[jonson3459]

Hello Audio Circle Members~
What a pleasure to meet you all! I am an avid music lover and an audio enthusiast. I also write for a audio/music magazine called PMA. PMA magazine is conducting a study to see if we can hear which is better between an mp3 and a 24/96 file in a side-by-side comparison. So far, the results are surprising. Right now the sample size is still small so, if we all participate, we’ll know the truth. It’s also a fun test. Please go to the page below to participate:

https://www.pmamedia.org/en/home/can-you-hear-the-difference

[DannyBadorine]

I have submitted my results.  I'm confident with the Vivaldi and Patrician Barber tracks but the Sabbath song was more difficult to tell than I expected.

[Norman Tracy]

Sees post title including "Blind Test..." from new member with 3 posts.

Immediate reaction "oh no not again, flame war incoming, prepare for AudioCircle signal to noise ratio reduction". 

[mcmusicman]

File too large for AV scan with Kaspersky.

[DannyBadorine]

Sees post title including "Blind Test..." from new member with 3 posts.

Immediate reaction "oh no not again, flame war incoming, prepare for AudioCircle signal to noise ratio reduction". 

Hahahahaha!! C'mon Norm.  Take the test!! It was much more difficult than I thought it would be and I still don't know the actual results.

[Rusty Jefferson]

Red Herring. It doesn't say where the 24/96 tracks originated. Upsampling 16/44 to 24/96 is not Hi-Res. If they started with 24/96 original files recorded digitally and made an mp3 copy of that file, it's quite easy for even untrained listeners to hear the difference. 

[I.Greyhound Fan]

No point in me taking the test.  In my stereo I can easily hear the difference between an MP3 recording and the same recording that was recorded in Hi-Rez.  MP3 is flat, dull and lifeless although some MP3 recordings are better than others.

[DannyBadorine]

Red Herring. It doesn't say where the 24/96 tracks originated. Upsampling 16/44 to 24/96 is not Hi-Res. If they started with 24/96 original files recorded digitally and made an mp3 copy of that file, it's quite easy for even untrained listeners to hear the difference.

I totally agree.  I took the test.  The Sabbath songs sound pretty much identical to me.  I'm not home to hear it on my hifi system but I listened on expensive in-ear monitors that should tell me the difference.  The last two tracks showed a little bit of difference, but normally when I compare MP3's with 96k/24bit it's pretty easy to hear the difference.  The differences weren't very clear so I'm wondering how they ended up at 96k/24 bit.  The Sabbath stuff was definitely recorded to tape originally so did they take the tape to 96k/24 bit?  Unlikely. Which means they upsampled the CD 44.1k/16 bit to 96k/24 bit.  That is far from being the same. 

[FullRangeMan]

Currently Iam working with outdoor MP3 files on HD, it barely works as a media, Its not possible to understand the words when there are vehicles passing by on the road. Definitely its a depressing test, linear PCM 24/44 its much better than MP3.

[Onadifferentnote]

I have a friend in the California Guitar Trio, in a recent album, he gave me four different files of the same song.  16/44, 24/96, 24/192, and 32/192.  He wondered if I could the difference in my system. My daughter played them in random order, 3 different times. 

Each time I chose the better file. Why?  My system has a high level of resolution and I have a trained ear.  If recorded correctly, I always prefer the higher resolution files which tends to have more air around the notes, more natural timbre, etc.
The notes just ring longer, more naturally.

When I hear people can't hear the difference, it does surprise me.
Cheers!
 

[Norman Tracy]

Hahahahaha!! C'mon Norm.  Take the test!! It was much more difficult than I thought it would be and I still don't know the actual results.

OK, I downloaded the files and listened. With clenched teeth in the case of the Black Sabbath. Not my cup of tea and they had to pick the most overplayed FM Rock song in the BS catalog? At least Ozzie could still form English words when it was recorded. I listened on Win10 workstation's sound card via Fostex T-50RP headphones.

