Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Windows Media Player 12 peculiar behaviour, mp3 ripping (Read 4583 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Windows Media Player 12 peculiar behaviour, mp3 ripping

Greetings!

I've noticed that it sometimes, for some albums, or even songs, doesn't even remotely offer the predicted 20.5kHz cutoff reserved for 320kbps files. Not sure whether it matters, but the file is Iron Maiden from the eponymous 1998 Remaster/Expanded CD.

I'm ripping directly from the CD, which isn't at fault (I ripped a test WMA lossless file and it was OK)

Example file, with comparison with iTunes rip (same 320CBR): https://i.imgur.com/yukJAol.jpg

Full screen spectrogram: https://i.imgur.com/S6AjIca.jpg

Thoughts?

Regards.





Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #1 –
Quote
, doesn't even remotely offer the predicted 20.5kHz cutoff reserved for 320kbps files. Not sure whether it matters,
It doesn't matter.  

I'm not sure if 320kbps automatically defaults to preserving those frequencies.    There's a LAME switch/option for that (if you're using LAME), but I don't remember how to do it.  And, you may get overall worse-sounding results if you start changing the settings, even if it gives you a better looking spectrum.

You can't hear the highest-frequencies in the context of other sounds in the background drowning them out.    If you can hear frequencies around 20kHz, you can only hear loud test tones against a quiet background.   

MP3 is lossy and it's going to throw-away some information.    There's no getting around that with lossy compression.     It's trying to throw-away sound that you can't hear anyway and it's usually best to allow it to what it does best.   If it preserves sounds that you can't hear, it might end-up throwing away something you can hear.     We know it's lossy, so try not to worry about what the spectrum LOOKS like...   It's the SOUND that matters.   

Quote
(I ripped a test WMA lossless file and it was OK)
Yes...   Lossless is lossless...  It's bit-identical to the original CD after decompression.


Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #3 –
The MP3 encoder that is used with Windows Media Player is one of the latest versions of frauhoffer encoder (the makers of MP3).
The MP3 encoder in iTunes came originally from another project, and later was made new inhouse (anyone is free to correct me on this, but i think it's correct).
Also, the MP3 encoder in iTunes was not considered to be top quality. I believe it received some tweaks since those days, but the main focus with iTunes has always been AAC.

So, even though iTunes graphics looks nicer, it might actually be the Windows Media Player encoded MP3 the one that sounds better.

As to why would a good encoder do what you see in the Windows Media Player graphic?
Well, MP3 was not thought as a full quality codec, and its design also tried to bridge two worlds that made it more inefficient still (as in trying to keep some structures similar to MP1 and MP2).
One of these design limitations exists in the upper band, the range from 16Khz to 22Khz (on a 44Khz audio).
In this band, there is no specific adjustment for quantization of data. Quantization is used to reduce precision when it is unneeded.

The consequence is that sometimes, specially if there are strong contents above 16Khz, it needs to modify a global quantization control, which in turn means that the whole spectum uses more precision than necessary, and so requiring more bitrate (or reducing quality if not enough bitrate is available).

LAME MP3 encoder has an option to ignore the need for more precision, and stores that upper limit with less precision, so that the rest of the audio maintains quality and doesn't bloat the bitrate.

Frauhoffer encoder seems to have some sort of ducking, which ignores the noise below a certain gain, and is only keeping content that is over that. That would explain why it looks like a filter, but still there is content above it.

Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #4 –
Essentially lossy encoding makes compromises, and throws away as much data as possible without compromising the sound. I trust LAME at V2 and V3 the most, but honestly I have no reason to doubt the quality of WMA. I doubt I could reliably hear the difference.
That's so plausible, I can't believe it.

Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #5 –
Thanks all for your replies, glad to know that it's probably normal. Even if my OCD is making me ponder whether I should re-encode my entire library or not. :-)

I tried listening to both the 320kbps version and the WMA one and I didn't really notice a difference at all (except maybe placebo), even with Sennheiser HD 650s and a dedicated sound card. Here's a spectro comparison: https://i.imgur.com/2EryY1s.jpg the green data seems highly congruent, compared to the blue one, but I believe that's mostly technically unused white noise (?).

I'm guessing I shouldn't take the following anymore at Bible-level value, it seems unadapted to WMP Frauhoffer, compared to Lame. (https://www.whatinterviewprep.com/prepare-for-the-interview/spectral-analysis/)

Quote from: [JAZ] on
The MP3 encoder that is used with Windows Media Player is one of the latest versions of frauhoffer encoder (the makers of MP3).
The MP3 encoder in iTunes came originally from another project, and later was made new inhouse (anyone is free to correct me on this, but i think it's correct).
Also, the MP3 encoder in iTunes was not considered to be top quality. I believe it received some tweaks since those days, but the main focus with iTunes has always been AAC.

So, even though iTunes graphics looks nicer, it might actually be the Windows Media Player encoded MP3 the one that sounds better.

According to EncSpot it's the same underlying Frauhoffer, but iTunes is "fastenc/mp3enc" compared to the "ACM/producer pro" of WMP. Don't know how accurate that is, tho.

The program is very old, though. DbPowerAmp seems to give "FhG (guess)" for both.

https://i.imgur.com/pRvjiif.jpg

Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #6 –
Starting to believe that spectral analysis can actually be misleading.

Here's a 320kbps live song from King Crimson file that looks like a 128kbps one: http://i.imgur.com/MvMFoMl.jpg

Same file ripped with Exact Audio Copy: https://i.imgur.com/K4TRxCM.jpg

Either that or the only thing that matters is the green data. In the original post, notice that iTunes mostly has congruent data with WMP when it comes to the green stuff, however iTunes seems to keep the blue data, while WMP mostly discards it at 16.5kHz.

Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #7 –
This is like the third or fourth topic that is based on the erroneous notion that spectral information is a reliable method of judging sound quality in the past few weeks.  There must be something in the water.

To save additional readers who are either unsuspecting or willfully ignorant: spectral analysis is not a reliable means for judging sound quality.  Those who wish to argue the contrary will have to take it to another forum as this is baked into the very core of this community.  If this sparks any addional interest, please review our Terms of Service.

Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #8 –
Quote from: greynol on
This is like the third or fourth topic that is based on the erroneous notion that spectral information is a reliable method of judging sound quality in the past few weeks.  There must be something in the w

...ikipedia? The diagram here:
https://en.wikipedia.org/wiki/Data_compression#Lossy_audio_compression
Meh.

(Not sure if everyone knows that two different mp3 encoders produce different CBR files at same bitrate. mp3 has matured so much that people may not anymore be aware that encoding (given the bitrate) is not specified by the format.)

Re: Windows Media Player 12 peculiar behaviour, mp3 ripping

Reply #9 –
People have all sorts of misapprehensions about all sorts of things.  I still see people who think that decompressing mp3 to wave gives back the original source without any loss of quality by virtue of the fact that the newly created wave file is bigger than the mp3 source.