Mp3 to WAV Converting

T.C. Elliott · May 07, 2012, 08:55:18 PM

Good luck. I had sound issues as well but once they were sorted they've continued to be sorted.

henwrench · August 16, 2012, 05:19:17 AM

So let me get this straight....by converting from mp3 to WAV, we're UNcompressing the file? Or simply turning an mp3 into a WAV. The two processes are completely different, surely?

Out of interest, I've followed the guidance here to carry out an experiment, and, well f*ck my old boots, it DOES appear to work. The converted file (mp3 - WAV) is certainly less restricted, more open sounding and the very high end shimmers as it should without the slight 'under water' warble that low-res mp3s can suffer from. This is pretty incredible. So I can download mp3s and convert them to 'full bandwidth' sound. Remarkable. Quite simply remarkable.
I am an Audiophile and my name is...

henwrench

Geir · August 16, 2012, 05:28:12 AM

I can't believe what you say Henny, tho I do believe you when you say it !!

Well there must be some explanation. One is that the (lame) decoder Audacity uses is better than the one you used to playback the mp3. ..... well that's about the only explanation I can think of !

henwrench · August 16, 2012, 05:52:57 AM

Quote from: Geir on August 16, 2012, 05:28:12 AMI can't believe what you say Henny, tho I do believe you when you say it !!

Well there must be some explanation. One is that the (lame) decoder Audacity uses is better than the one you used to playback the mp3. ..... well that's about the only explanation I can think of !

I'm using the Windows Media Player to A-B the two files, also ran some playbacks on the MBR itself. The mp3 file size is an average 4mb. The converted file is now coming in at 30mb. The resulting differences are incredibly subtle but undeniably audible. Like I say, strictly for SERIOUS audiophiliacs only.

I would be interested to know if anyone else wants to try this conversion and if an audible improvement can be detected...

henwrench

Speed Demon · August 16, 2012, 06:40:09 AM

I have frequency analyzers (software) that may be able to shed some light on this subject.
I haven't done anything with them yet so there will be a learning curve for me but I'll see what I can see.

henwrench · August 16, 2012, 06:52:12 AM

The bass response is slightly fuller and rounder and as I said, the very, very high end sounds sweeter and more defined. The stereo spread also sounds slightly wider. I must admit, I am sceptical of my own 'ears only' findings, only because the process is 'impossible'. The results are also more pronounced in the WMP as opposed to the MBR.

earwrench

Geir · August 16, 2012, 07:44:29 AM

Quote from: henwrench on August 16, 2012, 06:52:12 AMearwrench

Speed Demon · August 16, 2012, 08:05:30 AM

The highs always tell the story.
Without using bi-amping on the playback system, the low frequencies gobble up the highs, in their eternal quest for dominance of the sound spectrum.

64Guitars · August 16, 2012, 08:06:34 AM

Quote from: henwrench on August 16, 2012, 05:19:17 AMSo let me get this straight....by converting from mp3 to WAV, we're UNcompressing the file? Or simply turning an mp3 into a WAV. The two processes are completely different, surely?

Not sure if you realize it, but data compression and audio compression are two completely different things. As you know, audio compression limits dynamic range. But all that data compression is supposed to do (ideally) is to make the file smaller without affecting the sound at all. When the file is played back, software decompresses the data on the fly so that you get the same bits you had before the data was compressed. That's the ideal, and with lossless compression formats such as FLAC, that's what happens. But FLAC typically only reduces a file's size by about half. To get much smaller file sizes, lossy compression schemes like MP3 make a compromise. When compressing the data, they throw away bits that the compression software deems will not have a very noticeable effect on the sound quality. This discarding of data is combined with lossless data compression techniques to achieve files that are about 1/10th the original size. However, when the file is later decompressed for playback, it is impossible to get back the discarded data. This does have an effect on playback quality, which is why MP3s often sound crappy. When compressing files to MP3, several parameters can be adjusted which affect both the file size and the sound quality. For example, higher bitrates give better sound but the files aren't as small as they could be. Lower bitrates produce smaller files but the sound quality sucks.

So, to answer your question, converting a file to WAV format does decompress the file. However, this is only data decompression, not audio decompression (expanding). The bitstream that is stored in the WAV file should be identical to the bitstream that's produced when you merely play back the MP3. If it's not, then it's because the MP3 decoder you're using for file conversion (LAME) is different than the decoder in your media player (Windows Media Player, I assume) and one of them isn't doing its job properly (my guess is that WMP is the culprit). Try loading both the MP3 and the WAV file into Audacity and doing your A/B comparison there. Since LAME is used both for the playback of the MP3 and the conversion to WAV format, they should sound identical. If you're satisfied that they are the same and you suspect that WMP isn't playing your MP3s correctly, try the VLC media player. It's free and plays every media format under the sun.

By the way, you may be wondering how lossless data compression works. It seems incredible that you can reduce a file's size by half, yet get all of the original bits back on playback. Well, it's not magic. It's just substitution. There are many methods used in lossless compression, but as a trivial example, consider a file that contains a section of silence which is represented by 100 zero bytes. With WAV format, that silence would occupy 100 bytes in the file. But with FLAC, for example, the compressor might store a special identifier byte (let's assume it's 255) followed by a single zero, then the count (100). So it only has to store 3 bytes (255, 0, 100) instead of 100 bytes. When the file is played back, the decoder sees the special identifier (255) and knows that the next two bytes are the value and the count. So it inserts 100 zero bytes into the playback stream, making it identical to the original data stream.

Data archiving formats such as Zip and RAR use these same techniques. In this case it must be lossless. If you used lossy compression on an executable program file (throwing away some data), the program would not run after decompression because critical bytes would be missing. With lossless compression, it will run because the decompressed file is bit-for-bit identical to the original file.