i have two different mp3s from different sources, both 32bit, 256kbps. one is 441000hz and the other 48000hz.
What i’m wondering is which should sound better? I don’t know enough about audacity to know if you can tell by looking at the wavelengths… the top file (48000) shows a smaller wave girth, while the 44100 takes up way more space. what does this mean?
using sennheiser eh350s i surmised the following unprofessional observations
44100: ‘sounded louder’, cant put my finger on it
48000: ‘softer’, possibly less distortion, easier to listen to at max volume.
If you applied Audacity’s “amplify” effect (on default settings) individually to each track it will make them both the same “girth”, (and as loud as possible).
There may be some clipping distortion on your 41000Hz sample: parts of the waveform may have gone beyond +/- 1.0,
Using “amplify” (on default settings) may correct clipping distortion by reducing the amplitude so no part of the waveform is beyond +/- 1.0
If the amplitude goes beyond +/- 1.0 it will clip (flatten) the peaks and troughs of the waveform …
Altering the shape of the waveform = distortion.
Again if you apply “amplify” (default settings) to both tracks they will both have the same amplitude, (which will not exceed +/- 1.0),
then you will then be able to make a fair comparison between the two: they will both be equally loud.
If you avoid any clipping above the 0db levels and you are within 32-bit floating mode, then the amplification performed with 32-floating point is very, very precise. Unlike older analog tape technology (which was inherently non-linear at amplification due to the limitations and complexities of electronic gain circuitry) the new form of ‘pure digital’ amplification in Audacity has absolutely no downside and will not alter ‘quality’ of any wave files.
I have listened to many of my own recent recordings made with Audacity using very high samples rates (ASUS Essence ST soundcard) recording analog LPs at 192kHz/24 bit from a SOTA Star Sapphire turnatble through Audio Research tube preamp electronics, and I have always found the higher sampling rates provide “less leading edge” effect and therefore always sound somewhat softer (more LP-like) than lower sampling rate recordings, and the higher sample rate recordings also seem to provide the preception of much more detail, depth, and “space” around recorded instruments. The better the recording the more you are able notice this. Cymbals and other high frequncy instruments also show this off very well, as on lower rates (44.1 kHz) sound more like bursts of compressed air, somewhat mechanical sounding, versus a more vibrant and ‘live-scintillating’ sound from high sampling rates (192 kHz) that an analog recording (with infinite resolution) is better at capturing. Its the horrendous noise on LPs that prevents a pleasureable listening exprerience compared to CDs. New software like ‘Click-Repair’ has changed all that for the better and can clean up LPs to a very signifcant degree.
Regarding your two files from different sources, since they are from different sources there may be something else going on here (other than just sample-rates). Technically, the 48 kHz file should have a higher kbps (kilo-bits-per-second) data transfer rate than the 44.1 kHz, because by definition it must contain more data-samples to play in the same length of time…therefore the fact that both file ‘kbps’ are identical does not look quite right.
heres another case but more extreme. I think this example might have more to do with using different encoders? strangely they both have the same 320 kbps and both 18mb in size. I got them from different ends of the internet so i’m not positive.
I’ve uploaded two examples here if you’re interested in hearing the differences, I’d like to know what you think too. http://www.fileserve.com/file/VKQCxDp
You can import both to audacity, then click mute on one, then play it. then click the solo button on the one you’ve muted to switch back and forth between the two. different than the one below, but I picked a nicer song, same differences in the frequency.
The upper version of “super8” is normalized to -5dB, the lower looks like it’s normalized to 0dB, (could be clipping).
Again if you apply “amplify” (default settings) to both tracks they will both have the same amplitude, (which will not exceed +/- 1.0),
then you will then be able to make a fair comparison between the two: they will both be equally loud.
