We should rejoice that we can't really tell a difference (at least between 320 and uncompressed). This is a great technical feat, and saves data space, energy and streaming costs.
ive been telling people this for 20 years - I sit here with a pair of mackie HR 824's and blind tested myself many times...i cant hear difference between 320 MP3 and WAV
Totally, in a a good recording on 320 mp3 there's no way you'll be able to tell the difference from a wav/ aiff file. And the people that said they can hear the difference are lying. Trust me in a blind test no one can tell
I got 2/6 I was definitely guessing each time. To me the biggest take away from this is that if you really have to listen to the recordings so carefully, over and over again to be able to tell which one is not the 128 one... it means it really doesn't matter for your everyday consumption of music:D
have tried this a long time ago, i got it all correct.. and i didnt have any headset on. (actually didnt tried to listen to it at all.) i just wait who loads longer and that surely is the uncompressed wahahaha :D..
dont hate people.. i tried it first with my blon and i got 1 wrong.. but 320 and wave are almost the same or maybe the same. specially if the song is new records, its really hard to tell which is which..
I got 2 right, but I take pride in the fact that I could point out 128 in each one 😂 I always knew 320 and WAV would confuse me, because I never could tell the difference.
I got 2 out of 6 hits. Probably more by chance. Headphones: Beyerdnamic Lagoon ANC. I thought I had good hearing, but had reliably selected MP3 128Kb/s. To me, this is a sign that we have long become accustomed to the "bad" sound.
@@b3naqua Something about the vocal range in the song and how it's super apparent when the compression messes with the sound too much. It was in a section about the development of the MP3 format in the "Book How Music Got Free by Stephen Witt" which is about the history of music piracy. Just finished the book a few months ago, it was really good.
I can definitely tell the difference.. i am huge in car audio sound quality and what i learned was when your system is well tuned, you can make MP3 sound good but also make lossless sound much better...so, with that, lossless stands out as superior sound
Only if it's a bad mp3 (yes, you can rip a CD to 320kbps in the wrong way, bad encoder, bad settings). Use LAME encoder, use High Quality Stereo, 320kbps and noone can tell it from flac.
I really though i would be able to tell the difference MUCH easier. I can notice a difference, but barely and only if i try REALLY hard. Very impressive.
Everyone is capable of telling the difference between 128Kbps and 320kbps but there is no chance to hear the difference between 320kbps and lossless,320Kbps is a really good compression, actually if you look at the sound graphs in a special software,a mp3 at 320Kbps is extremely close to wav! I'm very impressed,these explains why is so hard to tell the difference between those two.
Hearing abilities in human beings are as varied from one individual to another as strength or speed. While I did notice some detail in the speakers, I tried to convince myself that there was no such difference - sure, I didn't want to have to re-rip my entire collection (more than 200 CDs) and encode it to FLAC - but then I bought a decent pair of headphones and decided to do some blind testing. and it became inevitable, I had to move to FLAC.
There is another A/B Test which was originally used by Tidal. I am 23 years old an never listen to loud audio. Even with mx B&W PX8 connected via usb i can't hear a single difference.
I’d even argue that 128 - 320 is irrelevant. If a set contains all 128, can’t tell, if a set contains some 128 and done 320, you’ll hear the difference. It’s all about perception, location and situation. In a controlled environment, yes, in a regular dive bar or local venue? No one will hear it. Ever.
People seem to forget that lossy codecs do not just twist some knobs and that's it; it's the result of research and science. Figuring out what can be taken. That is like "do you need infrared? No, as you can't see it?", just with ultrasonic content. FLAC is like 300 mph, high quality MP3 is 250 mph - but we all drive on a speedlimit of our brain, gear and ears, which is 200 mph. Of course, marketing says we need the 850 mph car, but logically --- nope.
If you're guessing you're not really correct. Anyway, I got 2 but both guesses. Though the 4 I miss were the next best. My take on this is that its not worth paying extra for lossless. At least for me.
