albums that sound really good or really bad as mp3s?

Discussion on new albums, developing listening skills, critical listening to others' work, as well as TOMB members' MP3 links, online recording critiques

Moderator: cgarges

percussion boy
carpal tunnel
Posts: 1512
Joined: Wed May 07, 2003 5:51 pm
Location: Bay Area

albums that sound really good or really bad as mp3s?

Post by percussion boy » Thu May 31, 2012 2:21 am

What albums, new or old, do you find especially good or bad as mp3s, compared to the original 16-44 CD wavs (or vinyl)?

What kind of changes do you notice, good or bad? Any theories why?

I listen to music mostly as 320 mp3s now, so this is a source of ongoing curiosity. Sometimes albums that were clear but not fancy on CD (e.g., Earth Wind & Fire's SPIRIT, which George Massenburg engineered) translate perfectly to mp3 and even sound a little better somehow . . . whereas other stuff that seemed simple and pleasant on CD, like old Joao Gilberto guitar/vocal/shaker cuts, will convert into something not so suave.

(There is also the question of CDs that already sound like bad mp3 conversions, e.g., Animal Collective's MERRIWEATHER POST PAVILION-- is this stuff remastered by Satan? But I digress...)

Thoughts?
"The world don't need no more songs." - Bob Dylan

"Why does the Creator send me such knuckleheads?" - Sun Ra
.
.
.
.

User avatar
Beat Poet
alignin' 24-trk
Posts: 50
Joined: Sun Jan 08, 2012 10:54 am
Location: Hertfordshire, UK
Contact:

Post by Beat Poet » Thu May 31, 2012 7:19 am

Hmmm, it's hard to say. I listen almost exclusively to MP3s now (iPod, laptop) and listen on headphones and think "this is damn good", then I whack the CD in and think "why the hell aren't I listening to this instead?".
Image

User avatar
fossiltooth
carpal tunnel
Posts: 1734
Joined: Sat Mar 24, 2007 3:03 pm
Location: Brooklyn, NY
Contact:

Post by fossiltooth » Fri Jun 01, 2012 2:18 pm

Double-blind scientific testing suggests that almost no one can ever tell the difference between 256kps AAC/MP3 and CD, and that absolutely no one has ever picked out a 320kbps file from a CD.

That's all I can say about that. I've heard rare stories of releases that were somehow encoded improperly, leading to sonic issues, but I don't know of any confirmed cases to share here.

chris harris
speech impediment
Posts: 4270
Joined: Tue Aug 12, 2003 5:31 pm
Location: Norman, OK
Contact:

Post by chris harris » Fri Jun 01, 2012 2:35 pm

^Yep.^ There's nothing at all about encoding to 320kbps mp3 that makes me do anything any differently in a mix.

User avatar
Brett Siler
moves faders with mind
Posts: 2518
Joined: Fri Dec 05, 2003 12:16 pm
Location: Evansville, IN
Contact:

Post by Brett Siler » Fri Jun 01, 2012 4:01 pm

For my own personal mp3 player, I'll usually encode it at 192khz. I feel it's a good balance between sound quality and file size. It's not as good sounding as vinyl or CD but it doesn't sound shitty. It just mildly hollower sounding. I'm not doing critical listening on my mp3 player though, just listening for funzies.

percussion boy
carpal tunnel
Posts: 1512
Joined: Wed May 07, 2003 5:51 pm
Location: Bay Area

Post by percussion boy » Fri Jun 01, 2012 8:37 pm

I'm not claiming to have golden ears when it comes to 320s vs. 44-16 wavs. I have definitely heard 256s translate weirdly in certain dynamic situations (e.g., acoustic plus voice plus shaker).

Do you all really notice no difference from the source audio at 128, though? Last time I bought mp3s off Amazon, that was the only download format iirc. Betcha more casual listeners hear 128s than 192s/320, or CDs for that matter.

So . . . if we include the more space-saving mp3 formats, what albums work or don't work for you, and how come?
"The world don't need no more songs." - Bob Dylan

"Why does the Creator send me such knuckleheads?" - Sun Ra
.
.
.
.

User avatar
fossiltooth
carpal tunnel
Posts: 1734
Joined: Sat Mar 24, 2007 3:03 pm
Location: Brooklyn, NY
Contact:

Post by fossiltooth » Sat Jun 02, 2012 9:57 am

percussion boy wrote:I'm not claiming to have golden ears when it comes to 320s vs. 44-16 wavs. I have definitely heard 256s translate weirdly in certain dynamic situations (e.g., acoustic plus voice plus shaker).

