Can anybody tell me what this waveform means?
- Thread starter rockne023
- Start date
moses
hardliner
Soundcloud waveforms are just fancy approximations, they look cool on the web, but their technical value is totally useless.
- To save bandwidth (and reduce processing time after upload), only the sum of both LR channels is displayed which can lead to strange visual issues with out of phase signals (they aren't visible at all).
- To save even more bandwidth, only the upper part of the waveform is used, the lower part is just a mirrored version.
- I'm not totally sure, full all waveforms seem to get normalized.
- All waveforms have the same width, to achieve this, soundcloud "stretches" them to fit the full width. This is done without proper resampling, so strange aliasing effects tends to show up in soundcloud (that aren't in the sound file).
It's pointless..
..even in super pro audio editors, the visible waveform is a very rough approximation. No matter how many pixels your screen has, the resolution of audio is million times higher. Same here, most audio editors won't waste cpu cycles to properly resample the waveform to the screen resolution in order to avoid aliasing effects. So, what you see has not much to do with the actual audio data.
Finally, don't make the mistake to assume that the visual representation of pcm data (raw audio waveforms) you see in audio editors is the actual audio waveform that drives the speakers. Here's the reason why, a digital audio system is described as [AD] -> [Digital format] -> [DA]. The true waveform is what comes out of the DA, not the "stepped" digital representation. Digital audio is perfectly smooth.
Beside spotting/fixing very basic problems (problematic DC content, need for trimming, fades, rough dynamic range estimation), "watching" music isn't a good idea at all. The same sound can be represented by an infinite amount of possible waveforms, it makes the whole thing even more pointless...
- To save bandwidth (and reduce processing time after upload), only the sum of both LR channels is displayed which can lead to strange visual issues with out of phase signals (they aren't visible at all).
- To save even more bandwidth, only the upper part of the waveform is used, the lower part is just a mirrored version.
- I'm not totally sure, full all waveforms seem to get normalized.
- All waveforms have the same width, to achieve this, soundcloud "stretches" them to fit the full width. This is done without proper resampling, so strange aliasing effects tends to show up in soundcloud (that aren't in the sound file).
It's pointless..
..even in super pro audio editors, the visible waveform is a very rough approximation. No matter how many pixels your screen has, the resolution of audio is million times higher. Same here, most audio editors won't waste cpu cycles to properly resample the waveform to the screen resolution in order to avoid aliasing effects. So, what you see has not much to do with the actual audio data.
Finally, don't make the mistake to assume that the visual representation of pcm data (raw audio waveforms) you see in audio editors is the actual audio waveform that drives the speakers. Here's the reason why, a digital audio system is described as [AD] -> [Digital format] -> [DA]. The true waveform is what comes out of the DA, not the "stepped" digital representation. Digital audio is perfectly smooth.
Beside spotting/fixing very basic problems (problematic DC content, need for trimming, fades, rough dynamic range estimation), "watching" music isn't a good idea at all. The same sound can be represented by an infinite amount of possible waveforms, it makes the whole thing even more pointless...
krushing
Moderator
Yeah. This isn't exactly hard to figure out even without any technical understanding of it, if you think about it a bit - if you take a vertical slice of a waveform like this, that's maybe 50 pixels high and understand that the sound could be just about anything...how could it possibly convey very much information about such things as "sound quality" or "mix quality"? Sure, you can tell if it's horribly clipping, but besides that, it's pretty obvious that it's a very crude approximation of amplitude and not much more.
LostProfit
New member
You can tell that it is not made by me. I say that because you can see hints of rhythm. Even though the wave might not be accurate to the last increment, what you can tell is something about the rhythm. Your brain can quickly identify patterns and you know almost subconsciously or again, being to approximate what it might sound like.
moses
hardliner
RedBeerd
New member
I use the common waveforms as indicators of breadth of dynamics (not as a measurement, but looking at a waveform, I can tell if one has a good range, generally).
But I Understand about no specific information being gleaned from one.
If one were to take waveforms in the DAWs, and make them an actual, measurable, observable tool, what would we want that tool to do. How could it help us?
PEace.
But I Understand about no specific information being gleaned from one.
If one were to take waveforms in the DAWs, and make them an actual, measurable, observable tool, what would we want that tool to do. How could it help us?
PEace.
Well that's the thing, waveforms don't give you a whole lot of information other than basic info on where the music is louder and softer and possibly the groove.
Why are you trying to glean more insight into this mix by looking at it?