• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Can You Hear the Difference between Uncompressed and MP3 Audio?

What did you hear?

  • Yes, distinguish all tracks uncompressed vs. MP3

    Votes: 18 25.4%
  • Yes, between 320 MP3 and uncompressed vs. 128 MP3

    Votes: 36 50.7%
  • No, it all sounds the same

    Votes: 17 23.9%

  • Total voters
    71
Joined
Jan 28, 2021
Messages
667 (1.08/day)
This is older but I haven't seen it before. NPR did put up a blind test of several tracks in different genres in uncompressed WAV, 320Kps MP3, and 128Kps MP3. I never did anything like this before but always thought about trying to set something up but its seemed like it would be pretty big ordeal. I had also always been pretty skeptical that there was really much of a difference between FLAC and high-bit rate compressed audio, and pretty skeptical I'd be able to hear anything given that 99% of what I listen to is modern music and not what the typical audiophile listens to in their dedicated sound room.
Annotation 2022-08-24 132944.jpg

Link to NPR test

I have to say I was pretty surprised when I took this, both in what I was able to hear, and also by how I heard things differently on different setups.

I ran though it twice on my livingroom speaker setup, a pair DIY Singularity speakers, NUC HTPC, and Pioneer A9 integrated amp with built in USB based Burr-Brown DAC, so pretty high-endish but not crazy (no room treatmeants or anything). On this setup I was able to pick out the Neil Young and Motzart lossless both times and it felt kinda easy, Jay-Z, and Katy Perry, Susan Vega I picked either 320 or lossless, Coldplay I really felt like I was guessing both times, this highly compressed song just didn't sound good on these speakers and I know I picked 128k once lol. This was done over the weekend and I only did it twice so I need to try again, which isn't enough to really indicate anything other than I could hear that 128k sounded worse than lossless and high quality MP3.

The other setup was my desktop setup which is a NuFroce Icon HDP hooked up USB, and Sony MDR-V6 and Grado SR80, so good but pretty basic (also have KRK V4s and JBL sub but I didn't try with that). I spent way more time with with this setup cause I could crank it late at night and ended up really only relying on results with the Sonys as the Grado's bass response was pretty pathetic and I wasn't really getting any other appreciable benefits vs. the Sonys in other areas, I probably need better headphones. I ran though this like 6-7 times with the headphones and the results where not at all what I expected. I really felt like I had to listen in more critical way vs. what just sounded obviously better on the speakers. I was not expecting that at all and frankly thought it would be a breeze on the headphones. Second is what I was listening for in the tracks on this setup wasn't really what I was thinking I would be hearing. In the Jay-Z track I was picking which track I thought had the best bass and I was able to pick lossless all but once, this was big surprise to me. Susan Vega was the same deal, lossless all but once and I was listening purely to the tone of her voice, none of the room reverb, or decay or anything like that, just voice. Katty Perry skewed more towards lossless but I had some 320 picks too and I felt like I was leaning on which track had the best vocals. Mozart was more of toss up between 320 and lossless and may have had a 128k pick, not sure whats going there as it seemed pretty easy on my Singularities. Coldplay was a mix but the more I went through the rounds I was able to pick 320 or lossless if I just listened to just Chris Martin's voice but I figured that out late in the game. The really weird one is with this setup was the Neil Young track and I have no explanation as to why this is the case but I actually picked 128K every time. Which means I wasn't making random picks but picking the worst quality track as the best sounding?, no clue as to what was happening there.

The tests take a while but I would encourage anyone that has ever been curious to try it and post your results. I will probably go through it again on both systems and actually track my results now that I know that I'm definitely able to hear something. Maybe swap out the DAC too as I recently picked up Schiit Modi and JDS Labs Atom.

The other take away I didn't expect is Katty Perry - Dark Horse sounds sick on my speakers so try it on high-end set of speakers if you have the chance lol.

Results System 1: Passively cooled Intel NUC > Pioneer Elite A9 integrated amp (built in USB Burr-Brown DAC) > Singularity tower speakers: Two runs 3/6 each time. Picked lossless on Neil Young and Mozart, 320 and lossless on Jay-Z and Katy Perry and Susan Vega, tossup on Cold Play.

Results System 2: Dekstop PC > Nuforce Icon HDP > Sony MDR-V6: Six runs about 50% overall picking lossless. Jay-Z picked lossless all but once, Susan Vegga lossless all but once, Katy Perry and Mozart 50/50 lossless and 320, Coldplay mostly lossless and 320, Neil Young 128 everytime.
 
