[Lord Chaos]

The difference I hear between compressed and uncompressed files isn't so much a matter of identifiable artifacts, as a matter of something missing. All of the music is there to a first approximation: cymbals, bass, voice, instruments... but it's as if everything is faded rather than painted. The sparkle is gone. I don't know how to describe it better than that.

 

[Cid]

Quote:

Originally Posted by EnOYiN /img/forum/go_quote.gif Funny how this thread does not evolve but just repeats itself.

Yeah, I know. I look at it this way.

People who use MP3 are posting here saying it's good enough and that anyone using lossless is just having the placebo effect.

People using lossless are posting saying anything less isn't good enough for their needs.

There is too many different opinions in this thread to result in anything remotely useful. ABX results are probably still not going to change other peoples opinions on formats, because at the end of the day it's the person listening who cares and if he/she can't tell a difference and you can, it won't change his choice on format.

I honestly hate threads like this because it's just the same every time. Some people debate lossless is the superior way to go, some debate MP3 is the way to go. It's just a complete clash of opinions and like I said before it's not resulting in anything useful.

I wish the OP would close the thread, because this isn't going anywhere and I don't see it changing.

 

[lordgibbness]

Quote:

Originally Posted by Lord Chaos /img/forum/go_quote.gif The difference I hear between compressed and uncompressed files isn't so much a matter of identifiable artifacts, as a matter of something missing. All of the music is there to a first approximation: cymbals, bass, voice, instruments... but it's as if everything is faded rather than painted. The sparkle is gone. I don't know how to describe it better than that.

Sorry to be pedantic, but ABX results?

 

[lordgibbness]

Quote:

Originally Posted by Cid /img/forum/go_quote.gif There is too many different opinions in this thread to result in anything remotely useful. ABX results are probably still not going to change other peoples opinions on formats, because at the end of the day it's the person listening who cares and if he/she can't tell a difference and you can, it won't change his choice on format.

All true. But as far as I can tell everyone who says they can hear a difference never ever produce ABX results. I mean, I KNOW that lossless is the best format quality wise because the sound has not been altered in any way, but that doesn't mean that you will hear the difference! Do the same people who can hear the difference betwen lossless and lossy also say they can hear the difference between CD, SA-CD and DVD-A ? I mean you do realise that CD is ONLY 44.1KHz with a 16-bit depth don't you!? Surely that can't sound as good as 192KHz with a depth of 24-bits? I must be able to hear the difference, right?

I'd like to add I am not against lossless in any way. It's a brilliant archiving tool and great for transcoding to different formats, but it's just silly when people like to chime in that they're ears are so good - "ooh, my golden ears".

 

[immtbiker]

Quote:

Originally Posted by lordgibbness /img/forum/go_quote.gif ....Do the same people who can hear the difference betwen lossless and lossy also say they can hear the difference between CD, SA-CD and DVD-A ? I mean you do realise that CD is ONLY 44.1KHz with a 16-bit depth don't you!? Surely that can't sound as good as 192KHz with a depth of 24-bits? I must be able to hear the difference, right? ...but it's just silly when people like to chime in that they're ears are so good - "ooh, my golden ears".

I'm a little confused by this statement. Are you saying that you can't hear the difference between SACD/CD/DVD-Audio or are you being sarcastic?

If you use a SACD player that outputs in true DSD rather than PCM, you are hear 50 times more information, and to me, the difference is huge. Everytime the analog curve goes from top to bottom to top, sampling 16 times compared to 24 times, provided you have a player that is not restricted to 16 or 18 bit, has got to yield inferior results.
When I listen to any .mp3 at a bitrate of 128kbps there is a 2 dimensional flat soundstage that has no depth. Take the same song at 360kbps and the whole soundstage blossoms. To me it's like the difference between am and fm and cd.

And on the note of saying that this thread is going in circles and not worth the discussion, basically is saying that at least half the threads at Head-Fi or any forum for that matter, are worthless. Isn't this discussion what it's all about? Varied opinions.

Also posting in a thread that a poster believes should be closed actually perpetuates the thread, doesn't it? How ironic.

