QUOTE(nightwing @ May 3 2006, 10:47 PM)

Run it through Audacity, compare that to any number of husky or other dog howls, and the sonograms look essentialy the same.
I can't find any that look essentially the same. Can you send me the file you compared it to? There are many dog howls out there to compare, but most of them are of a much higher pitch. The ones I've found with a lower pitch do not sound or look similar enough for me support your conclusions.
Here's a couple of pics from a recent attempts made in SoundRuler, first Stan's 4-7-2006 sound. They are not really very useful, but they are something to look at:
Click to view attachmentNow a dog from "Sounddogs.com" that had a deeper howling voice:
Click to view attachmentI don't know how to make these images show in the text here.. but looking at them will show you some differences between them. The sounds both appear to be relatively close to the microphone. The dominant frequency of the identified dog is recorded by SoundRuler to be 430Hz, the fundamental 215Hz. A human can make these tones, but you would have a hard time sounding just like a dog when you did it.
Stan's sound has a main frequency of 689Hz and a fundamental of 344Hz, again according to SoundRuler.
SoundRuler is a free application, now several years old (a new version is promised this year) that does have limitations. I think it was made for birders, for instance, and may therefore be based on assumptions that aren't so good for large mammals.
In SoundForge, a completely different program (that costs money) you can look at prettier sonograms, but I can't get captures of them easily. I can also watch the PowerGraph in real time, and see where all the power of the call is going. (It was by doing this that I noticed the strange features of some suspected bigfoot calls I mentioned more than a year ago, but that I have not been able to do anything of merit with since..)
By looking at the sounds in the SoundForge Powergraph window, I can see how the frequencies are interrelating. For instance, Stan's howl gets "dirty" several times, where the voice is no longer producing a clear note, but instead is moving into the growly "noise" range. We know that dogs can do that, but so can any other mammal that has enough voice, including man, and anything that roars. The dog file I have does a little of it at the beginning of its call too. So there is a similarity between them. But they do not match in where they use this modulation of the sound. You might not expect them to, but then we also all know the classic wolf call, and how that modulates. It's a classic because wolves do it, rather than using a completely chaotic repretoir of noises such that we can detect no pattern. This is not a major point, but so far as it goes, Stan's sound isn't matching what other dog calls I can find do in the way that it modulates its call. So it sounds like what a dog can do, but it sounds different from my dog examples. Make sense?
Here's another difference. You can see that both calls are very close to the microphone, as nature sounds go, because the high end information is well preserved above the noise-floor. I limited my view to around 4.2KHz so that I could see the fine details of what was going on in there, and both sounds gave me numbers in the 3.8-3.9KHz range. So insofar as distance from microphone goes, it appears there isn't a difference that makes a distinction. But the dog sound looks absolutely normal, hitting almost all its harmonics on the way up. These are the rough numbers in Hz:
420 648 850 1073 1269 1508 1720 2521 2997 3260 3417 3895
The fundamental frequency is at about 220Hz here, so every number above should be a multiple of 220. The differences are because of measurement errors and the fact we are using software that wasn't designed to do exactly what we're asking it to do. I have to hold the mouse cursor and read the number, write it down, etc. So 850 ought to be 860, or the fundamental was perhaps closer to the 215Hz that SoundRuler gave us (though SoundRuler was giving us the average, and I'm taking a snapshot of the "moving" sound, so we can expect this kind of difference).
Stan's sound is a little different.
662 971 1309 xxxx 1960 2270 2615 xxxx xxxx xxxx 3936
There are gaps, and the numbers don't match the fundamental as closely as the dog's sound did.
These are two important points, I think, to bear in mind. The sound
sounds weird to us for these reasons -- it doesn't conform to what we normally expect. I can explain the fact that the numbers don't match the fundamental, which ought to be around 340Hz according to SoundRuler, because the noise this animal is making is
dirtier than the sound the dog is making. It's much more like a roar than the dog's howl is, even at those points where it sounds relatively clear. What this means is that when I move my cursor looking for the peak, I have a lot of wiggle room. In figurative words, I choose the highest relative peak on a plateau. The frequency looks like a plateau because the animal was introducing distortion into its voice -- kind of like a guitar with a stomp box. Use enough of that distortion and you just get a noisey roar. Use a little, and it sounds cool. I don't know enough about dogs and their barks to say that dogs don't do this sometimes in exactly this way, but I don't have any sound files of them doing it.
Nightwing may have some and if he does I hope he'll share them with me.
About the gaps I am just confused. I can explain the upper gaps away due to distance, although that peak at 3.9KHz was a good one. I expected the other peaks to be there. Maybe if I focused down in on those frequencies, I'd find them, but then we'd still have to explain why they had been so damped when other frequencies hadn't. I checked some of Stan's other files of other animals in the same location, and it doesn't appear that there is any environmental reason for it. And I noted the same kind of gapping in the other files with this same animal doing its call. I don't know what it means, but I am entertaining the idea that it means something.. The gap where the 1.6KHz peak should be is a strange one.
I wish we had more good quality recordings of things we think are bigfoot so I could compare them looking for these features. As it is, the best of what I've been able to get is nowhere near the quality of these two files I'm using in this post. To compare them would be like comparing a picture in your hand with one you had to look at 100yds away through binoculars. You can't expect to get useful results.
I could say more but as usual, this has gone on long enough..
So while I won't come out and say, "Yes, this is it, this is a bigfoot call," I'm not writing these sounds off. I really want someone to give me a file that shows a dog making essentially the same noise, because Stan's sounds completely destroyed that hypothesis I had cooking last year. They don't show any double - fundamental features, except during the moments where echo reflects back into the sound in the file that was recorded furthest away. If this is really a bigfoot, my little theory is dead, or at least, in intensive care. But that's the way this thing works, and I'm ready to bury that theory and shed my tears if that's what comes of it.
OR, I'm ready to get the proof that Stan's sound is a dog.
I'm REALLY ready to get a clear sound file from someone who recorded a sasquatch that they saw making the call. Anyone got one?