Excellent.. I really do have tinears

[swalker]

So there I was a while ago, probably 6 months ago or so, listening to a Internet radio station being streamed in the ogg vorbis format. I was thinking how nice it sounded yet I thought it didn't have the same fine details of the ape version of the same song I had on my harddisk. Later that day I happened to be talking to the guy broadcasting and he mentioned that he would happily live with ogg vorbis files at the Q4 setting. Of course I felt a little snobby inside and told him I couldn't do that and blah but I did feel curious about my own ability to discern between files of different quality.

At that time I was going nuts with stripping all the CD's I had in APE and was thinking about buying a decent dedicated player. I was pretty sure I could tell the difference between, say, a Q6 ogg and a Q10 ogg file (or 192kbps and 320kbps in MP3). So with a degree of confidence I proceeded to rip a few of most often-played tracks in my collection and encoded them in various formats and quality settings. After making sure there is no noticable artifacts in the tracks and checking for quality differences, I asked my friends to convert them to wav and label them without letting me know. Soon I began to play all the tracks with a note on my hand.

Turned out, I couldn't really tell difference with any of those tracks with exception of a few low quality Mp3 (64kbps) and ogg (Q3) files and finding the q3 ogg vorbis file was actually a bit tougher than I expected. It surprised me a lot as I could swear I perceived the differences when I listened to them while knowing their formats and quality settings.

Since this could mean my system simply was not good enough to resolve the differences, I burnt the Wav files as an audio CD and brought it along with me when NotoriousBig graciously allowed me to audition his nice setup. I sat on the corner of the bed and listened intensively through 3 different pairs of headphones - my HD650's and ER4's as well as Biggie's RS1's which are three of better pairs of headphones discussed here in head-fi as far as I know. I listened and listened and...well I still couldn't tell any difference between q6 and q10 nor could I tell the difference between q10 and the original wav file.

I'm rather embarrassed to say that I couldn't tell any difference but having tin ears means 1)I saved tons of harddisk earlier today by converting all the music files into Ogg Q6 2)Somehow I don't think about my next audio upgrade as much. More money saved for food(yay)

So.. the moral of the story? I have proven to myself that I have tin ears

and it saved lots of harddisk space while saving me from the temptation to open up that little piggy bank to get a new CD player.

I remember there was an old saying the less you know -or dumber you are- the happier you get and I guess it works the same way in the audio world in some degree.

 

[pbirkett]

Excellent post, and for what its worth, you may or may not have tin ears, but one thing is for sure is I admire the honesty of your post. I cant personally tell the difference between MPC q6 and the original, and I happily admit I would struggle to tell lower quality settings too. I have tried this with various headphones and speakers.

The thing is, most audiophiles will claim they can tell the difference, but when confronted with a blind test, they will struggle to pick out the songs they are so comfortable at picking out in many cases.

At a bake off recently in my country, I compiled a CD with a 128K MP3, a 320K MP3 and a WAV, using music I knew they were relatively ok with, and on a system costing an RRP of about $8000.

To their disbelief, they really struggled. Some managed to pick out the 128's, but nobody picked out the 320's consistently. Yet in typical audiophile fashion, and behind the safety of a computer screen and hundreds of miles, they then went on a forum and claimed they could easily tell the difference, which was quite to the contrary to the results.

Its pathetic IMO, why can people just not admit compressed audio can sound good. Its their loss though, as I will get much more music on my hard disk than they will on theirs. Actually, I even knew some people who claimed to be able to tell APE files apart from the CD's and even more absurdely, WAV files. Unless something went seriously wrong somewhere on the line, I'd be asking how is this possible?

Quote:

I remember there was an old saying the less you know -or dumber you are- the happier you get and I guess it works the same way in the audio world in some degree

I may be alone, but I agree with this. I know if I had the best system that it would dictate what I listen to and would make me focus more on bad recordings, and they would be neglected. I've been there, I know, but thats just me FWIW.

 

[rauer]

I'm Team Tinear, also. There was recently a Hydrogen audio related test of lossy 128 kbps codecs. I did not go through the whole ABX procedure but briefly compared many different samples blindly. Frankly, I couldn't tell any apart.

