Home |
Search |
Today's Posts |
#1
Posted to rec.audio.tech
|
|||
|
|||
What the hell is this?
How do I turn http://i38.tinypic.com/33xu8p1.png into http://i34.tinypic.com/2zftt13.png
? And what did the perpetrator do to the waveform? It sounds exactly the same as the proceeding length. But If I amplify, then some parts of the weird-looking 3-second waveform in the beginning are inaudible. |
#2
Posted to rec.audio.tech
|
|||
|
|||
What the hell is this?
On Aug 8, 12:18 pm, Industrial One wrote:
How do I turnhttp://i38.tinypic.com/33xu8p1.pngintohttp://i34.tinypic.com/2zftt13.png And what did the perpetrator do to the waveform? It sounds exactly the same as the proceeding length. But If I amplify, then some parts of the weird-looking 3-second waveform in the beginning are inaudible. The first example clearly showws the first segment has a significant DC offset. The waveform also exhibits some slipped samples The second example shows that DC offset removed, and the clipped samples remain. To turn the first into the second, using any of the common editors about (e.g., Audacity, select that portion of the wave, and (usually) under the "Normalize" menu function you'll find an option to "Remove DC offsett." The more interesting question, if you're interested, is how did it get thatw ay to begin with. The most likely scenario is someone may have started with a 3-second audio file recorded in offset binary format (the default for 8-bit linear PCM wave files) and then appended a two'-s complement segment to that. An offset binary file takes has no "sign": It takes a normal signed two's complement representation (that, say, in 8-bit, goes from -128 to +127), adds 128 to it (so the result is no 0 to 255). It's possible that whatever editor was used to append these segment failed to renormalize the offset binary segment properly. |
#3
Posted to rec.audio.tech
|
|||
|
|||
What the hell is this?
On Aug 8, 11:04 am, wrote:
On Aug 8, 12:18 pm, Industrial One wrote: How do I turnhttp://i38.tinypic.com/33xu8p1.pngintohttp://i34.tinypic.com/2zftt13.png And what did the perpetrator do to the waveform? It sounds exactly the same as the proceeding length. But If I amplify, then some parts of the weird-looking 3-second waveform in the beginning are inaudible. The first example clearly showws the first segment has a significant DC offset. The waveform also exhibits some slipped samples Clipped, you mean? The second example shows that DC offset removed, and the clipped samples remain. To turn the first into the second, using any of the common editors about (e.g., Audacity, select that portion of the wave, and (usually) under the "Normalize" menu function you'll find an option to "Remove DC offsett." Kickass, this fixed it. Thanks. The more interesting question, if you're interested, is how did it get thatw ay to begin with. The most likely scenario is someone may have started with a 3-second audio file recorded in offset binary format (the default for 8-bit linear PCM wave files) and then appended a two'-s complement segment to that. An offset binary file takes has no "sign": It takes a normal signed two's complement representation (that, say, in 8-bit, goes from -128 to +127), adds 128 to it (so the result is no 0 to 255). It's possible that whatever editor was used to append these segment failed to renormalize the offset binary segment properly. You're saying that that 3-second sample is 8-bit while the rest of the audio is 16-bit? |
#4
Posted to rec.audio.tech
|
|||
|
|||
What the hell is this?
On Aug 8, 1:39 pm, Industrial One wrote:
On Aug 8, 11:04 am, wrote: It's possible that whatever editor was used to append these segment failed to renormalize the offset binary segment properly. You're saying that that 3-second sample is 8-bit while the rest of the audio is 16-bit? No, I'm saying that's one plausible way it could have gotten that way. |
#5
Posted to rec.audio.tech
|
|||
|
|||
What the hell is this?
On Aug 8, 12:00 pm, wrote:
On Aug 8, 1:39 pm, Industrial One wrote: On Aug 8, 11:04 am, wrote: It's possible that whatever editor was used to append these segment failed to renormalize the offset binary segment properly. You're saying that that 3-second sample is 8-bit while the rest of the audio is 16-bit? No, I'm saying that's one plausible way it could have gotten that way. I think it was done on purpose, 'cuz that 3-second part is where the DJ's name was whispered, so he wanted to make it harder to remove -- or something. Problem is, that part has a really high amplitude which forces the rest of the audio to be really low (btw, how can it have a higher amplitude yet have the same amount of decibels as the rest of the waveform?) |
#6
Posted to rec.audio.tech
|
|||
|
|||
What the hell is this?
On Aug 8, 3:12 pm, Industrial One wrote:
I think it was done on purpose, 'cuz that 3-second part is where the DJ's name was whispered, so he wanted to make it harder to remove -- or something. Problem is, that part has a really high amplitude which forces the rest of the audio to be really low (btw, how can it have a higher amplitude yet have the same amount of decibels as the rest of the waveform?) Because if the thing that measures decibels essentially has a high-ass filter function to it, the DC offset is irrelevant. |
#7
Posted to rec.audio.tech
|
|||
|
|||
What the hell is this?
"Industrial One" wrote ...
I think it was done on purpose, 'cuz that 3-second part is where the DJ's name was whispered, so he wanted to make it harder to remove -- Seems much more likely that somebody tried to boost the signal level (to hear the faint audio), but didn't do a very good job of it. |
Reply |
Thread Tools | |
Display Modes | |
|
|
Similar Threads | ||||
Thread | Forum | |||
Umm, what the hell | Audio Opinions | |||
What the hell is this? | Pro Audio | |||
What The Hell is an iHP-120 | Pro Audio | |||
what the hell is this? | Car Audio |