 

[Thelonious Monk]

i can't tell the difference between -v0 VBR and Lossless... at all. maybe this will change when i get a new DAC, but we'll see.

also, this is coming from a guy who can here the difference between 176.4khz and 192khz. the former is better to my ears...

 

[lordgibbness]

Quote:

Originally Posted by immtbiker /img/forum/go_quote.gif I'm a little confused by this statement. Are you saying that you can't hear the difference between SACD/CD/DVD-Audio or are you being sarcastic? If you use a SACD player that outputs in true DSD rather than PCM, you are hear 50 times more information, and to me, the difference is huge. Everytime the analog curve goes from top to bottom to top, sampling 16 times compared to 24 times, provided you have a player that is not restricted to 16 or 18 bit, has got to yield inferior results. When I listen to any .mp3 at a bitrate of 128kbps there is a 2 dimensional flat soundstage that has no depth. Take the same song at 360kbps and the whole soundstage blossoms. To me it's like the difference between am and fm and cd.

Just because there is more data doesn't mean you will hear it. The human ear is not that good! If they came up with a technology that sampled at 768KHz and a bit-depth of 1024bits you would not be able to hear the difference! Or maybe you would once you've spent $10,000 making yourself believe it.

Also the bit-depth is not the sample rate - it represents the dynamic range of a recording, but I'm sure that's what you mean, right?

And, yes 128kbps MP3s are not really of a good enough quality to compare to lossless, which is why the LAME presets were conceived. Have you tried transcoding your FLACs to LAME q0 and ABX-ing? With results?

 

[immtbiker]

I can understand that most humans can't hear over 14khtz so getting a Kappa tweeter that goes to 40k doesn't really help the music sound any better, but I know I can hear the difference between a redbook and DSD recording.
Without sounding too naive, I'm not hip to LAME q0, but I'm willing to read up on it and give it a try.
Does anyone have any links that they think would be really helpful in explaining all of the different trancoders and the software required? I'm very interested, but it seems the older a person gets, the harder it is to keep up with the latest and greatest.

 

[Rempert]

When it comes to mp3 encoding, the best starting point is here:

http://wiki.hydrogenaudio.org/index.php?title=LAME

 

[andyshedd]

320 kbs mp3 sounds slightly drier and softer than uncomressed. Not a completely bad effect, and less noticable (obviously) when listening on-the-go.

 

[BigSurSpoon]

Ok, here's my story.
The first 'high end' headphone I ever bought was the Grado SR-80. I bought it because I 'thought' that I could hear a huge difference. Then for jogging, I bought the senn pmx-200. At that time, I couldn't tell the difference between 128kbbs mp3 on the senns and ALAC/FLAC on the grados. I simply couldn't tell the difference. I kept telling myself that it was there to justify my purchase, but I couldn't hear it. But the longer I used the Grados, the more I could start to hear subtle differences. I then bought the shure e3c's and I could slightly hear the difference between them and the Grados. But still at this point, I couldn't tell the difference between mp3 and lossless. But after a couple of months of using the shures to play lossless (I put the lossless on my ipod due to placebo), I listened to some 320kbps mp3 lame 3.96, and I realized that the mp3 sounded slightly inferior. I have since upgraded my IEM to the UE super.fi 5 pro, and the difference between my ALAC and mp3 is quite dramatic to me. The mp3s sounded rougher, less detailed, and there was more sibilance. (I also started to notice the tighter bass that my total bithead supplied).

Remember, this is simply my story and may not apply to anyone else. I have just realized that all I needed was to train my ear to hear the differences between the lossy and lossless files. This may be a gift or a curse since I cannot fit all of my music on my iPod, and I'm likely to spend much more money on headphones and amps in the future. But in the end, I have realized that my ears have simply adapted to the higher end, but I cannot listen to low bitrate files on my UEs anymore, because I can't stand the sound quality.

But that's just me.

