[Metalomaniac]

I've been doing AB blind testing of 130kbps MP3s and HDTracks hires audio FLACs. And I'm not able to tell the difference.
 
I bought The Door's L.A. Woman and Rush's Presto from HDTracks. I compressed them using LAME's v5 settings, which yields a file of around 130kbps. As the source is the same, there's no issue with matching volume, different mastering, etc.
 
I put files through a filename randomizer, and put them on my "hi-res" MP3 player, Sony NWZ-F886 with FiiO Kilimanjaro portable amp, and KEF M200 IEMs.
 
Now, hi-res files are eighteen times the size of the 130kbps MP3s, and I was struggling to tell any difference in the audio. After going through the 30 tracks twice over a few months, it is clear that my success rate is purely random.
 
My gear is entry-level for people on this forum, but mid-range to high-range to anyone who's not. In any case, I would expect to be able to tell the difference between the audio quality of a low bitrate MP3 and hires audio even with this gear.
 
So, the question is: is my gear really that lacking, is it my hearing, or is hi-res audio snake oil? I've moved to Spotify, and am considering stopping to use the extreme quality to fit more music on my portable player as it seems I can't tell the difference anyway.

 

[willsw]

Short answer: the type of music, source file for the encoder, and encoder could yield a compressed file that sounds very close or basically indistinguishable from lossless (I'm not going to say anything more than "lossless"). Like, for instance, taking a very good source file and listening to that same file encoded at a lower rate with a quality encoder, as you've done. 
 
You could try finding a version of those albums that probably wouldn't mention the quality but the file for the album would be quite small, and using this as a comparison, if you really wanted to prove to yourself that there is less information in 130kbps than lossless. Or do some of the online tests that are around. I think most people would admit that they're not hearing a difference between high-quality 320kbps and lossless while walking around with their portable rig, and more music is preferable to more bits in that case.
 
As far as anything about 16-bit versus anything above it, head over to the Sound Science forum and I won't meet you there. 

 

[jodgey4]

LAME is pretty good these days. I keep everything as 256kbps+, and I can tell when its 256 but it doesn't bother me as much as like 192 or 128 does. You need really good gear to notice the difference.

 

[RRod]

So, the question is: is my gear really that lacking, is it my hearing, or is hi-res audio snake oil? I've moved to Spotify, and am considering stopping to use the extreme quality to fit more music on my portable player as it seems I can't tell the difference anyway.

 
In my own testing, material has mattered much more than gear. Search for "killer samples" on google to get a sense for the kind of material where differences can be heard more easily / at lower settings. Also note that lame (at least on my home box) uses different psycho-acoustic settings for CBR and VBR, so for instance there are killer samples that I can ABX at 320 CBR but not at V0.

 

[Metalomaniac]

Thanks for the responses, good to hear my gear and my ears are not - necessarily - totally useless
 
For listening to music, I guess I shouldn't be too concerned about bitrate, then. I'm considering upgrading my IEMs from KEF M200, and FLC 8S seems to be a good upgrade.
 
I take the source material point made above, though: it makes sense to pay extra for a master that is not marred by loudness wars, but there's probably not much sense in paying for 24 bit hi-def audio. Of course the two go hand-in-hand, but not always.

 

[hex182]

Just get lossless if you can and don't worry about it, it's better than not getting lossless because the it will be harder to not worry about it.

 

[Metalomaniac]

  Just get lossless if you can and don't worry about it, it's better than not getting lossless because the it will be harder to not worry about it.

 
Why would I get lossless if I can't tell the difference, it takes 10x more space, and costs much more?

 

[willowbrook]

Why would I get lossless if I can't tell the difference, it takes 10x more space, and costs much more?

Then you don't. You get 320kbps CBR mp3 files and be done with. However storage is so cheap these days that storing lossless files is not that big of a deal. Lossless file is not only about hearing that 20~20.5khz to 22khz (most people tops at 20), but it's also about hearing the full spectrum above the shelf that a lot of 320kbps mp3 files have. A lot of 320 webrips have a shelf at 16khz and some are 'true' 320 which extends to 20.5 kbps and have a slight shelf at around 16. Lossless is the best way to have a peace of mind. I've seen so many different type of spectrums that even though it is 320kbps, encoding settings change everything and FLAC is the way to go (if you really care). Personally, I just collect whatever I can find.

 

[hex182]

   
Why would I get lossless if I can't tell the difference, it takes 10x more space, and costs much more?

For the reason I stated above, piece of mind. Space, for most people, is not an issue as for the cost, I said if you can. I personally find that sometimes there is a difference between 320kbps and lossless, even if some of times it's just the mastering and sometimes there isn't.

 

[willowbrook]

 
For the reason I stated above, piece of mind. Space, for most people, is not an issue as for the cost, I said if you can. I personally find that sometimes there is a difference between 320kbps and lossless, even if some of times it's just the mastering and sometimes there isn't.

Peace of mind ^^

 

[hex182]

  Peace of mind ^^

My bad, at least I didn't write piss )

 

[Metalomaniac]

I received my FLC 8s a few weeks ago, and repeated the tests. Apart from IEMs, rest of my gear is unchanged. I used Simple ABX Tester Android app for the tests.
 
I could tell the difference between a FLAC and the same file re-encoded to 96kbps CBR MP3 quite easily. I didn't try higher CBR files as I don't think you can even find CBR files these days, and I did this test just to see if I'm totally clueless.
 
When I moved up to LAME V6 at around 120kbps VBR, I was already at random success rate.
 
I repeated the tests with the ABX High Fidelity site, which supported the findings above.
 
The two main purposes of this test were to find out if I should bother with expensive hi-res files, and which quality level on Spotify would work best for me. Clearly there's no point in spending extra on the hi-res files for me, and Extreme setting is way overkill. I can fit a lot more music on my Walkman using the lower quality setting in Spotify.
 
So rather then spending (wasting in my case) money on hi-res files, I will try to find recordings which have not been ruined by loudness wars. I'm guessing in some cases these are hi-res file if the masters are different from the CD or streaming. How to go about all that I'm not so sure...