The truth is that already at 192kbps it is no longer possible to perceive any difference from the lossless file, using a good codec (such as LAME). Our perception of frequencies from 16kHz upwards is drastically reduced, almost completely disappearing above 18kHz. The theory that we can perceive sounds down to 20kHz is only partially true; in the sense that if the sound at 20kHz is not covered by lower noises, we can (albeit with difficulty) to perceive it. The problem is that the higher frequencies are almost always covered by lower sounds, since our hearing does not perceive the entire audible frequency band in the same way: it focuses mainly on sounds between 2000 and 5000 Hz, that is, on the frequencies of a normal conversation. For this reason it doesn't make much sense to want to reach the highest possible bitrate at all costs, a good 192kbps is more than enough for the needs of the common user.
@@LetsRideIllinois Did you read what it says? It is worldwide "accepted" that 192 kBit/s is artifact free and "transparent". It is commonly agreed upon many people for more than two decades and thousands of experienced and skilled people that 192 kBit/s has no difference. Hence LAME has set 192 kBit/s as "Standard" and considers everything more a bloat effectively by adding unnecessary data to frequencies we don't "need" (as to why: Read the OP). I use VBR0 which is about 260 kBit/s and that is just because it has no lowpass filter. Those tiny bits of extra information are basically not worth the 30% increase of bitrate, but eh, if MP3, then the best setting^^
3/6, never once chose 128kbps. Studio headphones plugged straight into the motherboard. On the first one I just guessed. The cymbals and highs gave it away for me. Load times mentioned in the comments was imperceptible on my pc.
Glad it was helpful for you! Yeah, the differences are so slight/barely audible in most cases.
2 года назад
I can hear a significant difference on my Bluetooth speaker and headphones between the lossless and high quality setting on Apple Music even tho both getting compressed because of the SBC codec
4 out of 6 using galaxy buds 2 pro. Fast internet so everything loaded up about the same. The wave files were crisper, slightly louder and felt easier on the ears. Didn't think there was that much if a difference.
Thank you for this great video, Senior citizen, 75 yr old and MP3 128 sounds as good as High Ris files. That means I don't have to spend bundles of money on high res audio equipment and instead spend it on good media I like! Wish you could put out more videos on audio related things!! Good Job - David N.
It depends, ear buds or a cheap stereo might be bottlenecking the audio to a point that makes it very difficult to tell the difference even to a discerning ear. It might be a different story on high-quality headphones
Nice! Yeah i didn’t really pay too much attention until i tried lossless audio with some nice headphones. Now i can totally hear the difference between AAC and lossless and will crave lossless if i hear any sound degradation in music.
I'm no audio expert but I've had a fair share of experience with high quality files, so here's my two cents of this topic; I took the test using a Hifiman Sundara, Sennheiser HD600, and an iFi Zen Dac V2, I would play the 3 files on the test one time each using the Sundara, and then switching to the HD600 and doing the same thing, really focusing on the most minuscule details. I would eliminate the one that ended up sounding a tiny bit flatter, less body to the projection of sounds (which always ended up being the 128kbs). Then I would go back and forth between the 2 files I thought were the best sounding. It really took a while before my ears would adjust to the point where I would pick up a very very slight difference, that difference being instruments or sounds that are more on the higher frequency range retaining more of a "flare" as it fades off, and just this odd feeling that one file was just missing something the other file had, as if one was just a tiny bit "fuller" sounding. Doing this method, I manage to score 5/6, only missing on Jay-Z's 'Tom Ford', picking the 320kbs on it. My conclusion to this topic is that 320kbs is more than good enough for listeners with decently high end sounding headphones and wanting that high fidelity music experience. Uncompressed files are really only for people who would need such files like music producers who would want to mix and listen to songs to the finest detail they can achieve through their audio gear. For the gamers who obsess with high refresh rate monitors, it's the same comparison from 144hz to 240hz to 360hz. 144 to 240 definitely has somewhat of a noticeable difference, whereas 240 to 360 becomes super difficult for the common user, you have to really know what you're looking for to distinguish between them, and even then, it's not perfect, you could still mistake one over the other. My last and maybe controversial opinion even after all of this, 128kbs is more than good enough for every casual music/audio listener out there. It really takes quite a decent amount of focusing to even distinguish 128kbs from 320kbs and flac/wav. An amount of focus I doubt anyone does when they're just enjoying music or having it running in the background while doing other tasks.