Last edited:

ir_cow

Staff member
Joined
Sep 4, 2008
Messages
1,019 (0.20/day)
Location
USA
With the right speakers I can clearly here the difference between MP3, CD and SACD. Of course the master has be good as well :)

For example. original release of pretty hate machine from Nine Inch Nails is all on tape, but the transfer at the time was poor. Recently (as in 2017), it was remastered and sounds much better. Same master tapes, just reworks with modern methods of transfer.
 
Joined
Jan 28, 2021
Messages
667 (1.08/day)
With the right speakers I can clearly here the difference between MP3, CD and SACD. Of course the master has be good as well :)

For example. original release of pretty hate machine from Nine Inch Nails is all on tape, but the transfer at the time was poor. Recently (as in 2017), it was remastered and sounds much better. Same master tapes, just reworks with modern methods of transfer.
I grew up with MP3 so I knew what low quality MP3 sounded like, I just never sat down and tested any FLAC or other lossless. My stuff at the time was all pretty mid level at best or vintage so there's also that. I have pretty good speakers now but the electronics are probably the weak point at this point.

But yeah, the mastering is obviously a big deal, listening to the Coldplay track through the Singularities was rough didn't matter what the quality was, lol. I'm pretty much only familiar with NIN's later stuff but I may have to check that out, thanks!
 
Joined
Jul 20, 2020
Messages
183 (0.23/day)
System Name Gamey #1 / #2
Processor Ryzen 5 5600 / Core i7-9700F
Motherboard Asrock B450M P4 / Asrock B360M P4
Cooling Cryorig M9A / CM Hyper 212
Memory 16GB 3200 CL16 / 16GB 2666 CL13
Video Card(s) Sapphire Pulse 6600 XT / PNY GTX 1080 w/Raijintek Morpheus 2
Storage 1TB WD SN570 / 512G Toshiba RD400+2TB WD3D Blue
Display(s) LG 32GK650F 1440p 144Hz VA
Case TT Versa H18 / CM N200
Audio Device(s) Dragonfly Black
Power Supply Corsair CX500M / EVGA 650 G3
Mouse JSCO JNL-101k Noiseless
Keyboard Dell KB216p e-waste PRO!
Software Win 10, Throttlestop
Benchmark Scores 144fps in Minesweeper w/Vsync
On the orchestral, the uncompressed and 320mp3 sounded similar but the 128mp3 was flat. I guessed the uncompressed but it was a tossup. With Suzanne Vega's voice, the difference was subtle but I chose the 320mp3 over the uncompressed, also a tossup.

I chose the friggen 128MP3 on all the pop music. Dunno what that means...

Honestly everything outside of the orchestral was pretty much a tossup.

HiFiMan HE 400S, 14" Macbook Pro internal DAC. I'll try later with my DragonFly Black, curious if any detectable changes.
 
Joined
Aug 18, 2022
Messages
202 (3.96/day)
I got 3 out of the 6 correct. USB DAC to Denon amp and Cambridge audio minx min 12 speakers with a yamaha sub
 
Joined
Nov 13, 2007
Messages
9,224 (1.69/day)
Location
Austin Texas
System Name Espresso Machine
Processor 12600K @ 5.2Ghz 1.235v
Motherboard MSI 690-I PRO
Cooling 280MM AIO
Memory 32 GB DDR5 6000 MHZ 36-36-36-95
Video Card(s) MSI Ventus RTX 3080 UV @1810mhz 800mv
Storage 2x 2TB WDC SN850, 1TB Samsung 960 prr
Display(s) LG CX 48" 120hz, Samsung 28" 4k 144hz
Case SLIGER S610
Audio Device(s) Bose Solo
Power Supply Corsair SF750
Mouse Superlight wireless
Keyboard 65% mini hyperspeed wireless
Software Windows 11
On the orchestral, the uncompressed and 320mp3 sounded similar but the 128mp3 was flat. I guessed the uncompressed but it was a tossup. With Suzanne Vega's voice, the difference was subtle but I chose the 320mp3 over the uncompressed, also a tossup.

I chose the friggen 128MP3 on all the pop music. Dunno what that means...

Honestly everything outside of the orchestral was pretty much a tossup.

HiFiMan HE 400S, 14" Macbook Pro internal DAC. I'll try later with my DragonFly Black, curious if any detectable changes.

Pop music is a bit like the audio version of minecraft graphics -- you really dont gain much as the resolution increases :laugh:. You probably start noticing more artifacts.
 

ir_cow

Staff member
Joined
Sep 4, 2008
Messages
1,019 (0.20/day)
Location
USA
To be fair, the difference between CD and SACD/96-24 FLAC often comes down to the smallest of details. Each instrumental is more distinct and the highs are generally not a abrupt cutoff. Rather it rolls off and the whole track is less distorted and muddy. Once again it needs to a good mix from the master. Otherwise it will sound exactly the same.

