320kbps or not?

Hi Rudolfensis,

You are comparing different things here. Your graphs are showing the power per frequency which only depends of the song you are looking at. It only means that the second song have more high pitch sounds than the previous one (would be also true to say that the encoding of the first song may have also drop high pitch frequency during the conversion).

In the other hand, kbps refers to a bitrate, meaning how much bit per second you actually use to encode your songs. More is the better but this is not shown in your graphics. If you want to know the bit rate of any given songs, simply do a get info on it and iTunes will show it to you in the Summary tab.

I don't think there are true criteria for telling if a song has been re-encoded from a different bitrate other than hints. For example, your 128k encoding cuts off at 11kHz, your 320 cuts off at 16kHz. I don't think if the 320k had been encoded from the 128k that it would invent those higher frequencies.

Read up on mp3 encoding. Compression involves certain shortcuts, one of which is getting rid of freqencies at the edges of the normal hearing range. I'm guessing here but maybe the 128k setting chops off at a lower frequency than the 320.

One way to check this is top run a few tests on a CD rip and compare.

Rudolfensis wrote:

I want to know if the 320kbps song is a true 320kbps song, or something that I converted into 320kbps from a lower bitrate — 128, 256 or whatever.

If both are the same songs, you can make the assumption that the second should have a better quality since it will reproduce high frequency information that was missing in the first file.

However, you cannot assess if was encoded from the original of if it is a conversion from a previously encoded 128kbps files. In the first case, your high frequencies signal would have been encoded from existing data while in the second case, it would have been statistically interpolated from an already compressed signal. In both case you will see high frequency signal but the first one would definitively sound better... or closer to the uncompress file. Therefore, looking at the frequency spectrum will not allow you to conclude what was the original signal. In fact, I don't know if there is only other ways that earing the file...

To verify if what I'm saying is true or not (I may be wrong) you can actually make the test of converting your 128kbps file to 320 and look if it ends up adding information above 12kHz.

Take a couple of tracks. Rip them at 320k and 128k, then convert the 128k to 320k. Compare the three graphs for each track. If there is an obvious pattern to tracks that have been upsampled then it should become clear.

tt2

Limnos wrote:

I don't think if the 320k had been encoded from the 128k that it would invent those higher frequencies.

I recall from my course in signal treatment that you can actually use statistical algorithm to interpolate higher frequency that have been discarted. (That what my hypothesis that converting to higher bit rate will do but this would need to be verify.) The result is really not as good as compressing from the original of course but still sounds up a little better.

In any cases, as I was sayin, you will not be able to assess what was the source of a compress files by looking at a spectrum. If you doubt your source, encode again from your CD and listen to the files. Your ears will notify the difference... and if not, who cares then. 😝

Do a RIP at 128 then convert it to 320 and compare it to one of the 320 you are unsure about.

I suggest you simply reRIP the CDs you do have then deal with the others after that.

Is there a reason you are RIPping to MP3 as opposed to AAC?

AAC will retain more and if you need to convert to MP3, you will have less loss than the other way around.

Hi again,

No use to start converting multiple time. Remember what you want to assess; original -> 128kbps vs original -> 128kbps -> 320. Compare the two result even if you are doing the test with any of your CD.

At the end, if you want to decide weither or not it worth the cost of purchasing better quality, try to ear the difference. For exemple, if you only playback on small amplified speaker, you may still not being able to discriminate which is which. In that case, would it worth to pay for it again?

As Chris pointed out, AAC encoding is far better than MP3.

Don't forget the iTunes store is only going to give you 256k ACC, not 320k, so the improvment won't be as good as a re-rip.

tt2

Rudolfensis wrote:

Hi,

Do you mean to convert the 320k song to 128, and then back to 320 and compare?

No.

Take a CD (any CD) and reRIP a song at 128 kbps.

Then convert that new 128 kbps RIP to 320 kbps.

Look at the freq analysis of each of these two files.

Does the 128 kbps file stop at 11khz?

Does the converted to 320 kbps file stop at 16khz? or 11khz?