Okay, DUMB question of the day… and feel free to even laugh at the notion.
I have an OLD project from the SX2 days, that I’d love to do a remix on. The problem is, some of the audio files became corrupt on their storage medium over the years, and now a lot of the crucial vocal and other files are missing, assumed dead.
Is there any way to “reconstruct” an audio file from these .peak files that represent them? Logically it seems possible but I have no idea of the science that would go behind it.
Thanks for any info, and again, feel free to have a bash at me.
Since those are really just tiny images (bitmaps the AI tells me) I doubt they have anything close to the necessary resolution to recreate the audio - even if there was some kind of tool to do it with.
I lost all of my early reel-to-reel tapes during a house move & really the best thing to do is move ahead with whatever is still available. I assume you have no backups from back then. Do you have a mix that is readable? You could use that with SpectralLayers to separate elements.
Aren’t these used to just display the audio waveform? There’s no actual audio data in them? If you delete them, Cubase will just recreate them.
VEGAS uses the same technique (.sfk files), presumably other programs do, also.
They are indeed. But I heard tell that through some miracle of tech one can actually RECONSTRUCT the actual .wav file from this. Although I heard it comes out quite imperfect. But, imperfect would be better than LOST FOREVER.
Might try the whole Spectralayers thing, though I’m skeptical on how it can possibly separate stems out from a Stereo mixdown with as many layers as this has. We’ll see.
is there software that can convert a bitmapped image into an audio waveform
Yes, there are a few software options that can convert a bitmap image into an audio waveform:
Bitmaps & Waves by Victor X: This program converts bitmap images to sounds and vice versa. It uses an inverse Fourier transform to achieve this1. It’s a simple tool available for Windows.
Sound Art by Stone Voices: This free application developed by Vasily Makarov allows you to convert images to sounds. It’s designed for Windows and is part of Stone Voices’ suite of free VST plugins2.
I tried changing the extension on one file from .peak to .bmp but the graphic editor couldn’t open it.
Those programs just interpret the images as if they were a spectrum display, that is, do an inverse FFT, with the x-axis being time, y-axis being the frequency determined by the sample rate, and color being sample value. A waveform image won’t yield the result you expect…
While it should be technically possible to convert the image of a waveform to samples if you interpret x as time and y as sample value with the height of the image being 0dbFS, the result probably won’t be what you expect. You simply don’t have enough values on the x axis to accommodate for a common sample rate like 44.1kHz. For example, take an image that is 1024 pixels wide, you’ll get something like 0.02 seconds of audio. Reduce the sample rate to 1024Hz, then you’ll at least 1 second of audio, but band limited to 500Hz…
(Assuming that the .peak files are actually a bitmap format. Who knows, they could be vector based).