Oh yeah.. you also needs like expensive headphones ($300) or speakers, otherwise it will sound the same no matter the source. Speakers can be a bit tricky. Price doesn't mean better but I can tell you that there is a massive difference between Klipsch and Polk speakers to something a bit more pricey like KEF or GoldenEar. I'm rocking these old KEF XQ40's for my front mains. Bought it super cheap. Craiglist FTW. GoldenEar is better, but outside my budget atm.
 
Joined
Jan 5, 2006
Messages
13,755 (2.25/day)
System Name AlderLake / Laptop
Processor Intel i7 12700K / Intel i3 7100U
Motherboard Gigabyte Z690 Aorus Master / HP 83A3 (U3E1)
Cooling Noctua NH-U12A 2 fans + Thermal Grizzly Kryonaut Extreme + 5 case fans / Fan
Memory 32GB DDR5 Corsair Dominator Platinum RGB 6000MHz CL36 / 8GB DDR4 HyperX CL13
Video Card(s) MSI RTX 2070 Super Gaming X Trio / Intel HD620
Storage Samsung 980 Pro 1TB + 970 Evo 500GB + 850 Pro 512GB + 860 Evo 1TB x2 / Samsung 256GB M.2 SSD
Display(s) 23.8" Dell S2417DG 165Hz G-Sync 1440p / 14" 1080p IPS Glossy
Case Be quiet! Silent Base 600 - Window / HP Pavilion
Audio Device(s) ALC1220-VB + ESS ES9118 DAC / Realtek onboard + B&O speaker system
Power Supply Seasonic Focus Plus Gold 750W / Powerbrick
Mouse Logitech MX Anywhere 2 Laser wireless / Logitech M330 wireless
Keyboard RAPOO E9270P Black 5GHz wireless / HP backlit
Software Windows 11 / Windows 10
Benchmark Scores Cinebench R23 (Single Core) 1936 @ stock Cinebench R23 (Multi Core) 23006 @ stock
Sorry, didn't have the patience for this test, but I had the first one correct...:D
 
Joined
Nov 16, 2007
Messages
877 (0.16/day)
Location
Hampton Roads
Processor Xeon 5650
Motherboard SABERTOOTH X58
Cooling Fans
Memory 24 GB Kingston HyperX 1600
Video Card(s) gtx1060
Storage small ssd
Display(s) Dell 2001F, BenQ short throw
Case Lian Li
Audio Device(s) onboard
Power Supply X750
Software Mint 19.2
Benchmark Scores not so fast...
As said many times, garbage in, garbage out.

And that said, NY has a healthy obsession with superior sound on the masters. Most classical labels are the same way....you must put the best audio into the recording. That's prob why you chose the lowest bitrate, as it didn't really matter.

I haven't done the test, but .mp3s have this strange quirk of putting a wierd spatial inflection on "s" hiss type sounds, it seems at any bitrate.

I have three (four including car Pioneer) different sound systems...two in the shop and one in the living room. LR is spatial simulated surround from Yamaha soundbar and sub. The room acoustics make everything sound equally crappy.

The shop systems are powerhouses. JBLs and Cerwin Vegas pushed by old Sony AV receiver and Pioneer THX, respectively. Audio sources are flacs ripped from vinyl or CD, and and 44.1kHz 128kb/s mp3s. I haven't tested the DACs in the Pioneer yet, but thia thread gives me the motivation to do so. Noise is nonexsistant. But the good sources (perfect engineering, nice mix levels, etc) are great to listen at any volume.

But, there is a band called Kyuss that has the shittiest recordings, and the music wants to be played loud but it cant, while King Tubby, all kinds of noisy can be played loud.

I will do the test in a couple weeks, I promise.
 

ir_cow

Staff member
Joined
Sep 4, 2008
Messages
1,019 (0.20/day)
Location
USA
With my cheap computer headphones, I can't hear the difference at all in these tests. I even picked 128kbps a few times lol. My second choice was always uncompressed and never 320 Kbps.