 

[doofus]

Headphones: AudioTechnica ATH-AD700
Source: ASUS A8N-SLI Motherboard On-Board Sound
Music: Various tracks from the Ultrasone CD (Lossless)
Player: foobar2000 0.9.4.2
Playback Method: ASIO plugin for foobar2000 (I've also tried kernel streaming and the normal playback method)
Drivers Used: nVidia NVMixer Audio
MP3 Codec: LAME 3.98 alpha build 11 @ 115kbps/V6/fast


I can hear changes in quality right up to 115kbps but don't notice any difference between 115kbps and lossless. Does anyone have any idea on what part of my test system might be affecting the results? The source is horrible and normally hisses a little bit when idle (background noise). The nVidia NVMixer drivers are loaded and are used in the process, which are probably decreasing quality a lot in the process.

Would it help to burn off a CD with a track in 85kbps, 115kbps and lossless (converted back to CDDA) and try this on an old Luxman amplifier (much better quality than this 'new PC technology' - and it is 20+ years old!)?

Also, does the listening volume play a role in this test? Without a sound level meter, is there some sort of rough way to get a rough "standard headphone volume" I could work from?

 

[EnOYiN]

Quote:

Originally Posted by immtbiker /img/forum/go_quote.gif I can understand that most humans can't hear over 14khtz so getting a Kappa tweeter that goes to 40k doesn't really help the music sound any better, but I know I can hear the difference between a redbook and DSD recording. Without sounding too naive, I'm not hip to LAME q0, but I'm willing to read up on it and give it a try. Does anyone have any links that they think would be really helpful in explaining all of the different trancoders and the software required? I'm very interested, but it seems the older a person gets, the harder it is to keep up with the latest and greatest.

I posted these links before in this thread, but I suppose I can do it again.

Here they are:

http://ff123.net/index.html

http://ff123.net/training/training.html

This site will explain a lot of things about mp3 encoding and the flaws in that. The second one will explain artifacts in mp3. HydroGen Audio will provide enough information about this subject as well.

Edit: another link; the lame wiki from hydrogen
http://wiki.hydrogenaudio.org/index.php?title=LAME

 

[EnOYiN]

Double post for some reason. Sorry for that.

EditII:
Quote:

Have you tried transcoding your FLACs to LAME q0 and ABX-ing? With results?

Yup. It's in this thread actualy. If you read it I am sure you will find it.

Quote:

Would it help to burn off a CD with a track in 85kbps, 115kbps and lossless (converted back to CDDA) and try this on an old Luxman amplifier (much better quality than this 'new PC technology' - and it is 20+ years old!)?

I doubt it. The new PCs will most likely have a better DAC than something older.

Quote:

Also, does the listening volume play a role in this test? Without a sound level meter, is there some sort of rough way to get a rough "standard headphone volume" I could work from?

For a ABX test you should apply a gain. Replaygain or mp3gain will do this for you. Mp3 will always have a different volume (louder) than a file encoded with FLAC.

 

[Rempert]

Quote:

Originally Posted by doofus /img/forum/go_quote.gif Headphones: AudioTechnica ATH-AD700 Source: ASUS A8N-SLI Motherboard On-Board Sound Music: Various tracks from the Ultrasone CD (Lossless) Player: foobar2000 0.9.4.2 Playback Method: ASIO plugin for foobar2000 (I've also tried kernel streaming and the normal playback method) Drivers Used: nVidia NVMixer Audio MP3 Codec: LAME 3.98 alpha build 11 @ 115kbps/V6/fast I can hear changes in quality right up to 115kbps but don't notice any difference between 115kbps and lossless. Does anyone have any idea on what part of my test system might be affecting the results? The source is horrible and normally hisses a little bit when idle (background noise). The nVidia NVMixer drivers are loaded and are used in the process, which are probably decreasing quality a lot in the process.

This is a bit confusing, because 115 kbps suggests CBR, whereas -V6 is VBR. A 115 kbps CBR encoding will usually be a lot easier to ABX than a -V6 one, even if the -V6 file is smaller.

On-board sound is not ideal, but to be honest it may not matter. -V6 is very easy to ABX on some songs, but on many songs it is darned hard.