[FFmpeg-user] aevalsrc question

Muhammad Faiz mfcc64 at gmail.com
Mon Dec 19 17:08:34 EET 2016


On 12/19/16, Adam Puckett <signalsender103 at gmail.com> wrote:
> On 12/19/16, Nicolas George <george at nsup.org> wrote:
>> L'octidi 28 frimaire, an CCXXV, Adam Puckett a écrit :
>>> When you say it "works" for you (I have to double the ^ to escape on
>>> Windows),
>>
>> My condoleances.
>>
>>>	    does it produce this file?
>>
>> Yes, approximatively. Are you just hearing Vorbis artifacts?
> No, because when I ffplay the same command, or generate uncompressed
> files, I hear the same thing. I've checked the math and nothing
> appears to be wrong according to the documentation, because when t =
> 10, 10/10 = 1, 262*2^1 = 520, so the formula should produce a 520 Hz
> sine at the end, but the resulting frequency at the end is close to 1
> kHz. This is puzzling.

The formula is wrong.
Consider sin(262*2^(t/10)*2*PI*t).
phase, ph(t) = 262 * 2^(t/10) * 2 * PI * t
angular freq (in radians/s), w(t) = dph(t)/dt = 262 * 2 * PI *
(2^(t/10) + log(2)/10 * 2^(t/10) * t)
freq(t) = 262 * (2^(t/10) + log(2)/10 * 2^(t/10) * t)
so, at t = 10, freq(10) = 262 * (2 + log(2) * 2) = 887 Hz

Thx


More information about the ffmpeg-user mailing list