From the moment the youtuber relisten the recording, I already have a great idea that regular people would not recognize a difference from hires to apple musics regular songs
In internet browser , you can’t set up the sound , so the audios will be upsampling or down sampling ,this depend how your computer audio set up is configured it , I tried with a external dac
and that in order to compare them you have to have headphones like his or similar... which I don't have, my headphones sound very good... but you probably won't notice the difference between WAV and MP3 320kbps...
My takeaways from this are : - If it was recorded well, it will still sound good , even compressed. - Audio compression has come a long way - If you aren't hardcore you won't really notice the difference, but it is there and some people will appreciate it.
I got 6 out of 6. Definitely the load times, stage, brightness and midrange was a factor for me. I also noticed the volume of the lossless is a bit low or some would say muffled. But that was a good test.
The only song I was 100% sure which one was which was Dark Horse because of the short calm section which made it quite apparent. Everything else was guess work when it came to 320 to lossless. 128 was easier to get.
3/6 for me and i feel like it's just luck. Personally i've always felt that these lossless audio arguments are a complete waste of time majority of people will never really be able to tell the difference. If it's not even a night and day difference then what's the point the only people that would be defending lossless audio are complete audiophile nerds that care too much abt the minor differences because the reality is most people do not care.
I did the test before watching you and only correctly identified 2/6 but i was using Bose QC2 over bluetooth so 🤷♂️ Im probably going to hold off paying extra for premium streaming for a while 😂 Great stuff as always Ben, you looking great 👌
you should do another test with ABX testing.. it's WAY harder. basically instead of guessing which sounds best you are now guessing which 2 tracks are the same. ie: is track X same as A or B? There is a site abx digitalfeed which does this and even at 96kbps vs 320kbps mp3 its sometimes hard.. and I'm running a Sennheiser HD 6XX with Schiit stack.. I was super surprised. in a regular 128kbps vs 320kbps test I could say 320kbps was better most of the time in a single blind test. But doing the double blind was extremely hard.
I picked the 320 4/6 and lossless 1/6 and I had no idea for the Neil Young, it just sounded flat and quiet on all 3 of the files to me and I never listen to that sort of genre so it was a complete guess on that one.
I tested with SB X7 limited edition DAC + Fostex th900 and answered all 6 correct. Difference was audible. But then I took same test on Mac using Apple stock headphones. I couldn't tell any difference. So I failed the test 1/6 correct. It appears that investing in high end audio equipment really makes a difference in listening experience.
depends i suspect they are using newer compression algorithms on the web site, i remember back in the day anything downloaded at 128 was immediately noticeable as hot garbage its literally what pushed me into seeking higher quality recordings and higher quality gear.
I got kinda the same results in this test but some time after this test i got obsessed with very high quality and high bitrate dsd versions of my favorite albums and i gotta say you can tell the difference there. If you cpmpare a 12000 kbps dsd version of a song and compare it to a 320 mp3 file, you are going to tell the difference much easier. The dsd sound much better in my experience. So i think its about how big is the gap. Because almost always you are comparing the same track and the same version and also on the same audio interface and with the same headphones and if the gap is big enough you gonna notice. This test kinda proved that the gap between a 1.4k wav file and a 320 mp3 is so small that you can't tell right away and have to listen and compare certain detail many times to just make a good guess and still you are not gonna be correct 100 percent of the time.
I tried this NPR test a few times already (before I found this video). I like how it randomises the tracks each time, so you can just try it again and again. My results have been good. I can pretty much always spot the 128 file, but sorting 320 from lossless is much tougher. The test results screen claims that you are more likely to hear differences using better equipment - in my case samson headphones plugged into a shitty HP laptop may not be ideal.
I am 51, And Category 2 on my most recent hearing test, Which is hampered slightly by Tinnitus. I got 3/6. I got two wrong at 320 kbps , And 1 at 128 kbps.
i don't know what magic they did...i've had 128kbps mp3 files before but they're not as good as the ones used in the test...i could clearly discern that my 128kbps is not good compared to my 320kbps versions 😅
Depends on the player. Not all hardware is created equal - some media players have better internal amps, DACs, etc that color/distort the sound in better/worse/unique ways. Higher end stuff is generally better, but not necessary all the time. I’ve used tons of cheaper equipment that sounds just as good to me. Experiment!
