Are larger files supposed to sound better?

[Aldbaran]

I've always heard MP3 files don't sound as good as other file types because they compress the data to save space and sacrifice quality. Well, I found a song I like and decided to do a test to see if it would be worth it to use larger WAV files for my most favorite songs. The MP3 version is around 5MB, and then WAV version is 33MB. I played the song through my stereo system and listened intently to see if there was a difference in quality between the two. I honestly couldn't discern any difference at all. Is there some other way to increase the quality? Perhaps a file type other than WAV?

[poisondeathray]

I've always heard MP3 files don't sound as good as other file types because they compress the data to save space and sacrifice quality. Well, I found a song I like and decided to do a test to see if it would be worth it to use larger WAV files for my most favorite songs. The MP3 version is around 5MB, and then WAV version is 33MB. I played the song through my stereo system and listened intently to see if there was a difference in quality between the two. I honestly couldn't discern any difference at all. Is there some other way to increase the quality? Perhaps a file type other than WAV?

Your comparison is flawed because you're looking at different generations. You need to start with the original source (before the mp3)

Consider :

Original source => mp3

vs

Original source => wav

vs

Original source => mp3 => wav


If you take the mp3 an convert it to wav, you're just "uncompressing" or decoding the mp3. They will and should sound the same. You're just increasing the filesize for the same quality

If you take the original source, and convert it to wav, it will sound like the original source. Which is better than your mp3 which uses lossy compression and throws out bits to achieve that compression . You can never get back the bits that were thrown away, that's why it's called "lossy" compression

[Aldbaran]

I've always heard MP3 files don't sound as good as other file types because they compress the data to save space and sacrifice quality. Well, I found a song I like and decided to do a test to see if it would be worth it to use larger WAV files for my most favorite songs. The MP3 version is around 5MB, and then WAV version is 33MB. I played the song through my stereo system and listened intently to see if there was a difference in quality between the two. I honestly couldn't discern any difference at all. Is there some other way to increase the quality? Perhaps a file type other than WAV?

Your comparison is flawed because you're looking at different generations. You need to start with the original source (before the mp3)

Consider :

Original source => mp3

vs

Original source => wav

vs

Original source => mp3 => wav


If you take the mp3 an convert it to wav, you're just "uncompressing" or decoding the mp3. They will and should sound the same. You're just increasing the filesize for the same quality

If you take the original source, and convert it to wav, it will sound like the original source. Which is better than your mp3 which uses lossy compression and throws out bits to achieve that compression . You can never get back the bits that were thrown away, that's why it's called "lossy" compression

I guess I wasn't clear that when I did the comparison, I made it both MP3 and then WAV (separately) from the original source. There's a site that allows files to be saved from Youtube and saved as whatever file type you choose. I found a video of a song I liked and saved it as an MP3, and then again as a WAV. Here's the site I used: https://www.onlinevideoconverter.com/mp3-converter

[redwudz]

There are some 'lossless' audio formats such as flac you might try that will be smaller than WAV audio.
Just make sure your audio equipment can play it back.
You can also likely change the encoding settings for MP3s and improve quality.

[poisondeathray]

If you couldn't hear the difference, then there is no reason to increase the quality of the mp3. But you could use higher bitrates to increase the quality of the mp3 encode if you wanted to

Either your ears, or equipment are not discerning enough, or that specific source did not have certain frequencies or tonality range to begin with - and that's ok if its for your ears.

That link is for a youtube converter. So I'm assuming your "source" was a youtube video, or some other site like vimeo ? If so, that imples the source already was quite lossy, because all the sites on that list re-encode the audio, usually with AAC .

Just be aware , when you use higher quality sources, you need to use better mp3 settings , or higher quality compression than mp3, or higher bitrates before "you can't tell the difference" . ie. you're starting with a low quality youtube source, where many frequencies have already been discarded - so it's "easier" to reproduce or get similar results at lower bitrates. It's already been "culled", so to speak. If instead you started with a high quality source with full range of frequencies, it requires more bits to achieve a similar level of quality where a listener "can't tell the difference" .

[Aldbaran]

If you couldn't hear the difference, then there is no reason to increase the quality of the mp3. But you could use higher bitrates to increase the quality of the mp3 encode if you wanted to

Either your ears, or equipment are not discerning enough, or that specific source did not have certain frequencies or tonality range to begin with - and that's ok if its for your ears.