3 out of 6. 3 choices were uncompressed (correct) and 3 where 128 Kbps. Katty Perry sounded nearly the same on all three but the tossup with 128 or uncompressed. I picked the wrong one. JayZ and Mozart: Piano Concerto where easier. waay to much clipping going on in 128 Kbps. If you have shit audio in the first place, Lossless 96/24 tracks aren't going to sound any better :)
 
Last edited:
Joined
Jun 27, 2019
Messages
1,066 (0.89/day)
Location
Mid EU
System Name I don't name my systems.
Processor i3-12100F
Motherboard ASUS PRIME B660-PLUS D4
Cooling ID-Cooling SE 224 XT ARGB V3 'CPU' / 3x Raijintek Auras 12 + 3x Cooler Master MF120L black case fans
Memory 2x8GB G.SKILL Ripjaws V DDR4 3200MHz
Video Card(s) Asus TuF V2 RTX 3060 Ti @1950 MHz Core/980mV Undervolt
Storage 3 TB Toshiba P300, 1 TB WD Blue, 1 TB Kingston A2000 NVMe, 256 GB Adata Spectrix s40g NVMe
Display(s) 29" 2560x1080 75 Hz / LG 29WK600-W
Case In Win 101c Black
Audio Device(s) Onboard + Hama uRage SoundZ 700
Power Supply Seasonic CORE GM 500W 80+ Gold
Mouse Canyon Puncher GM-20
Keyboard SPC Gear GK630K Tournament 'Kailh Brown'
Software Windows 10 Pro
Can't say that I notice much but thats a given since I'm using a ~60$ headset with on board sound on my B660 mobo. :laugh:

Tho it might be my imagination but the 320 Kbps sounds a bit different than the 128 but I can't really put my finger on it, I guess if I wasn't trying to hear the difference I wouldn't even notice.
Uncompressed vs 320 nope can't notice anything with my setup.

To be honest its no surprise, after all I'm peasant enough to listen music on Youtube and don't mind it at all.:oops:
 
Joined
Jan 3, 2021
Messages
998 (1.55/day)
Location
Slovenia
Processor i5-6600K
Motherboard Asus Z170A
Cooling some cheap Cooler Master
Memory 16GB DDR4-2400
Video Card(s) IGP
Storage Samsung 850 EVO 250GB
Display(s) 2x Oldell 24" 1920x1200
Case Bitfenix Nova white windowless
Audio Device(s) E-mu 1212m PCI
Power Supply Seasonic G-360
Mouse Logitech Marble trackball, never had a mouse
Keyboard Key Tronic KT2000, no Win key because 1994
Software Oldwin
Pop music is a bit like the audio version of minecraft graphics -- you really dont gain much as the resolution increases :laugh:. You probably start noticing more artifacts.
So true - that Coldplay sound is like it went through some mobile phone compressor & codec chip with a low-res Realtek logo on it before it even reached the master tapes. It also changes quality from annoyingly dull to annoyingly bright in the middle of that half-minute sample, an effect I used to hear when setting the azimuth on my cassette player. I can't imagine I could tell WAV from 128k (or 64k, for that matter).

I haven't done the test, but .mp3s have this strange quirk of putting a wierd spatial inflection on "s" hiss type sounds, it seems at any bitrate.
mp3s often ruin the sound of cymbals, that's what I usually detect first when I listen to 128k-160k files. But even that isn't always true. I guess it's worse when music is more complex, meaning that the encoder must make more compromises.
 
Joined
Jan 28, 2021
Messages
667 (1.08/day)
On the orchestral, the uncompressed and 320mp3 sounded similar but the 128mp3 was flat. I guessed the uncompressed but it was a tossup. With Suzanne Vega's voice, the difference was subtle but I chose the 320mp3 over the uncompressed, also a tossup.
Thats exactly how I felt about Mozart and Neil Young listening through the speaker setup. 128k sounded pretty dead and was easily eliminated and it was more careful listening to what I thought was 320 and lossless.
Pop music is a bit like the audio version of minecraft graphics -- you really dont gain much as the resolution increases :laugh:. You probably start noticing more artifacts.
Thats what I expected to happen but thats not what I got. I picked out the lossless Jay-Z track all but once from six runs and all but maybe once 320k or lossless with Dark Horse.
To be fair, the difference between CD and SACD/96-24 FLAC often comes down to the smallest of details. Each instrumental is more distinct and the highs are generally not a abrupt cutoff. Rather it rolls off and the whole track is less distorted and muddy. Once again it needs to a good mix from the master. Otherwise it will sound exactly the same.
Interesting. Did you verify that with a blind listening setup? I mean I trust what I'm hearing but the psychology and the power of suggestion is real thing regardless.
Oh yeah.. you also needs like expensive headphones ($300) or speakers, otherwise it will sound the same no matter the source. Speakers can be a bit tricky. Price doesn't mean better but I can tell you that there is a massive difference between Klipsch and Polk speakers to something a bit more pricey like KEF or GoldenEar. I'm rocking these old KEF XQ40's for my front mains. Bought it super cheap. Craiglist FTW. GoldenEar is better, but outside my budget atm.
All my speakers are DIY so few are going to even know what the fuck they even are. The Singularities are pretty serious speakers though so I have a feeling my weak point is either electronics or the room. I already have plans to building two different higher-end bookshelfs, one with all Peerless drivers from their higher-end range and one with Dayton RS woofer and Peerless/Vifa tweeter. Also have a Schiit Saga+ pre and parts to build a IcePower class D amp as well as parts to build a Nelson Pass ACA class A amp.

