advertisement


Honestly - can you hear the difference between lossy and lossless?

Funk

pfm Member
Following on from another thread about streaming sound quality, it lead to wondering if I could genuinely put my money where my mouth is about whether I can tell the difference between Spotify quality and lossless. Seems I got a mixed bag on this test:

http://abx.digitalfeed.net/spotify-hq.html#

iJnZ5PTl.png


Anyone else prepared to take the test and share their results...?

It also seems to suggest that a well-mastered recording (such as Daft Punk) will still sound very good indeed even at 320kbps OGG.

Oh, and if there are reasons why an ABX demo in a browser is pointless please feel free to say why - I'll have to assume it's valid until told otherwise!
 
Even 128 kbps MP3 can be hard to identify as long as you avoid the corner cases that broke it like castanets.
All lossy codecs have their weaknesses, but I would not expect Daft Punk samples to be one of them
 
I don't need to participate in this kind of thing. There's all kinds of challenges inherent.

The only test that counts is in my system with tracks I know. I would also hesitate to categorise anything by Daft Punk as music let alone a decent recording.

I can easily hear differences. Having said that, 320kbps is indeed very good.

If you can't hear differences then don't bother with decent speakers, amps and source, as you won't be able to tell the difference between them.

I'm obviously being provocative, but there is some seriousness to my comment. There's no point in putting together a decent system if you're going to feed it a sub optimal input and you can't hear the difference.
 
I don't need to participate in this kind of thing. There's all kinds of challenges inherent.

The only test that counts is in my system with tracks I know. I would also hesitate to categorise anything by Daft Punk as music let alone a decent recording.

I can easily hear differences. Having said that, 320kbps is indeed very good.

If you can't hear differences then don't bother with decent speakers, amps and source, as you won't be able to tell the difference between them.

I'm obviously being provocative, but there is some seriousness to my comment. There's no point in putting together a decent system if you're going to feed it a sub optimal input and you can't hear the difference.
In a nutshell!
 
I think it’s rather like in cooking, where a small amount of flavour is added to enhance a sauce but not enough to identify. The differences between lossy and lossless may at times be difficult to consciously quantify but the perception of sound is very subtle. The increased enjoyment of a more lifelike presentation may be subconscious. You cannot measure everything.
 
there are no controlled conditions in the test as everyone will be listening on different equipment and listening for different things, result tells us nothing
 
I would not consider my chances to tell the difference with most 320kbps material to be very good, especially not if it's OGG, which is better than MP3.

I read an academic study once, but can't find the link anymore - both OGG and AAC outperformed MP3. I can't remember the used bitrates anymore but in the tests MP3 was transparent with about 90% of the material and with OGG and AAC it was more like 98% or 99%.

Disc space is cheap enough that I compress with lossless FLAC when I rip my own disks and I also decode HDCD and store those tracks as 24 bit. But if I someone was to secretly rerip them with 320 kbps OGG without telling me, I am convinced I would never notice just by the sound alone.
 
I think it is sometimes hard to discern great differences but this is probably a consequence of the quality of the production. Some tracks sound very listenable. As above, I notice it more in the soundstage. Spotify tends to have a narrower, more forward presentation, similar to car radio. Qobuz higher res tends to open out and sound less tense in comparison. But I don’t diss Spotify. It has a broad catalogue and is very listenable. As long as the sound is good enough for me to appreciate the Music, that’s all I’m bothered about.
 
Telling 128 kbps mp3 from lossless can be anything from easy to impossible depending on the content and the encoder. 320 kbps mp3 and more advanced codecs at sufficiently high bit rates are generally very difficult to distinguish. I still prefer to get my music in lossless formats simply because you never know when something will trip up the encoder and cause an audible glitch.
 
Ease of differentiation is very material dependent. There are several types of sonics that the MP3 codec, even at 320, doesn't do well.
 
128 kbps isn't hi-fi quality & I'd never use it for serious listening. 320 can be pretty good, but as always, it's down to the mastering; some vinyl can sound awful. I always flac for my downloads where possible, but I have tons of 320 too.
 
I can easily hear differences.
How do you know?
Tim
If only there were some kind of test... :D

I should note that my PC is my digital front-end, the test was done through my system as I would listen to it daily. What it shows me is that in my case there is absolutely zero point in bothering with anything greater than 44.1/16 as I can only hear the difference with lossless around 2/3 of the time.

I do note that there have been a lot of assertions and excuses yet no other posts with any results though..! :D If people are that confident they can 100% tell the difference then surely this is an ideal way to prove to yourself that you can. Arguing you can 'only do it with tracks you know' doesn't really cut it either I reckon; if your ears are that capable then you'll be able to tell with any material I'm sure.

I've been bold enough to post that I don't have perfect ears - anyone else..? :)

All that being said, given how cheap disk space is I'll always default to FLAC over MP3 or other formats. After all, you can't ever put back in future what's already been removed.

Edit:

If you can't hear differences then don't bother with decent speakers, amps and source, as you won't be able to tell the difference between them.

I'm obviously being provocative, but there is some seriousness to my comment. There's no point in putting together a decent system if you're going to feed it a sub optimal input and you can't hear the difference.
I'd wager the differences between speakers and amps is far easier for people to identify, they're designed to sound different to each other. A lossy/lossless challenge is far more subtle and nuanced as they're designed to try to sound the same (or as close to it a possible) so spotting the differences is trickier.

I've not been afraid to 'hold my ears to the light' as it were and see if I can hear what I like to think I can - and a third of the time, it seems I can't.
 
If you can't hear differences then don't bother with decent speakers, amps and source, as you won't be able to tell the difference between them.

I'm obviously being provocative, but there is some seriousness to my comment. There's no point in putting together a decent system if you're going to feed it a sub optimal input and you can't hear the difference.

The rising tide lifts all boats, it's silly to suggest that if one cannot tell the difference between 320 and flac that a better system won't make both sound better.

I would be surprised with the better system one couldn't hear the difference with music with a decent dynamic range but that's another matter. Mastering is still the most critical issue I feel. A well mastered 320 will most likely sound better than a poorer mastered flac.

I did the Mark Waldrep high res challenge and (like most people) could not reliably differentiate between 16/44 and higher resolutions.

https://pinkfishmedia.net/forum/thr...rence-between-96-24-and-44-16-and-mp3.233495/

.sjb
 
I've been bold enough to post that I don't have perfect ears - anyone else..? :)
I most certainly don’t - anyone of my age that says they do is bonkers.

Also I often wonder how much of what we think we hear is in fact our brains filling in information for us for better or worse - rather like they cover up our blind spots in a way we haven’t yet worked out completely.
 
I took the same test a month or so ago when Robert posted a similar test and got 64 percent. I was guessing most of the time and certainly would not pay more money for lossless. Whatever Spotify Premium is serving up is good enough for me as it sounds great through my system. I spend many hours a week listening to it!
 


advertisement


Back
Top