Do you all really notice no difference from the source audio at 128, though? Last time I bought mp3s off Amazon, that was the only download format iirc. Betcha more casual listeners hear 128s than 192s/320, or CDs for that matter.

So . . . if we include the more space-saving mp3 formats, what albums work or don't work for you, and how come?
That's true: some listeners can tell the difference between 128kbps and higher resolutions formats. The data suggests that the majority of regular listeners (even musicians and audiophiles) can't, but I'd guess that many of the better engineers could pick them out in a blind test.

You can test yourself here if you're curious: http://mp3ornot.com. I know at least several people (myself included) who can get 100% all day long on a test like this. But I also know some very good listeners with great taste who can't.

It's kind of irrelevant though, because 128kbps is a dying, if not dead format. Nowadays, the only time we have to deal with awfully low-res sound clips is while watching YouTube at low bitrates. I mean, c'mon -- Both video and audio combined together at 240kbps? That's way worse than 128 mp3s ever were. Those rates approach Napster-era levels of audio fidelity when you might download a song at 64kbps or lower.

I don't think you can really mix "for" such an appallingly low bitrate though. It kind of makes everything sound equally dreadful. Either deal with it or don't, I say!

percussion boy
carpal tunnel
Posts: 1512
Joined: Wed May 07, 2003 5:51 pm
Location: Bay Area

Post by percussion boy » Sat Jun 02, 2012 3:14 pm

fossiltooth wrote:
percussion boy wrote:I'm not claiming to have golden ears when it comes to 320s vs. 44-16 wavs. I have definitely heard 256s translate weirdly in certain dynamic situations (e.g., acoustic plus voice plus shaker).

Do you all really notice no difference from the source audio at 128, though? Last time I bought mp3s off Amazon, that was the only download format iirc. Betcha more casual listeners hear 128s than 192s/320, or CDs for that matter.

So . . . if we include the more space-saving mp3 formats, what albums work or don't work for you, and how come?
That's true: some listeners can tell the difference between 128kbps and higher resolutions formats. The data suggests that the majority of regular listeners (even musicians and audiophiles) can't, but I'd guess that many of the better engineers could pick them out in a blind test.

You can test yourself here if you're curious: http://mp3ornot.com. I know at least several people (myself included) who can get 100% all day long on a test like this. But I also know some very good listeners with great taste who can't.

It's kind of irrelevant though, because 128kbps is a dying, if not dead format. Nowadays, the only time we have to deal with awfully low-res sound clips is while watching YouTube at low bitrates. I mean, c'mon -- Both video and audio combined together at 240kbps? That's way worse than 128 mp3s ever were. Those rates approach Napster-era levels of audio fidelity when you might download a song at 64kbps or lower.

I don't think you can really mix "for" such an appallingly low bitrate though. It kind of makes everything sound equally dreadful. Either deal with it or don't, I say!
You sure are burning a lot of calories to argue that my original question was pointless. :? :shock:
"The world don't need no more songs." - Bob Dylan

"Why does the Creator send me such knuckleheads?" - Sun Ra
.
.
.
.

chris harris
speech impediment
Posts: 4270
Joined: Tue Aug 12, 2003 5:31 pm
Location: Norman, OK
Contact:

Post by chris harris » Sun Jun 03, 2012 8:11 am

I think it's like trying to consider any other unlikely variable. Some people are going to listen on stereos where one speaker or the other is broken. That's no reason to avoid hard panning. My experience tells me that many, many people will listen on stereos with "smile" graphic eq curves. I don't compensate for that. I've seen tons of "loudness" buttons just left in the "on" position. I don't take that into consideration when I mix.

I think that the point that Justin is making is that while you can consider many listener variables, you cannot account for them all in your work. I think that you should do what's always been done in mixing: get to know your monitoring setup, and make the best sounding mixes you possibly can. Once it leaves your studio, you have no control over how people will listen. It's a wonderfully fucked up, unknowable, random, level playing field.

FWIW, if you do care about compensating for people who don't care enough to get something better than 128kbps mp3s to listen to, then you should avoid the high percussive stuff like shakers, tambourines, etc... Those things, as well as some modulation effects like phasers/flangers/chorus, take the hardest hit in low res mp3 conversion.

User avatar
vvv
zen recordist
Posts: 10158
Joined: Tue May 13, 2003 8:08 am
Location: Chi
Contact:

Post by vvv » Sun Jun 03, 2012 9:53 am

FWIW, lemme mention that hard-panned delays can also be problematic in low-res *.mp3's.