I was pretty regularly nailing lossless on the Jay-Z and Susan Vega tracks with my $75 MDR-V6, granted the NuForce DAC/Amp as like $400 (when new) but I don't think there was a lot in that. And I do have to say I was listening more diligently with the headphones to even weed out the 128k whereas that was easily picked out through the speakers on most tracks, and it all (aside from Coldplay) was just way more enjoyable on the speakers. I think now that I know what to listen for I could probably hear the same things I heard on Singularities that I did on the headphones.

To be honest its no surprise, after all I'm peasant enough to listen music on Youtube and don't mind it at all.:oops:
A lot my music listening is YouTube also and yeah, the other take away is I really don't need lossless or even high bit rate MP3 to enjoy music even on relatively high-end gear.
 
Joined
Mar 20, 2022
Messages
213 (1.05/day)
Processor Ryzen 5 5600X
Motherboard Asus B550-F Gaming WiFi
Cooling Be Quiet! Dark Rock Pro 4
Memory 64GB G.Skill Ripjaws V 3600 CL18
Video Card(s) Gigabyte RX 6600 Eagle (de-shrouded)
Storage Samsung 970 Evo 1TB M.2
Display(s) 2 x Asus 1080p 60Hz IPS
Case Antec P101 Silent
Audio Device(s) Ifi Zen DAC V2
Power Supply Be Quiet! Straight Power 11 650W Platinum
Mouse JSCO JNL-101k
Keyboard Akko 3108 V2 (Akko Pink linears)
Their little "visual" illustration where the images become grainier is very misleading. In reality it's never such a blatantly obvious degradation in quality.

It's like trying to differentiate PNG and JPG image formats. It's possible but you've really got to concentrate while doing back-to-back A-B tests to notice any subtle differences.
 
Joined
Jul 19, 2006
Messages
43,380 (7.32/day)
Processor AMD Ryzen 5900x
Motherboard MSI x570s Carbon Max WiFi
Cooling Corsair AIO 360. Phantek T30 fans.
Memory 32GB G.Skill 3600Mhz CL14
Video Card(s) Zotac RTX 3070 Ti Trinity OC Watercooled
Storage SSD's
Display(s) MSI MAG322CQR
Case Lian Li PC 011 Dynamic
Audio Device(s) Schiit Modius DAC, SMSL SP200 amp. HiFiMan HE400i Headphones PreSonus Eris 4.5" monitors.
Power Supply Corsair H1000i
Mouse Razer Viper V2 Pro
Keyboard Fantech Maxfit 67
Software Windows 11
I missed it with the Neil Young song (just so so much mids), but I could differentiate the uncompressed audio on the others using what's in my system specs with the powered monitors.
 
Joined
Feb 22, 2016
Messages
489 (0.20/day)
Processor Intel i5 8400
Motherboard Asus Prime H370M-Plus/CSM
Cooling Scythe Big Shuriken & Noctua NF-A15 HS-PWM chromax.black.swap
Memory Crucial Ballistix Sport LT
Video Card(s) ROG-STRIX-GTX1060-O6G-GAMING
Storage 500GB 850 Pro
Case Cooler Master MasterCase Pro 3
Power Supply Seasonic S12II 520w
Software W11 Pro
Where to even begin.