That link is for a youtube converter. So I'm assuming your "source" was a youtube video, or some other site like vimeo ? If so, that imples the source already was quite lossy, because all the sites on that list re-encode the audio, usually with AAC .

Just be aware , when you use higher quality sources, you need to use better mp3 settings , or higher quality compression than mp3, or higher bitrates before "you can't tell the difference" . ie. you're starting with a low quality youtube source, where many frequencies have already been discarded - so it's "easier" to reproduce or get similar results at lower bitrates. It's already been "culled", so to speak. If instead you started with a high quality source with full range of frequencies, it requires more bits to achieve a similar level of quality where a listener "can't tell the difference" .

You're probably correct about Youtube being a lower quality source to begin with. That would explain why saving something from it as a WAV file instead of MP3 didn't make any difference. I'm new to all this.

[jagabo]

It also depends on what bitrate you use for the MP3 encoding, which MP3 encoder you use, and even the properties of the particular audio. And for subtle differences (at higher bitrates) you have to know what to listen for.

[aedipuss]

LOL all mp3 and cd format wav suck as far as audio quality goes.

[JVRaines]

I guess that settles it.

[hello_hello]

LOL all mp3 and cd format wav suck as far as audio quality goes.

CD format wave also sucks as far as audio quality goes??? What's the reasoning behind that claim?

[shans]

I've always heard MP3 files don't sound as good as other file types because they compress the data to save space and sacrifice quality.

Your comparison is flawed because you're looking at different generations. You need to start with the original source (before the mp3)

Consider :

Original source => mp3

vs

Original source => wav

vs

Original source => mp3 => wav


If you take the mp3 an convert it to wav, you're just "uncompressing" or decoding the mp3. They will and should sound the same. You're just increasing the filesize for the same quality

If you take the original source, and convert it to wav, it will sound like the original source. Which is better than your mp3 which uses lossy compression and throws out bits to achieve that compression . You can never get back the bits that were thrown away, that's why it's called "lossy" compression

Hitherto, I used to extract audio files like AAC, AC3, MP3 etc. from movie videos to WAV / WAVS (in case of 5.1 channels) by using different programs, load them into Audacity with FFmpeg plugin, do normalizing and / or compression, save them again to wav / wavs, re-encode them to the desired audio format enhancing the bit rate and mux them with video with different / same containers. I did the above exercise using various programs depending upon the format. I used to do the above invariably for all movies to enhance the dialogue sound quality. I was under the impression that I had not deteriorated the audio quality by following the above procedure. After seeing this post, I realized that I have been following the unwanted steps.

Hereafter, I will straightaway load the video in audacity which will, however, decompress them as wave/waves, do the normalize/compress function and save them to the desired audio format so that I can avoid the extra steps. Please correct me if I am wrong anywhere. Thanks for detailed explanation.

[Ennio]

If I understand your latest post correctly, I doubt if you will gain any "quality". It seems to me that you are editing audiofiles that are already encoded in a lossy compressor in both cases. Whether you first extract and unpack them to pcm, or directly load the compressed files into Audacity does not make a difference for the program, I guess. I strongly believe that whatever editing you do, Audacity will do this in uncompressed pcm domain anyways. To anyone, please correct if if I'm wrong here.
Therefore, after import of a lossy compressed audiofile, I think the first thing Audacity (and I think most editors) will do is decompressing it to pcm and all editing steps will be in the pcm wave. So what would be the benfit then if you would extract the audio to uncompressed pcm yourself?

Furthermore, after editing, of course when you re-encode your editing result into a lossy compressor you'll have the dis-benefit of losing actual data (read: quality) for the second time.

Please inform me if I misunderstood something in your latest post.

Maybe, in many cases you (your ears) wouldn't be able tell the difference in the end. My experience with mainly music is that not every file "reacts" the same when going from lossless to lossy.
Each time you'll have to listen and compare if the quality / space trade-off is worth it. Maybe life is too short for this...

[shans]

If I understand your latest post correctly, I doubt if you will gain any "quality". It seems to me that you are editing audiofiles that are already encoded in a lossy compressor in both cases. Whether you first extract and unpack them to pcm, or directly load the compressed files into Audacity does not make a difference for the program, I guess. I strongly believe that whatever editing you do, Audacity will do this in uncompressed pcm domain anyways.
...