I note that a cuppla years ago I posted about problems with panned tambo's and no one knew what I was talking about!

I guess I'm ahead of the curve when it comes to the fucked-up ... :twisted:
bandcamp;
blog.
I mix with olive juice.

percussion boy
carpal tunnel
Posts: 1512
Joined: Wed May 07, 2003 5:51 pm
Location: Bay Area

Post by percussion boy » Sun Jun 03, 2012 11:15 pm

chris harris wrote:I think that the point that Justin is making is that while you can consider many listener variables, you cannot account for them all in your work. I think that you should do what's always been done in mixing: get to know your monitoring setup, and make the best sounding mixes you possibly can. Once it leaves your studio, you have no control over how people will listen. It's a wonderfully fucked up, unknowable, random, level playing field.
That all makes sense to me. And I find both you and Justin have a lot of great thoughts on recording.

I really was just looking to see what other people's experiences had been in the mp3 universe, as listeners. Presumably no one's tripping about how their favorite music sounds at whatever level of mp3 compression, since no one's piped up about it.

Re the obsolescence of 128s, I have to buy some reference mp3s off Amazon tomorrow and will be curious to see what format they're offering this month.

One day there will be peace on earth and 24-96 downloads for all, if civilization lasts.
"The world don't need no more songs." - Bob Dylan

"Why does the Creator send me such knuckleheads?" - Sun Ra
.
.
.
.

smtimecharlie
audio school graduate
Posts: 13
Joined: Fri Jul 24, 2009 6:11 pm

Post by smtimecharlie » Mon Jun 04, 2012 7:37 pm

When I listen to streaming Rhapsody (which I do via iPhone or to preview albums), I'm pretty sure it's 128 kbps. And I've thought about this topic too, it seems like Ryan Adams stuff doesn't sound good as mp3 and I'm not sure why or if it's my imagination. Personally, with my home recorder-grade mic's, the AKG C1000 really crunches up in a bad way in a final mp3. There are those who hate this mic in any format, I know, but I've wondered whether different mic's hold up better to compression.

MoreSpaceEcho
zen recordist
Posts: 6677
Joined: Wed May 07, 2003 11:15 am

Post by MoreSpaceEcho » Wed Jun 06, 2012 1:26 pm

i pretty much never listen to mp3s, but i do know that whenever i hear something i mixed on bandcamp's streaming (128kbps) i want to die. sounds so bad, especially on high frequency transients...anything approaching 0dbfs turns into distortion, and the whole mix gets this ugly metallic sheen to it. hate it.

in general i'd just say records that have a little headroom to them are likely to sound better as mp3/aac.

percussion boy
carpal tunnel
Posts: 1512
Joined: Wed May 07, 2003 5:51 pm
Location: Bay Area

Post by percussion boy » Wed Jun 06, 2012 10:34 pm

smtimecharlie wrote:Personally, with my home recorder-grade mic's, the AKG C1000 really crunches up in a bad way in a final mp3. There are those who hate this mic in any format, I know, but I've wondered whether different mic's hold up better to compression.
I wonder sometimes about how compression in general relates to mp3s that seem distorted.

Can't put my finger on it, but I wonder if recordings that go too far either way can freak out the lower-rent mp3 rates -- ie.,--

-- problem mp3s from WAVS that were previously hypercompressed at mastering, before the m3 process re-compresses them--this kind of aggressive mastering often goes hand in hand with the low headroom Mr. Echo mentioned. ...or...

-- problem mp3s that were not compressed enough at the level of individual tracks during the mix, so that the untamed transients of particular instruments (think tamborines again) somehow confuse the mp3 encoder about how to compress the ensemble as a whole, and the file distorts.

I'm thinking out loud, not sure there's a sweeping conclusion to make here.
"The world don't need no more songs." - Bob Dylan

"Why does the Creator send me such knuckleheads?" - Sun Ra
.
.
.
.

User avatar
vvv
zen recordist
Posts: 10158
Joined: Tue May 13, 2003 8:08 am
Location: Chi
Contact:

Post by vvv » Thu Jun 07, 2012 4:29 am

One observation: in my DAW, when I convert to *.mp3, certain transients spike a little.

For that reason, I always mix down to -.3 dBu rather than 0; this seems to eliminate the overs that I get from the conversion process.
bandcamp;
blog.
I mix with olive juice.

Post Reply

Who is online

Users browsing this forum: No registered users and 25 guests