the graph for one thing. What are the numbers along the left and bottom edges. How many octaves is included in the display? Do a screen capture.
What is the resolution of the analysis? I can make any sine wave look like that with a sloppy analysis sample rate.
Actually the left is level and the bottom is frequency. However Wavosaur doesn’t provides any value on the X & Y axis. But, the value should be at 60Hz @-10dBFS & 7kHz @ -22dBFS. The Sample rate is at 44.1kHz and the Buffer size is 1024. I realised that if I increase the buffer size to 2048, I have a better result (less flutuation). Note: This result is a persistence FFT analysis.
Any help is appreciate. Thanks!
Sorrie, I got mixed up with my earlier post. The value should be 3kHz @0dBFS level…
The strange looking results are likely to be due to the method that Wavosaur is using to analyse the audio.
I’m not familiar with what methods Wavosaur uses, but if you try the spectrum analysis in Audacity 1.3 you will see that you get vastly different displays depending on both the size, function, axis and algorithm used.