Audio is similar to display. Just because a display boasts is hdr capable doesn't mean it will do so well. For example, the buzzword HDR Ready and HDR 400. Both these terms should be read as "yes we can in fact read hdr metadata. But the fuck you want us to do with it?"
I got 3 out of 6. On 5 of them I was able to find the 128 while on 1 I chose 128. I really needed to try hard to hear the differences though, even between 128 and 320. I used audio-tehnica m40x headphones plugged directly into the laptop for the test.
not "kilobytes" but "kilobits" per second, that's the unit which is measuring the bitrate. If you want exact kilobytes per second, you need to divide by 8, so 128kbits is actually only 16 kilobytes per second
Did it myself, 2 out of 6 and I made sure to preload the songs after the first one so it wouldn't fudge the results on load time which made it pretty obvious. Legit could not tell a difference whatsoever. I actually preferred the 128kbs because it seemed to round out the plosives better
I took the test and because I don't have the fastest internet speed, I could identify the highest quality sample by the delay after pressing the play button - the highest quality sample took the longest to download :( Guys at npr should address this issue, there are ways to solve this.
its very interesting , the real question now that im thinking is since it look like there is not a lot of difference between 320 and uncompress), would wired headphone/DAC/amplifier play a bigger role in getting better sound especially vs bluetooth headphone technology and prowess today? I really wonder what is a difference today with top of line Bluethtooth headphone and lets say an Hifiman wired Ananda/XS whatever with a dac . Is the gap really that mindblowing or its just s 320 vs uncompress difference.
Good Q: I think it's subjective, like balancing the convenience of bluetooth headphones w the technical specs and output they can deliver vs the less-processed wired headphones experience. imo they both are good for different use-case scenarios. I love bluetooth for when I'm mobile, walking around the house, gym, etc. But if I'm sitting or stationary for a while, I love wired headphones for that lil bit of extra oomph in sound quality.
4/6 two tracks I got wrong were 320 bit mp3s I used my shitty cheap but honestly decent sound quality philips shp 1900 at %50 volume and decided on feel
i got all but to correct then went back and found all but one low quality intentionally, side note if you click try again at the end they scramble the playlist. i tested on hd599 and beyer tygr 300r with my pc built in dac at 192khz 32bit(something in the software broke and the audio started phasing and distorting). switched my fiio dac set to 384khz 32bit later i will test on my studio monitors and hd 660s and gl2000 to see if more details stand out. but trust me if you do a blind ab comparison on youtube it will stick out clear as day because their 128 compression sucks hell everything that is not 2k video gets smacked with compression.
I tested it on a song by Metallica - One (high res 24 bit 192 khz flac, filesize is... 317 megabytes) Converted it myself to 8/12/16/20/... until 70kbps mp3 with abr presets, then from 70 to 128 with quality v9..v5 presets. Below 40 the sound was complete garbage, below 60 kbps the difference was still obvious, most obvious that it was just very quiet and compressed. I started to barely hear a difference at 70 kbps vs flac, and again it was mostly in the volume - mp3 was a bit more quiet. But at this stage I already felt like I was mostly just guessing and imagining a difference (other than the volume) After 70 kbps the difference was practically non-existant for me. And I can safely say that on 128kbps there is no difference at all vs high res flac. My sound card is asus xonar essence stx and headphones are beyerdynamic dt-990 pro 250 ohm
If we invest 1.5-2k in hifi gear, I think it’ll be easy for us to differentiate between lossless and compressed audio. So, yeah. Some people will notice the difference, but the majority of people who listening to music just casually won’t benefit from uncompressed file.
No u can't..its placebo..its a tactics to buy hifi gear..if we invest that money on medical aid for improving our hearing..we could at least that money would be useful and worth
trust me ive got a very expensive bang and olufson setup and i can't hear a difference i almsot feel robbed blind and stupid for buying it i played lossless and mp3 on it cant tell a difference think its all placebo myself and its like speaker companies selling snake oil....but people will swear if you had this this and this setup it will sound better the number of speaker setups ive had is stupid to the point I don't give a shit anymore Ive spent thousands and thousands searching for the next setup dont be me. Your money is better in your pocket.