  • There is no wrong or right. Hearing, background noise, many many things vary in wide amounts within your own and everyone else's environment.
  • On the face of it this is a simple test. Reality is not kind to that assertion.
  • Bias, you have it. You have accomodated hearing things a certain way from certain transducers. This determines what sounds better or worse.
  • Compressed music, as in compressors were used in production, sounds better in the car than on a good home system.
  • Uncompressed can have wide variances in the loudness and more accurately portray location of each instrument that sound better at home.
  • Compressed music makes Mp3 or wav sound nearly identical because so much information was thrown away in production.
  • MP3 are frequency limited to 16KHz. CD quality lossless files (flac/wav/etc) are frequency limited to 22.05KHz. Humans typically hear 20Hz-20KHz.
  • No recording sounds exactly like the live event. Furthermore, same file on headphones, home stereo, and car stereo will sound different.
  • Almost everything newly released that isn't acoustic was shaped to be more pleasing on lower end equipment and in lossy formats.
  • Digital is tricky and highly complex. Youtube sounds great on nearly everything despite outputting lossy frequency limited audio.
  • Big corporate budgets and the best minds constructed WAV and are what set youtube apart. As always, follow the money.
  • Flac is versatile. You can compress files to be smaller or not at all. On better equipment you will begin to find it is slightly inferior to wav.
  • Momentary listening is highly fallible. Over time these test files will reveal their true worth if revisited with some attempt to listen seriously.
  • Memory across time is fallible. 30 seconds is more than enough. Relax and takes notes if your ADD is going to contradict itself.
  • There is no wrong or right to be assessed. One format is the least technically accurate and another is the most technically accurate.
 
Joined
Jun 16, 2022
Messages
20 (0.18/day)
Location
Hungary
System Name UNDERVOLTED (UV) silenced, 135W limited, high energy efficient PC
Processor Intel i5-10400F UV @ 65W/35W PL1/PL2 LIMIT
Motherboard MSI B460 Tomahawk
Cooling Scythe Ninja 3 rev.B @ 620RPM
Memory Kingston HyperX Predator 3000 4x4GB UV
Video Card(s) Gainward GTX 1060 6GB Phoenix UV 775mV@1695MHz, 54% TDP (65 Watt) LIMIT
Storage Crucial MX300 275GB, 2x500GB 2.5" SSHD Raid, 1TB 2.5" SSHD, BluRay writer
Display(s) Acer XV252QZ @ 120Hz
Case Logic Concept K3 (Smallest Full ATX/ATX PSU/ODD case ever..)
Audio Device(s) Panasonic Clip-On, Philips SHP6000, HP Pavilion headset 400, Genius 1250X, Sandstrøm Hercules
Power Supply Be Quiet! Straight Power 10 500W CM
Mouse Logitech G102 & SteelSeries Rival 300 & senior Microsoft IMO 1.1A
Keyboard RAPOO VPRO 500, Microsoft All-In-One Keyboard, Cougar 300K
Software Windows 10 Home x64 Retail
Benchmark Scores More than enough Fire Strike: 3dmark.com/fs/28356598 Time Spy: 3dmark.com/spy/30561283
To be honest its no surprise, after all I'm peasant enough to listen music on Youtube and don't mind it at all.:oops:

Youtube uses OPUS that could reach a very good quality. Depends on the uploader you hear good quality audio or not.
But even a well parametered encoded mp3 could reach the listening quality of original CD. And only can spectrum analyzer can decide which does have more audio information but not human ears.
But, sadly many mp3 around the internet converted with some simply wizard program for dumbs with no chance to set paramters in the mp3 encoder. For mp3 encoding one of the best is LAME. EZ to make 192kbit mp3 that very hard to decide it is mp3 or the original CD. (Of course some material need more bitrate, but also EZ to encode them to VBR...)
I recommend to try LAME. I already convert most of my CD and i just listen them in mp3 while the CDs are in safe place.
 
Joined
Apr 6, 2021
Messages
504 (0.92/day)
Location
Bavaria ⌬ Germany
System Name ✨ Lenovo M700 [Tiny]
Cooling ⚠️ 78,08% N² ⌬ 20,95% O² ⌬ 0,93% Ar ⌬ 0,04% CO²
Audio Device(s) ◐◑ AKG K702 ⌬ FiiO E10K Olympus 2
Mouse ✌️ Corsair M65 RGB Elite [Black] ⌬ Endgame Gear MPC-890 Cordura
Keyboard ⌨ Turtle Beach Impact 500
Man, that's really a great find. :) Bookmarked.

I got 3/6 correct with my low/mid grade combo (AKG K702 & FiiO E10K Olympus 2).
1, 4 & 6 was a miss, picked the 320kbps versions, though. Lousy 128kbps was easily spotted right away. I always had the impression there was no difference when comparing album downloads, so I sticked with 320kbps (or VBR, which I didn't hear a difference to 320kbps). But have to say back then my audio gear was a lot worse.

Have to say with my gear the difference between uncompressed & 320kbps is hardly noticeable. HD surround mixes (DTS-HD or Dolby TrueHD) for movies on the other hand is night & day, on the surroud system. Still remember how I got the chills from Fast & Furious 1 or The Dark Night. Huge difference.
 
Joined
Dec 8, 2020
Messages
1,817 (2.72/day)
What happens when you can tell which is 128k and ignore that 1, but can still tell the difference between 320k and uncompressed, but don't which of the two is uncompressed.

