Sunday, August 14, 2005

Dagdag Boses Theory Refuted

Posted by Alecks Pabico 
PCIJ

AT DENR Secretary Michael Defensor's press conference yesterday, I brought up the issue of the tapes that Press Secretary Ignacio Bunye presented to the MalacaƱang press corps last June 6 because these — both the allegedly "original/unaltered" and "spliced" versions — also contain the "yung dagdag" portion, which Defensor alleges to have been inserted in the Paguia tape. Except for the deleted utterance of the word "Namfrel" in the supposedly "unaltered" version, this portion was found to be free from splicing by an independent sound expert who shared with us his analysis of the Bunye tapes.

Since there is reason to believe that all the versions that have come out have the three-hour recordings as source, I asked how come the findings of Jonathan Tiongco, Defensor's so-called expert, say otherwise. Could it be that the track he asked to be analyzed had been tampered with?

If the examination made by the same independent sound expert who analyzed Bunye's tapes (and the "Chavit X-tapes") is to be considered, there is reason to doubt Tiongco's findings. Based on the FFT (Fast Fourier Transform) Spectrograph analysis he has done on the Paguia tapes, our source has found no discrepancy or discontinuity in the " yung dagdag" conversation.

An FFT Spectrograph analysis was used to show any discontinuity or splicing done on the audio clips. The FFT Spectrograph shows the frequency-time domain of the audio being analyzed where the intensity of the red pixels on the graph shows changes in the volume of a particular frequency in the audio clip. Time is represented by the x-axis, the frequency by the y-axis with the lowest frequency found at the bottom of the graph. Splicing will be very much apparent if there are breaks in the red graph color distribution across the x-axis.

Below are his findings (click on the links to download larger images of the corresponding plates):

paguia-plate1-1.jpg

Plate 1 shows the FFT Spectrograph analysis of Paguia's version of the entire "yung dagdag " conversation. On this graph alone, you will not find any discrepancy/break or discontinuity in the conversation.

paguia-plate2-1.jpg

Plate 2 shows the FFT Spectrograph analysis of Paguia's version of Pres. Gloria Macapagal-Arroyo's portion where she mentioned " yung dagdag, yung dagdag." In this analysis, one will never find any break or splicing in the clip. The conversation is very much "continuous" and there is no abrupt change on the background frequencies.

paguia-plate3-1.jpg

Plate 3 shows the FFT Spectrograph analysis of the three-hour tapes particularly on the part where Arroyo mentioned " yung dagdag, yung dagdag." Analysis shows that the clip is pretty much unaltered. Not a single "break" or discontinuity is found here.

Tiongco's findings also mentioned that the "yung dagdag" portion was altered by means of varying the speed of the playback. But our source says that if that was the case, the background should also change in "pitch." The following plate images illustrate his point:

paguia-plate4-1.jpg

Plate 4 shows the actual waveform presentation of the "yung dagdag" portion in the three-hour tape. Notice that the background including the GSM radio buzz is consistent all throughout the recording. If there were changes in the speed or pitch in the clip, there should also be a corresponding change in the background. But in this case, the background is very much consistent on both Garcillano's and Arroyo's parts.

paguia-plate5-1.jpg

Plate 5 shows a zoomed version of the Garcillano part's background. This is actually the GSM radio noise heard on a typical cell-phone conversation.

paguia-plate6-1.jpg

Plate 6 shows a zoomed version of Arroyo's " yung dagdag" background. This also shows the same noise found on Garcillano's part and the pitch has never changed. It thus disproves allegations that the speed of the conversation was altered. A simple test of just listening to the background will show that the background noise frequency/pitch is pretty much consistent. So if there had been speed alterations, the background should have been affected and will in fact show as frequency variations in any FFT analysis.