Author Topic: Timelines and Routes, Thread 2  (Read 75744 times)

0 Members and 1 Guest are viewing this topic.

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #345 on: April 17, 2013, 10:49:32 PM »
Honestly, just use your ears.  I think you're overcomplicating this process.  Listen to it, and it's pretty obvious that he's saying "You'll see my truck".  That's why it's important that we're using the same source, so we know we're listening to the same thing (not factoring in the speakers/headphones we're using to listen).

Also, I can't imagine how anybody got "The keys are in the ignition" from the crosstalk section there.  It doesn't sound even remotely like that.

I did honestly just use my ears, and what I heard was "That's my truck." You honestly used your ears and heard "You'll see my truck." When LeeM honestly used his ears he heard "They'll see my truck." I believe the Audacity waveforms support my ears' version. Audacity is free and easily obtained. You can try for yourself saying "You'll see my truck" in a manner that approximates GZ's rhythm and inflection, and see whether it produces an identifiable blip for both "you'll" and "see." I tried many times and couldn't produce a waveform similar to the one I showed. When I tried "That's my truck," it was similar. 

I agree the next utterance sounds nothing like "The keys are in the ignition," but I don't doubt that's what someone honestly heard.
« Last Edit: April 17, 2013, 10:56:44 PM by MJW »

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #346 on: April 17, 2013, 10:55:30 PM »
I know of no evidence that this program is any good. All that I've heard about it suggests that it isn't.

Perhaps you can elaborate on what you've heard. I've used it for many purposes, and find it very useful.

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #347 on: April 17, 2013, 11:16:10 PM »
I'm trying to determine if the file/source you're using is the same as the one I'm using.  The file I'm using is 8,032,066 bytes, and it's a .WAV extension file.

Yes, same file type, same size.

Offline nomatter_nevermind

  • Hero Member
  • *****
  • Posts: 5447
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #348 on: April 18, 2013, 12:07:16 AM »
Perhaps you can elaborate on what you've heard.

I should have been clearer. I haven't heard anything except what has been said on the forum.

My point is just that I don't have any reason to believe the software is better at identifying sounds than my own ears.

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #349 on: April 18, 2013, 12:38:16 AM »
I should have been clearer. I haven't heard anything except what has been said on the forum.

My point is just that I don't have any reason to believe the software is better at identifying sounds than my own ears.

I recently mentioned Audacity for two reasons. The first in response to:
Quote
It's even harder to understand what's being said if you keep repeating anymore than that 1 second of tape. so one can always get used to holding down the rewind button(don't click) at axiom to keep repeating that 1 second portion, if you don't have the time or software/program.

I'm sure you'll admit playing a selection in a loop in Audacity is preferable. I assume you'll also admit there's an advantage to being able to slow down the audio without changing the pitch.

The second was to show the waveform of the phrase "? my truck." The program isn't trying to identify sounds; it's showing what the sounds look like. If you read the WAV file into Audacity, you'll see it looks the same, and you can verify that I've correctly labeled each component. If you're experimentally inclined, you can try recording various phrasings of "That's my truck," "You'll see my truck," and "They'll see my truck," to see if any of the versions with two syllables for the disputed section only have a single "blip," or whether, as I've found, both syllables can be clearly seen in the waveform.

Offline nomatter_nevermind

  • Hero Member
  • *****
  • Posts: 5447
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #350 on: April 18, 2013, 01:11:18 AM »
The program isn't trying to identify sounds; it's showing what the sounds look like.

Yes, strictly speaking, the program itself doesn't identify sounds. It's a tool that is used for that purpose. I was speaking loosely, but the distinction doesn't affect my point.

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #351 on: April 18, 2013, 01:30:45 AM »
Yes, strictly speaking, the program itself doesn't identify sounds. It's a tool that is used for that purpose. I was speaking loosely, but the distinction doesn't affect my point.

If your point is seeing a visual representation of the sound isn't an aid in determining what was said, I think it's an extremely weak point. Both you and Departure claim that the best approach is to just listen. But just listening has led me to believe GZ says one thing and you to believe he says something else.  That's as far as listening can take us, because your confidence that your version is correct does nothing to change my mind, and vice versa.

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #352 on: April 18, 2013, 01:42:16 AM »
I'll add that in this instance, I don't think it makes any difference whether GZ said "That's my truck," "You'll see my truck," or "They'll see my truck." They all mean the same thing.

Offline nomatter_nevermind

  • Hero Member
  • *****
  • Posts: 5447
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #353 on: April 18, 2013, 02:01:10 AM »
If your point is seeing a visual representation of the sound isn't an aid in determining what was said

That isn't my point.

ETA:

Quote
Both you and Departure claim that the best approach is to just listen.

I don't recall saying that.
« Last Edit: April 18, 2013, 02:06:33 AM by nomatter_nevermind »

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #354 on: April 18, 2013, 02:09:28 AM »
That isn't my point.

Combining:
 
My point is just that I don't have any reason to believe the software is better at identifying sounds than my own ears.

with
Yes, strictly speaking, the program itself doesn't identify sounds. It's a tool that is used for that purpose. I was speaking loosely, but the distinction doesn't affect my point.

lead me to believe that was your point. If not, what is your point?

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #355 on: April 18, 2013, 02:23:08 AM »
I don't recall saying that.

What you don't recall saying is that the "best approach [to determining what GZ said] is to just listen." You said, "I don't have any reason to believe the software is better at identifying sounds than my own ears." I think my paraphrase was pretty accurate. It certainly summarized what I thought (and think) you meant.

Offline nomatter_nevermind

  • Hero Member
  • *****
  • Posts: 5447
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #356 on: April 18, 2013, 02:55:34 AM »
If not, what is your point?

I don't know if I can say it any more clearly than I have. I'm content to leave this part of the discussion where it is.

Offline Departure

  • Jr. Member
  • **
  • Posts: 39
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #357 on: April 18, 2013, 07:34:32 PM »
My point is just that I don't have any reason to believe the software is better at identifying sounds than my own ears.

I don't feel the software is necessarily better at identifying speech than my own ears.

Also, to MJW.  The audio file of GZ's NEN call is being fed through a cell phone, to the dispatcher's location, undergoing who knows what kind of compression or other modifications before it became the WAV file we have access to.  When you are saying things yourself, I imagine you are not duplicating the recording process that GZ's voice was subjected to that night.  You're using a desktop microphone wired directly into your computer most likely, no?  Not exactly an apples to apples comparison.

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #358 on: April 18, 2013, 08:35:55 PM »
I don't feel the software is necessarily better at identifying speech than my own ears.

Also, to MJW.  The audio file of GZ's NEN call is being fed through a cell phone, to the dispatcher's location, undergoing who knows what kind of compression or other modifications before it became the WAV file we have access to.  When you are saying things yourself, I imagine you are not duplicating the recording process that GZ's voice was subjected to that night.  You're using a desktop microphone wired directly into your computer most likely, no?  Not exactly an apples to apples comparison.

All true, but that doesn't covert two syllables into one.

Offline MJW

  • Hero Member
  • *****
  • Posts: 1304
  • Rate Post +0/-0
Re: Timelines and Routes, Thread 2
« Reply #359 on: April 18, 2013, 09:20:55 PM »
To see what two one-syllable words, the second with a long "e," look like, consider:

it's raining, and he's just... At about 22:15
staring at me.  At about 48:00
to check me out. At about 1:22:55
them call me? At about 3:49:40

The long "e" is always clearly visible. Long "e" is a "tense" vowel. Tense vowels are longer in duration and more powerful acoustically that other ("lax")  vowels.

 

Site Meter
click
tracking