I think hearing the original uncompressed, the guessing which of the 3 was the original you heard could produce different results.

----

The setup: Realtek S1220-A > DTS DCH driver > Optical > Logitech Z906 Class-D amp.
 
Last edited:
Joined
Jan 28, 2021
Messages
667 (1.08/day)
Bias, you have it. You have accomodated hearing things a certain way from certain transducers. This determines what sounds better or worse.
Certain speakers and headphones make different music sound better or worse but I'd expect that the track that is encoded in the highest quality would always sound the best regardless of what you are playing it back through.
Compressed music makes Mp3 or wav sound nearly identical because so much information was thrown away in production.
It really depends. The Jay-Z track is highly compressed (low dynamic range kind of compression) and I was able to pick out the lossless 75% of the time, and at least 50% of the time on the Katy Perry track.
MP3 are frequency limited to 16KHz. CD quality lossless files (flac/wav/etc) are frequency limited to 22.05KHz. Humans typically hear 20Hz-20KHz.
I thought high-bitrate MP3s from modern encoders like LAME covered 20-20,000Khz?

Unless you are 13 years old its highly unlikely you are hearing 20Khz. Doesn't really matter though as you are hearing various harmonics of higher frequencies in the music so even if you are not hearing everything in the upper ranges you are still hearing the material. Thats why audio engineers can keep designing really good sounding gear and making good mixes well into their 60's and beyond.
Digital is tricky and highly complex. Youtube sounds great on nearly everything despite outputting lossy frequency limited audio.
Youtube sounds fine but sounds worse than my high bitrate MP3 encodes. I guess I don't know for sure if its the same master but it sounds like the same mix.
lac is versatile. You can compress files to be smaller or not at all. On better equipment you will begin to find it is slightly inferior to wav.
Why would FLAC be inferior to WAV?
1, 4 & 6 was a miss, picked the 320kbps versions, though.
FYI, the tracks are randomized each time you load the page.
 
Joined
Aug 6, 2020
Messages
598 (0.75/day)
since lame alt preset standard, I haven't been able to tell the difference

But before that you needed the whole 320 to get many tracks to be transparent.

I don't have the patience to do these listening tests anymore
 
Last edited:
Joined
Jan 3, 2021
Messages
18 (0.03/day)
The difference between compressed MP3 128kbps and uncompressed FLAC is the same as between PCM and DSD.
 
Last edited:
Joined
Feb 22, 2016
Messages
489 (0.20/day)
Processor Intel i5 8400
Motherboard Asus Prime H370M-Plus/CSM
Cooling Scythe Big Shuriken & Noctua NF-A15 HS-PWM chromax.black.swap
Memory Crucial Ballistix Sport LT
Video Card(s) ROG-STRIX-GTX1060-O6G-GAMING
Storage 500GB 850 Pro
Case Cooler Master MasterCase Pro 3
Power Supply Seasonic S12II 520w
Software W11 Pro
I believe this set of test files is intended to offer a slightly more subtle set of circumstances than my attempt to lay down a fairly simplified set of base knowledge sufficiently covered.

Certain speakers and headphones make different music sound better or worse but I'd expect that the track that is encoded in the highest quality would always sound the best regardless of what you are playing it back through.

Encoding doesn't add quality that was never there to start with. Think of the file format as a package that does a better or worse job of protecting what is stored inside it. Lossless packaging intends to not just keep what it stores in, but also everything else out. In the modern developed world this is a whole lot.

It really depends. The Jay-Z track is highly compressed (low dynamic range kind of compression) and I was able to pick out the lossless 75% of the time, and at least 50% of the time on the Katy Perry track.

Encoding to a lossy format can and does change the sound. Per the statement above lossy also means gainy. Intentionally limiting dynamic range by pushing everything upwards to near clipping is one way to hide noise and other interference that does find its way in. When you then overload the middle range (KP) there is less contrast to differentiate with (Jay-Z). Needless to say us and the animals are hardwired to place immediate mortal stress upon interpreting sounds of a threatening nature even inside a rap track.

Again, making what sounds good to you on your equipment is a highly developed response the labels are deeply invested in creating. I really wanted to point out how drastically they attempt to make the only release of certain material sound better through catering to the capabilities of known responses by commonly available equipment. If it sounds better to you on a $100 set of headphones than during a demo of $3000 headphones you are not wrong. Noting what expectation bias is can be equally important to anyone taking this test seriously.

Relax and enjoy what you find impactful.

I thought high-bitrate MP3s from modern encoders like LAME covered 20-20,000Khz?