Great video, for me the codec is somhoe more important, i love to listen to Spotify via my browser, but i can't stand the desktop app. It sound so different to me. Maybe because the browser has more bandwidth? I dont have premium maybe thats also a reason it sound better on the browser.
to me loseless sounds more clear and in my car the two sound way different to me and that loseless is better when I drive. So i am dropping spotify for Deezer
Def can hear the diff but not on a cheap headphone or stereo system. Try a US6k headphone system or a US10k stereo system ( maybe even higher if you can afford it ) . It’s night and day.
I did thest test had only 3/6 lol, for me it really depends on the song, for some songs I can tell the difference and for others I just can't. Dark horse and Mozart Piano was impossibe to guess for me. Anyway I will stay with mp3 320kbps, I am not an audiophile nor do I have an expensive audio setup, mp3 320kbps is good enough for me.
For me, Tom's Diner had less reverb in the 128, which was the giveaway for me. On Niel Young, the orchestral part had more impact, more dimensionality on the high quality ones. Same goes for the piano piece, the downward arpeggio sounded tinnier and less full on the 128 sample. The Jay Z one was really hard for me, but the vocals sounded a bit more real on the 320 and WAV. Dark Horse was the most difficult for me, I ended up choosing 320 (as I did for the last half of the test) and honestly I probably just got lucky that I didn't choose 128 for this one. (edit) Upon revisiting the Dark Horse sample, the first loud "CLACK" has more snap, impact and realness to it when it's not 128.
Well it also depends if you connect your headphones in the front of your pc or back. The back is always better but since my cable doesnt go that far i had to do it at the front. Headphones quality is also that determines if you can hear the difference. Got a 200 euros (21% vat incl) headphones which can go wireless aswell so its not a headphone to get the best audio experience. So with this only, I guessed 3/6. It was hard to discern between the 320kpbs and uncompressed but if everyone had better equipment like a 500 euro headphone and external audio converter for your pc, then yeah everyone would be able to figure out. Its just the equipment.
i got 3/6 and tbh i think it really depends on A) the genre but...... B) your actual audio equipment.... i think that with a good pair of headphones or earbuds that in honestly you really cant tell the difference. I have a pair of Astro a40's (Gaming headset) and with it comes not only Astros own EQ but also Dolby Digitals EQ as well which makes nearly any song sound fantastic. Even the most raw Black metal sounds awesome i also picked the 3rd option for Dark horse as well lmao
@@b3naqua i have inherited my dads musicians ear, but don't have his hearing loss yet. i'm also mildly autistic, so psychoacustics may be less effective (as it is for people with hearing damage)
I was initially disappointed at the fact that you didn't include the audio clips bc of copyright, then I remembered that RU-vid can't even stream audio at 320kbps
I got 2/6 on my stereo set and 4/6 on AKG Q701s with the Behringer UMC404HD. Though in another A-X-B test I got pretty much 50/50 and it felt much harder. I do have some EQ and a crossfeed filter on the headphones, so I might try without those later on.
The other test was a bit different, it had samples A and B which were always different, but X which was the same as one of the others. The challenge was to figure out is X the same as A or B. It also had you try the same samples multiple times to get an average result.
RU-vid audio is 128 kbps aac so no one can tell the difference at all. Your audio was all leveled when you uploaded. People are so full of crap in the comments. Compression uses perceptual encoding meaning it’s impossible on a scientific basis for most humans to hear a difference because it uses a psychoacoustic algorithm that only removes frequencies we struggle to an extreme degree to try to pick out and most people cannot. 128 kbps is enough to rip music, however I will say that the difference is that the higher bit rate track will sound somewhat more dynamic and will be slightly louder than the compressed version, so you can play it on a bit lower volume and still hear everything, like 1 notch lower. Lol
Tried this, got 2 first time and 0 the next time. I listened through wireless earbuds from my phone though. If i get wired earphones and try it would it work properly, or does listening on your phone compress it anyway?