I fully agree with you that I lose quality in re-encoding the audio file after editing. Earlier, I was doing it by using various programs and now after reading the post #2 will do the same only by using audacity. My aim is to enhance the sound quality of speech which is very less at some and louder at others. So, I want to enhance it by normalizing/compress at the cost of losing quality.

[pandy]

I've always heard MP3 files don't sound as good as other file types because they compress the data to save space and sacrifice quality. Well, I found a song I like and decided to do a test to see if it would be worth it to use larger WAV files for my most favorite songs. The MP3 version is around 5MB, and then WAV version is 33MB. I played the song through my stereo system and listened intently to see if there was a difference in quality between the two. I honestly couldn't discern any difference at all. Is there some other way to increase the quality? Perhaps a file type other than WAV?

Well, look at your statement - you always heard but you don't perceive any difference so what do you think about this?
MP3 is good for most, average consumers - perhaps 90% or more humanity will not perceive MP3 as "don't sound as good as other file types" - quality especially perceived one is outcome of many factors and most of us will not perceive any difference between MP3 and WAV unless guided by someone very experience with special recordings (to expose some coding limitations) in very good listening conditions.

By WAV i mean lossless source used for MP3 creation - your example as stated already in this topic is just MP3 decoder test (assuming that different decoders are used by you in real time listening and offline decoding).

[poisondeathray]

Hereafter, I will straightaway load the video in audacity which will, however, decompress them as wave/waves, do the normalize/compress function and save them to the desired audio format so that I can avoid the extra steps. Please correct me if I am wrong anywhere. Thanks for detailed explanation.

In theory all decoders should output the same data.

So the only negligible difference is what algorithm is used for for dithering during bit depth conversions . Most audio editors work in 32bit float internally. In audacity , preferences=> quality => dither . If you decompressed audio in another audio editor, and it used a different dithering algorithm, then you can get different results. You will not be able to "hear" the difference, and it will be impossible to detect those types of differences with human ears

[shans]

In audacity , preferences=> quality => dither . If you decompressed audio in another audio editor, and it used a different dithering algorithm, then you can get different results. You will not be able to "hear" the difference, and it will be impossible to detect those types of differences with human ears

Thanks. In audacity, preferences=> quality => Dither: There are four options, viz. None, Rectangle, Triangle, Shaped. Which one do you suggest? The default is 'None" There is another parameter listed under 'High-quality conversion' Dither which also has the same four options. The default is found to be 'Shaped'. I have not changed anything.

[poisondeathray]

I would leave it default, but you can test them and see if you can "hear" the difference. You won't be able to at normal bitrate ranges (for very low bitrates with lossy compressed audio you might). But dithering differences are detectable with various programs

[Cornucopia]

Triangular is recommended for intermediate stages, Shaped is for final (not to be re-edited) output.

Scott

[hech54]

Beauty is in the ear of the beholder.

[Hoser Rob]

Dither is very good if you're doing something like converting 24/192 WAV/FLAC to 16/48 lossless. For converting to .mp3 I'm not so sure it's worth the bother.

Is mp3 worse than FLAC. Technically, yes, of course, demonstrably, no argument.

In practice it depends on the original recording quality.

Most CDs and virtually major label pop for the last 15 years or so are compressed to the point of absurdity. Search "loudness wars". It doesn't make any difference that mp3 loses low level information. There's isn't any low level info there. Fortunately with the advent of streaming they're stopping this. Streaming sites almost always have normalization turned on at the site bby default now. There's no point in all that compression now. Thank God.

Also, and this is one of the big dirty secrets in audio, most recording studios don't use high quality microphones. They're expensive and too delicate, and one of the first priorities for pro audio gear is that it won;t break when you drop it. These small basement studios so common now almost never have high quality mics. A couple of those would cost more than all their other studio gear put together.

I keep my ripped audio files in FLAC format unless they sound bad, which is a lot of them. Then I'll happily convert them to mp3. I use constant bitrate at 320K/sec usually.

Can I tell the difference between the FLAC and a 320k mp3? On a good recording, you bet. It's easy to test. Just convert it to mp3, load the original and the mp3 into my music player consecutivelly, and switch between them. Hit next enough and you'll forget which file is playing and the levels will be the same, without which any comparisopns are meaningless. But the difference is obvious.

And the difference between 320K and 128K mp3 is quite noticeable unless the recording is REALLY bad. Like 70s reggae from the Lee Perry studios ... the worst recorded great music I know of.