Unless you are 13 years old its highly unlikely you are hearing 20Khz. Doesn't really matter though as you are hearing various harmonics of higher frequencies in the music so even if you are not hearing everything in the upper ranges you are still hearing the material. Thats why audio engineers can keep designing really good sounding gear and making good mixes well into their 60's and beyond.

MP3 have trouble storing frequencies above 16KHz which eventually resulted in some answers with higher bit rates. Effectively this means they allow unwanted elements to pervade across a further range while losing more information (lossy) as well. Size was their major benefit over lossless formats. What I don't want to leave unexposed is the human element, artistry, that can be poured into making the best use of a chosen media format. Someone who makes MP3 to their utmost is not worried about the constraints or benefits of other options they didn't choose.

Hearing acuity and full range response are audiophile qualities. Some people still have better than 20/20 eyesight at a rather advanced too. This is where the normal person starts thinking they are crazy and believe me it drives those inflicted with such high level sensitivity in one sense crazy as well! Audio pros are a mixed bag depending on how well they protected their hearing. The great leveler being what is locked inside their heads doesn't actually require that great of hearing to conceptualize inside a high level of experience.

By this point you should realize I broadly generalized where it was safe to do so. For the purpose of exposing what will determine success in this social experiment. Limiting discussion out of the gate with high level maths and theory serves nobody.

Youtube sounds fine but sounds worse than my high bitrate MP3 encodes. I guess I don't know for sure if its the same master but it sounds like the same mix.

Local playback compared to streaming essentially. There are many reasons this might be the case.

When you listen to these high bitrate MP3 you open a player program which nowadays does a fair amount in cooperation with the OS to lessen the effects of other processes etc. Chances are youtube follows a much simply chain to reach your ears. Upload quality is another example already mentioned. Don't overthink it but realize your hardware and software chain can be optimized for one or the other. Doing both with realistic means? Third party software that handles both streaming and local playback produce a more even output.

Why would FLAC be inferior to WAV?

Which is to ask why do CD's still use Redbook PCM which is effectively WAV. The non-audible problem WAV solved immediately is security in a constantly changing landscape over the duration of all content previously and yet to be released. Apple's favored lossless format needed to meet this exact same accountability to secure distribution (which they still use for DRM). Flac is a good thing lying outside the immediate realm of corporate interests taking hold if it with their interests. WAV should be considered more of an institution.

Better than 40 years of development and refinement is also part of the answer by now. Flac is a much newer open source solution created to deal with digital delivery and storage (greatly reduced file size). Until very recently bandwidth was too limited for even flac streaming (tip of the iceberg logistically and technically). Not so long ago consumer hardware also struggled to unpack flac files as well as they handled reading uncompressed WAV that have been used since the inception of digital.

The best answer is verify with your own ears. If you hear a difference it may not be present on every file or every release of the same material. With exceptional quality recordings that were turned into exceptional quality lossless flac and wav files which are played back to back on exceptional quality and well tuned equipment there is less difference with every software or firmware update. Where you encounter discrepancies that are clearly audible is mostly reprinted/remastered/redistibuted albums. When an album has never been out of print you find some timeframes or factories or files the distributor provided did a bit better at transferring the music.
 
Last edited:
Joined
Aug 12, 2020
Messages
698 (0.89/day)
2/6 only identified Katy Perry and Coldplay using my speaker + integrated audio and had to listen carefully to even make a meaningful attempt.
I think I'm fine with 128 kbps as is :roll:
 
Joined
Sep 17, 2014
Messages
17,183 (5.84/day)
Location
The Washing Machine
Processor i7 8700k 4.6Ghz @ 1.24V
Motherboard AsRock Fatal1ty K6 Z370
Cooling beQuiet! Dark Rock Pro 3
Memory 16GB Corsair Vengeance LPX 3200/C16
Video Card(s) MSI GTX 1080 Gaming X @ 2100/5500
Storage Samsung 850 EVO 1TB + Samsung 830 256GB + Crucial BX100 250GB + Toshiba 1TB HDD
Display(s) Gigabyte G34QWC (3440x1440)
Case Fractal Design Define C TG
Audio Device(s) Situational :)
Power Supply EVGA G2 750W
Mouse XTRFY M42
Keyboard Lenovo Thinkpad Trackpoint II
Software W10 x64
Give me a FLAC vs any quality MP3 and ill pick em out one by one. Even regardless of speakers; Ill pick them out over mobile phone speakers 9 out of 10. The dynamic range versus static full range is enough.

128kbps is even easier, one hihat can tell it all.
 
Top