Fake 24bit files?

Ok so I downloaded these "24 bit" files and compared them to my CD rips. Here it is:



I'm far from a pro but why does it seem there's more data on the 16 bit CD rip than the supposedly 24 bit file? You can see on the graph that the waves go all the way up on the CD rip and not as much on the 24 bit one. Thoughts?

 

I would presume there are a lot fake 24bit files.

 

Hmm, I don't think you will be able to judge the PCM sample's word length (i.e. 16 or 24 bit) simply viewing a track's spectrogram (you can have a pretty good idea of a sampling rate of the original session recordings and/or a final mix, but not the word length, IMO). It looks like your first waveform was likely stemmed from material sampled at (or resampled to) 44kHz.

 

No, you would not be able to determine bit-depth simply with that spectrogram.

All the graph suggests is as testikoff noted, the 24-bit version looks like some kind of filter was applied to attenuate frequencies >20kHz compared to the 16-bit version. Each sample could still be of higher resolution.

(Reminder folks - my 24-bit vs. 16-bit audio test still open for survey submissions! See the blog...)

 

just out of curiosity, can you hear any differences between the two files?

 

The good news is that none of this really makes any difference.

 

My thoughts are that you should learn how to read these graphs and learn a bit about digital audio. This tells you absolutely nothing about bit-depth.

 

Agree with other posters. This analysis can't tell.

 

What are the sizes of the 18bit and 24 bit files. The 24bit version should be a lot bigger

 

I know it can't tell the bit rate, but it looks as the 24bit one might not even be a lossless file lol... Here's what a 128 kbps MP3 looks like in this program... The lower the quality, the lower it gets on the graphic:

 

Interesting

 

I know how a 128 kbps file looks. It still tells you nothing about its bit-depth.

If someone converted an mp3 to 24-bit, and you downloaded it, oh well...

 

The 128kbps mp3 screenshot only tells you that frequencies above 16kHz were discarded, which is a way of losing data to reduce filesize.
Your cd vs. 24 bit file comparison only shows there is slightly less energy in the upper frequencies for the 24 bit one, which could be a mastering thing. If the treble was boosted for the cd, that is what you'd see.
It tells you nothing about bit depth.

 

There is only one possibility: the source material was restricted to around 16,000 cycles, thereby cutting off any frequencies above that point.

 

In general, the lower the bitrate, the lower the setting of the low pass filter frequency, so the lower it looks graphically. This is one way of reducing file size.

But this is not always the case. During encoding, the user could have forced the low pass filter to a higher frequency. This would mean a sacrifice in quality in the lower frequency components for a specific bitrate (which is more critical). The MP3 encoder has to deal with more information, yet compress to the same output file size.

 

If it truly "matters" so much, simply listen to them, and you should easily be able to tell which is which. ( hopefully not knowing which is which)

If you are not able to tell them apart easily, it must not matter all that much.

 

https://docs.google.com/file/d/0B2ojIyyHowC8VmJXUjc2VDdnRUU/edit?usp=docslist_api

What if anything does this graph say in terms of this actually being a 24bit audio file?

 

Nothing. No bit information, just a frequency -> level plot. It might be a DSD file, though, or it might have been originally. That bit on the right looks like modulation noise.

 

I did similar graphs for other digital files. All of my standard 44.1 CDs have information up to 20,000Hz. My 24/96 and 192/24 files show info into the 30,000-50,000 range and some even above that. THEN there are those like the chart below, where there seems to be a space and then more info. As you can see I'm no technician, but I just want to confirm that what I have is indeed 24bit music, and not surreptitiously upconverted from CD quality.

 

The truth is that you can't know unless you get the info straight from the studio or artist. Good hunting!

 

I see you're using Adobe Audition. If you open a 24-bit file it should show at the bottom of the screen in the statusbar. I see you cut off the screenshots so the statusbar isn't showing, could you check what it says in the statusbar?

 

What bit rate the file is identified as means nothing either, necessarily. It's easy to convert 16 bit audio to however many bits you want...just zero all the low bits. Doesn't mean you're getting any more info, of course.

 

But in this case it is even less info, that's what's weird.