Maybe I'll begin downgrading before I've really started the upgrading.

 

[adhoc]

Quote:

Originally Posted by swalker I remember there was an old saying the less you know -or dumber you are- the happier you get and I guess it works the same way in the audio world in some degree.

ignorance is bliss?

and great post too pbirkett - i agree with you fully. there once was this thread over in portable audio where people were claiming to be able to hear the difference between lossless and HQ lossy on an ipod, and i just went

.

 

[Distroyed]

The type of music makes the difference here. With, say trance or psy, I can't tell the difference at all. But with an ambient piano solo (Harold Budd, Brian Eno, et al), the difference is actually rather obvious. The quieter the music is, the more the nature of the compression is brought out.... which is why I distrust vbr rips. They got it backwards.

 

[adhoc]

Quote:

Originally Posted by Distroyed The quieter the music is, the more the nature of the compression is brought out.... which is why I distrust vbr rips. They got it backwards.

i'm sorry, but afaik the vbr algorithm does not look to volume to determine how much bits to allocate an audio frame - it looks at the relative complexity of the audio signal.

 

[noir]

my 2 cents:

well ripped and compressed 256kbit and 320kbit are virtually impossible to telll apart from the originals.

But i still keep getting flac/ape's only. I have some memories of badly encoded, and even worse ripped CDs... the mp3s just sounded awful.

Most Ape/Flacs i get are correctly done EAC rips, coming with the EAC log.. so at least i i know that i got a good rip.

This might sound crazy tho... but as i invested loads of money into a setup... i only want to feed it flawless source material... maybe its some self-contentiousness... but i am that way

 

[commando]

Good honest post swalker. Makes me wonder why I bother using FLAC when it's so large. Maybe i'll do my own blind test some time and see if it's worth the 15GB or so on my work HDD

 

[adhoc]

Quote:

Originally Posted by commando Good honest post swalker. Makes me wonder why I bother using FLAC when it's so large. Maybe i'll do my own blind test some time and see if it's worth the 15GB or so on my work HDD

my apologies for the OT post, but i just got a 160gb seagate to accomodate my flac collection!

 

[noir]

you're lucky, i nearly filled my 250gigs

 

[commando]

Quote:

Originally Posted by adhoc my apologies for the OT post, but i just got a 160gb seagate to accomodate my flac collection!

What's "OT" - off topic? That seems pretty on topic to me... and my comiserations on the loss to your wallet

 

[Sycraft]

Ya, I generally find that even on pretty stellar hardware 256k+ is CD quality, as in you can't reliably tell the difference. There actually was a listening test conducted to this end on reference grade hardware that found the same thing. I can't find a link though.

The interesting is that since both MP3 and OGG are 32-bit internally, you can encode from a 24-bit source and it will use the extra data, given a high enough bitrate. You can thus get files that are subjectively better than CD quality though smaller.

The one thing I will warn is that if you want to process the music, stick with lossless compression. The lossy stuff does rely on properties of human hearing so that which isn't perceptable can be made so with further processing. EQ is pretty safe but if you like playing with reverbs or such you are better off going lossless.

I personally go lossless since I have the space. Realistically there's no gain over 256k, and for most popular music 160k or less would be fine (given that it has no dynamic range thanks to WAAAAY over agressive use of peak limiters) but hey, means I won't have a problem in the future if I do decide to do some reprocessing.

I keep wating for the day when computers are fast enough and someone writes a program to do realtime, very high quality, room simulation so you can reprocess normal recordings to binaural on the fly. I do have software that does this, but it all has a problem in one way or another. Maybe I should just take an impulse in a listening setup I like and do it that way.

 

[Steve999]

....

 

[commando]

Some of you guys might like to join team diminishing returns

 

[adhoc]

Quote:

Originally Posted by noir you're lucky, i nearly filled my 250gigs

perhaps i should mention that i bought the 160gb to complement my existing 3x80gb western digitals, bringing the total storage in my primary computer to a whopping 400gb.