I listened A/B'ing both the full minute of each selection and also little sections that contained interesting sonic details. It’s been a few years since I knowingly listened to MP3s and it is impressive how close they are. After getting reacquainted with the sound I zeroed in on the fine detail of instrumental texture and the nature of sibilance and notes’ trailing off sounds. For each of the three I wrote down which I thought was the MP3 based on a slight coarsening effect heard in the three areas mentioned. I was 100% correct in identifying high-res vs. MP3.

I know my identifications were correct because after I wrote down my listening results I loaded all six files into Adobe Audition. Using Audition’s Frequency Analysis display set to Linear View 0 – 48k Hz it’s immediately apparent the MP3 by the brick wall filter cutting off everything above 21k Hz and the occasional complete drop outs between 15k to 20k when the MP3 algorithm decides that band can be discarded for awhile to save bits. While in Audition I also noted the Black Sabbath cut was taken from CD as both tracks frequency band cutoff above ~22k Hz. The ‘hi-res’ BS rolled off slightly more gently vs. the MP3. The Patricia Barber and Vivaldi were both sourced from true 96k or greater sources. Those two high-res tracks have the characteristic mostly noise with occasional (or nearly constant in the case of the Vivaldi) musically correlated signal in the 20k to 48k band. And in the case of the Barber a constant noise peak 22.7k Hz at -72 dB. I think every studio has a video monitor or SMPS buzzing along at ultrasonic frequencies. Well, not all, the Vivaldi was just music up there. It is interesting how much energy from musical overtones there are in string instruments in the 20k to 40k Hz band.

I also checked the levels of the A vs B samples and they did get those matched to within a few tenths of a dB for each of the three pairs.

[DannyBadorine]

That is excellent!

[Letitroll98]

Excellent post Norman, most helpful work that I've seen in a while on these pages, and we've got some great posters here.  Because no one's losing any money here I'll assume the op is sincere.  Is it possible that the differing recording chains might be on purpose?   Wouldn't one learn more by comparing various lower res chains upsampled to 96k as well as native 96k files to MP3?  And everyone seems to have a universal reaction that the files Norman id'd as low res weren't much different than MP3, and the native hi res files were easy to pick out.  And that tracks well with non blind comparisons, the issue now appears settled.

[RDavidson]

No point in me taking the test.  In my stereo I can easily hear the difference between an MP3 recording and the same recording that was recorded in Hi-Rez.  MP3 is flat, dull and lifeless although some MP3 recordings are better than others.

Yep. Nail on the head. Sometimes the difference is obvious. Sometimes it is less so.

At the end of the day, who the hell even cares? We have the ability to stream and download hi res much more easily than we could 20 years ago. It cannot be argued that hi res isn't technically better than mp3, so why even debate? This is a waste of progress. Sorry, but I hate mp3's (well...all lossy formats in general) and just want them to go die finally. They had their place and purpose and have run their course. I want better and I want everyone, whether they know it or not, to also listen to better.

[FullRangeMan]

Well, one of the problem with MP3 low sound quality is that the weaker sounds are removed by the compression algorithm, while with linear PCM this does not happen even with low res PCM.

This makes MP3 unsuited to hi-fi only to cell phones audio, these people will not even notice the difference because they have never heard a tube amp, DSD or 24/192.

[Tom Bombadil]

I recently read a post from an owner of a PS Audio Directstream DAC, which upsamples all audio formats into DSD256.  He hadn't listen to MP3 for a while but wanted to compare after getting a software upgrade.  He wrote that he was very surprised by how good the Spotify 320K MP3 track sounded.   Upon close listening in his system he could hear differences but he expected the differences to be immediate and very obvious. 

The Directstream DAC was once the Stereophile Product of the Year and was listed on their A+ Recommended list.  It is now better than it was at that time due to software upgrades. 

[Kw6]

I tried Tidal,  Deezer, Amazon HD and decided to stick with Spotify  for it's  ease of use  and their large international  music content!