Re: [nfc-l] missing first few milliseconds of spectrograms made in Raven

[Bill Evans]

Thanks for the explanation Dean and Laura, and to everyone else for chiming in. 
 I haven’t made spectrograms of short call notes for publication since the 
Flight Call Guide back in 2001. I used Canary back then and the method for 
generating spectrograms must have been different because there was no such 
delay apparent.

Bill



From: Bill Evans 
Sent: Thursday, January 22, 2015 7:40 AM
To: Night Flight Call List 
Subject: [nfc-l] missing first few milliseconds of spectrograms made in Raven

When I open a short night flight call wav file in Raven, the waveform window 
begins at 0 sec but the accompanying spectrogram window doesn't begin at zero 
-- it begins after a few milliseconds. See attached example.

It’s not something that one would note unless they are working with very short 
sound files, and perhaps wish to use such a short spectrogram in a publication.

Anyone else notice this possible bug in Raven?

Bill Evans
--
NFC-L List Info:
Welcome and Basics
Rules and Information
Subscribe, Configuration and Leave
Archives:
The Mail Archive
Surfbirds
BirdingOnThe.Net
Please submit your observations to eBird!
--
--

NFC-L List Info:
http://www.NortheastBirding.com/NFC_WELCOME
http://www.NortheastBirding.com/NFC_RULES
http://www.NortheastBirding.com/NFC-L_SubscribeConfigurationLeave.htm

ARCHIVES:
1) http://www.mail-archive.com/nfc-l@cornell.edu/maillist.html
2) http://www.surfbirds.com/birdingmail/Group/NFC-L
3) http://birdingonthe.net/mailinglists/NFCL.html

Please submit your observations to eBird:
http://ebird.org/content/ebird/

--

Re: [nfc-l] missing first few milliseconds of spectrograms made in Raven

[Laura C. Gooch]

Bill,

The Cornell folks could probably answer you more definitively, but I 
think this is not a bug, but rather a result the way spectrograms are 
generated. The spectrogram comes from sampling and analyzing successive 
chunks of the waveform, so until enough time has gone by to encompass 
the initial "chunk" of the waveform there's not enough information to 
generate the spectrogram. You can change the delay by adjusting the 
spectrogram parameters, and you should be able to calculate what the 
delay will be as a function of those parameters, but there will always 
be a lag.

The lag is small enough that it won't be noticeable unless you are 
looking at a very short file, like an NFC. It may be that different 
programs make different decisions about how the spectrogram lines up 
with the wave form, so that the lag his masked in some programs.

Hope that helps.

Laura


On 1/22/2015 7:40 AM, Bill Evans wrote:
> When I open a short night flight call wav file in Raven, the waveform 
> window begins at 0 sec but the accompanying spectrogram window doesn't 
> begin at zero -- it begins after a few milliseconds. See attached example.
> It’s not something that one would note unless they are working with 
> very short sound files, and perhaps wish to use such a short 
> spectrogram in a publication.
> Anyone else notice this possible bug in Raven?
> Bill Evans
> --
> *NFC-L List Info:*
> Welcome and Basics 
> Rules and Information 
> Subscribe, Configuration and Leave 
> 
> *Archives:*
> The Mail Archive 
> 
> Surfbirds 
> BirdingOnThe.Net 
> *Please submit your observations to eBird 
> !*
> --


--

NFC-L List Info:
http://www.NortheastBirding.com/NFC_WELCOME
http://www.NortheastBirding.com/NFC_RULES
http://www.NortheastBirding.com/NFC-L_SubscribeConfigurationLeave.htm

ARCHIVES:
1) http://www.mail-archive.com/nfc-l@cornell.edu/maillist.html
2) http://www.surfbirds.com/birdingmail/Group/NFC-L
3) http://birdingonthe.net/mailinglists/NFCL.html

Please submit your observations to eBird:
http://ebird.org/content/ebird/

--

RE: [nfc-l] missing first few milliseconds of spectrograms made in Raven

[Dean L Hawthorne]

It’s not a bug, it’s a deliberate design decision.  The first spectrogram frame 
takes N samples of audio data to compute; where should we say that the 
spectrogram frame is centered, on sample 0 or in the middle at sample N/2?  It 
was decided that Raven would center the spectrogram frames on the center of the 
time window which comprises them, that is, at N/2.  The width of a spectrogram 
frame is given by the hop size, so the only time the gap will not appear is 
when the hop size is equal to N (0% overlap).

Dean

From: bounce-118726705-10098...@list.cornell.edu 
[mailto:bounce-118726705-10098...@list.cornell.edu] On Behalf Of Bill Evans
Sent: Thursday, January 22, 2015 7:40 AM
To: NFC-L
Subject: [nfc-l] missing first few milliseconds of spectrograms made in Raven

When I open a short night flight call wav file in Raven, the waveform window 
begins at 0 sec but the accompanying spectrogram window doesn't begin at zero 
-- it begins after a few milliseconds. See attached example.

It’s not something that one would note unless they are working with very short 
sound files, and perhaps wish to use such a short spectrogram in a publication.

Anyone else notice this possible bug in Raven?

Bill Evans
--
NFC-L List Info:
Welcome and Basics<http://www.northeastbirding.com/NFC_WELCOME>
Rules and Information<http://www.northeastbirding.com/NFC_RULES>
Subscribe, Configuration and 
Leave<http://www.northeastbirding.com/NFC-L_SubscribeConfigurationLeave.htm>
Archives:
The Mail Archive<http://www.mail-archive.com/nfc-l@cornell.edu/maillist.html>
Surfbirds<http://www.surfbirds.com/birdingmail/Group/NFC-L>
BirdingOnThe.Net<http://birdingonthe.net/mailinglists/NFCL.html>
Please submit your observations to eBird<http://ebird.org/content/ebird/>!
--

--

NFC-L List Info:
http://www.NortheastBirding.com/NFC_WELCOME
http://www.NortheastBirding.com/NFC_RULES
http://www.NortheastBirding.com/NFC-L_SubscribeConfigurationLeave.htm

ARCHIVES:
1) http://www.mail-archive.com/nfc-l@cornell.edu/maillist.html
2) http://www.surfbirds.com/birdingmail/Group/NFC-L
3) http://birdingonthe.net/mailinglists/NFCL.html

Please submit your observations to eBird:
http://ebird.org/content/ebird/

--

[nfc-l] missing first few milliseconds of spectrograms made in Raven

[Bill Evans]

When I open a short night flight call wav file in Raven, the waveform window 
begins at 0 sec but the accompanying spectrogram window doesn't begin at zero 
-- it begins after a few milliseconds. See attached example.

It’s not something that one would note unless they are working with very short 
sound files, and perhaps wish to use such a short spectrogram in a publication.

Anyone else notice this possible bug in Raven?

Bill Evans
--

NFC-L List Info:
http://www.NortheastBirding.com/NFC_WELCOME
http://www.NortheastBirding.com/NFC_RULES
http://www.NortheastBirding.com/NFC-L_SubscribeConfigurationLeave.htm

ARCHIVES:
1) http://www.mail-archive.com/nfc-l@cornell.edu/maillist.html
2) http://www.surfbirds.com/birdingmail/Group/NFC-L
3) http://birdingonthe.net/mailinglists/NFCL.html

Please submit your observations to eBird:
http://ebird.org/content/ebird/

--