" Warning message: In n_recs * f : NAs produced by integer overflow" 90Mb Limit for soundscapespec?

[Seggie Fault]

Where Wav file is > ~90 Mb (48k, 16bit, stereo files)

> test<-readWave("test~91Mb.wav")

>soundscapespec(test,  plot=TRUE, col="darkgreen")

Warning message:
In n_recs * f : NAs produced by integer overflow

Smaller wav files do not exhibit this behaviour; further testing reveals the cut-off to be about 90 Mb.

Which is an unfortunate length for some musical audio.

Any chance of a slight expansion on the type to something a bit bigger?

Many thanks :)

[Jérôme SUEUR]

Dear Seggie,

One step of the process of computing this frequency spectrum consists
in scaling the STFT by the number of FFTs * sampling frequency.

With a long sound this number might overpass the integer limits of R.

A solution could be (not tried) to reduce the number of FFTs by
increasing the FFT length (wl) and/or decreasing the overlap between
successive FFTs (ovlp).

HTH

Jerome

Seggie Fault <simon...@gmail.com> a écrit :

> Where Wav file is > ~90 Mb (48k, 16bit, stereo files)
>
>
>

> *> test<-readWave("test~91Mb.wav")>soundscapespec(test, plot=TRUE,
> col="darkgreen")*
>
> *Warning message:*
>> *In n_recs * f : NAs produced by integer overflow*

>>
>
>
> Smaller wav files do not exhibit this behaviour; further testing reveals
> the cut-off to be about 90 Mb.
>
> Which is an unfortunate length for some musical audio.
>
> Any chance of a slight expansion on the type to something a bit bigger?
>
> Many thanks :)
>

> --
> You received this message because you are subscribed to the Google
> Groups "seewave" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to seewave+u...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.
>



--
Muséum National d'Histoire naturelle
UMR CNRS 7205 Institut Systématique, Evolution, Biodiversité (ISYEB)
45 rue Buffon - 75005 Paris - FRANCE
Tel. + 33 1 40 79 33 98
http://isyeb.mnhn.fr/SUEUR-Jerome
https://www.researchgate.net/profile/Jerome_Sueur

[Seggie Fault]

I see.
Redefining the problem better on my side will certainly be the next step then; please consider the matter closed

Perhaps a caveat in the notes about a maximum file size - or a similar note about R integer limitation - would not go amiss to prevent future queries.

Best regards

You received this message because you are subscribed to a topic in the Google Groups "seewave" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/seewave/Fk4u4sBoGjw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to seewave+u...@googlegroups.com.