ERROR in Plot Spectrum?

[Rick944S]

Hi this is my first post here. I've been stumbling along the path to mastering Audacity sporadically over the years.
I just used the Generate>Tone command and placed a beep at 440hz, amplitude 1. When I analyze that segment it indicates a peak of 6db.
Shouldn't it return 0db?

Screen capture attached
Cheers
Rick
WIN 7 64
Audacity 2.0.6
Install from downloaded exe file

[steve]

Shouldn't it return 0db?

Ideally, yes.

If I generate a 440 Hz tone, amplitude 1.0, duration 30 seconds with all other settings in Audacity at the default values, I see a peak in Plot Spectrum that is very close to 0 dB.
What would I need to do to get the results that you are demonstrating?

[Rick944S]

Ohhh, the only difference is that I just did it on a section of track that had audio imported from WAV file, drag selected a roughly 10 second selection and created the beep - then did the analyze on that same selection.
Now when I just tried it on a new file, new Audacity session, I got the same (correct) results you got. Hmnmnmm.

[Gale Andrews]

Ohhh, the only difference is that I just did it on a section of track that had audio imported from WAV file, drag selected a roughly 10 second selection and created the beep - then did the analyze on that same selection.
Now when I just tried it on a new file, new Audacity session, I got the same (correct) results you got. Hmnmnmm.

Might you have simply changed the selection and forgotten to replot?


Gale

[steve]

I just did it on a section of track that had audio imported from WAV file

That'll be it.
If the track was a stereo track (which I'm guessing it was), then generating a full-scale sine tone will produce two identical sine tones, one in each channel. The spectrum, analysis shows the sum of all selected channels, thus two identical channels will show as 6 dB greater than one channel.

[Gale Andrews]

I just did it on a section of track that had audio imported from WAV file

That'll be it.
If the track was a stereo track (which I'm guessing it was), then generating a full-scale sine tone will produce two identical sine tones, one in each channel. The spectrum, analysis shows the sum of all selected channels, thus two identical channels will show as 6 dB greater than one channel.

I forgot it didn't average multiple channels, so I added a note about that to the Manual.

Though I now recall someone on [email protected] a while ago arguing it was a bug that it summed stereo pairs.


Gale

[Rick944S]

Interesting. I can't speak to the intended functionality of that analyze command because I still find it baffling at times, but then I'm still a newb.
Thanks though!

[steve]

Though I now recall someone on [email protected] a while ago arguing it was a bug that it summed stereo pairs.

If that was a very long while ago, it may have been me. I questioned the behaviour and accepted the developer's explanation of why it was designed that way.

[Gale Andrews]

Though I now recall someone on [email protected] a while ago arguing it was a bug that it summed stereo pairs.

If that was a very long while ago, it may have been me. I questioned the behaviour and accepted the developer's explanation of why it was designed that way.

It was someone else with a lot of scientific qualifications after his name.

What was the developer's rationale?


Gale

[steve]

What was the developer's rationale?

basically that it is as good a way as any for plotting the spectrum of multiple channels. There are pros and cons to any method.