Skarecrow, you know your isht! That was a great post.
 
Ypok, I'm sold!  I'm going to buy some more expensive phones in the near future.  

Quote:

peskypesky said:

Skarecrow, you know your isht! That was a great post.
 
Ypok, I'm sold!  I'm going to buy some more expensive phones in the near future.  

Click to expand...

Thanks, I try. I've been into mp3 since the voodoo days back when we had to do everything by command line and you basically had 2 encoders to choose from and they both sucked. I haven't been paying as much attention lately since, in all fairness I use FLAC exclusively on my home network simply because I have the space for it and it makes later transcoding easier, with aoTuV vorbis on my portable, but I take a look at the state of lame about once a year and I haven't seen many major advances since 3.97, which in itself wasn't that big a step over 3.90.3, the previous standard. LAME has been very very good for a very long time now (Well, comparatively speaking). So long as whoever encoded your music (if it wasn't you) did so with with a decent bitrate (192kbps+) on some modern version of LAME and they used good ripping habits, I have no problem recommending headphone upgrades.
 
Today's Mp3 has sort of the same reputation as domestic (USA) automobiles. They are currently very fine products, as good as any of the competition, trying to live down a bad reputation they (deservedly) acquired years ago. in 1998, 128kbps Mp3s really WERE horrible, but the rub is that high bitrate mp3s were horrible too since the rippers didn't error correct and the encoders didn't have a good model to properly use the extra bits where they'd matter. I don't know where people got the idea that high bitrate automatically makes something better. Today's 128kbps mp3s probably sound better than the 256/320kbps mp3s of the late 90s. Today's high bitrate stuff (now that the models and tweaks/tuning are better) are just that much better. And the REAL benefit is that even when you can hear an audible difference from the source, what you're hearing isn't necessarily bad... just different. It may not be accurate, but it won't usually dull your enjoyment like it once would.
 
as a matter of fact I remember reading an article a year or two ago that said that in double blind testing, the results showed that in many instances today's teens and young adults actually prefer the mp3 versions of music because what few compression artifacts that do exist, they're accustomed to.

 
Quote:

eric0531 said:

While I understand where you are coming from, if the current benchmark for digital audio is the CD why would you want to accept inferior lossy formats when you buy music downloads?  Almost all of my music listening is with MP3 files (either LAME VBR -0 or fixed at 320) and I've done enough double-blind tests to be comfortable that I can't tell the difference between those and lossless on my system.  But I still want to have the full lossless version of the file archived, because perhaps someday I'll have a good enough system that I could tell the difference.  Like others have said, storage is cheap.  I don't want downloads to *not* be available in MP3 format, I would just like to have lossless available since that's the format I want to download whenever possible.  
 

Click to expand...

I think the fault with your argument lies in the first sentence, and specifically the word "inferior". I don't consider high bitrate MP3 inferior in any way. I just have no problem with it and don't believe I ever will have a problem with it, for archiving or otherwise, just as I've never had a problem with minidisc. You mention a CD standard, and in my view that's often where the problem lies, not in the format but the implementation; there are still too many indifferent sounding CDs. If I had a chance to improve any part of the chain, it would be the training of audio engineers (or should that be acoustical architects?).
 
However, I do have one problem with MP3, and that's the fact that it inserts a pause wherever it detects a new track, even when there may be no pause in the music. Don't know about popular, but in classical music there are numerous works which have track markings but no actual pauses (one movement symphonies, for instance). Can't get around this problem, and not from lack of trying.        
 
 
 

Quote:

pp312 said:

 

However, I do have one problem with MP3, and that's the fact that it inserts a pause wherever it detects a new track, even when there may be no pause in the music. Don't know about popular, but in classical music there are numerous works which have track markings but no actual pauses (one movement symphonies, for instance). Can't get around this problem, and not from lack of trying.        
 
 
 

Click to expand...

 
Well, you can. sorta. sometimes.
 
Some modern encoders (LAME specifically comes to mind) embed offset information for that exact problem. Basically the long and short of it is that when you're encoding an MP3, there are specific block sizes, and if your song ends in the middle of a block, the rest is padded with 0s, basically silent information. It shows up as a brief pause of silence on track changes. Modern encoders detect when this happens (which is most of the time actually), and embed a big warning flag that is supposed to signal to the software/hardware playing the mp3 "hey, blank info due to a partially filled block right here! it is Xms long, just skip over it and get to the next track!". The problem is that since this isn't written into the mpeg-2 level 3 iso standard, there is no reason it needs to be implemented by the authors of playback software/hardware unless they feel like it, which unfortunately most of them don't seem to.
 
They do exist though. Try playing back a performance/live cd encoded by lame on foobar2000 some time. I think winamp supports it too. I'd be floored if itunes supports it, but then I haven't used itunes since like 2004 anyway so I don't know what it does and does not do anymore.

Thanks, Skarecrow. Though it's very interesting reading about the dialogue that takes place between different encoders and software, I was really looking for a solution for when I transfer my music files to a media player for playing elsewhere.

Quote:

pp312 said:

 
However, I do have one problem with MP3, and that's the fact that it inserts a pause wherever it detects a new track, even when there may be no pause in the music. Don't know about popular, but in classical music there are numerous works which have track markings but no actual pauses (one movement symphonies, for instance). Can't get around this problem, and not from lack of trying.        
 

Click to expand...

The inability of a lot of players to deal with this is a real pet peeve of mine.  My first DAP was a Rio Karma which did support gapless MP3 playback, something that was rare at the time.  I would never, ever even consider buying a DAP that could not play MP3 as gapless (I don't do a lot of classical, but live concerts are a music staple of mine and nothing ruins the mood more than a clunky gap either between songs or at a transition point where songs should flow seamlessly from one to another).  
 

I'm currently listening to MP3 320 kbps on some flat studio grade speakers in an acoustically dampened room with no external noise sitting right in the sweet spot using a marantz av receiver with a 32/192 DAC, 108dB SNR, .008% THD  and i still cannot hear the difference...... if i could hear one it would definitely be with this setup, it brings EVERY imperfection out. DBT is the science i am referring to, nobody on the face of this earth in a DBT setting has been able to tell a difference between high bitrate mp3's and lossless codecs.

i made a thread about this awhile back showing a frequency analysis graph which proved this point. MP3's at 256kbps are reconstructed perfect with the only difference showing up on the graph that anything above 20khz is deleted, which makes no difference because we cannot hear higher then 20khz. at 320kbps, the information above 20khz is still intact.
Quote:

tws1 said:

After the MP3 has been through its reconstruction filters, etc, and with the complexity of the modern MP3 encoders, I wouldn't worry about FLAC vs. MP3, as they are going to be extremely similar. After all, assuming that the FLAC has been created from a 44.1kHz CD the extra accuracy is the detail of the wave every 441000th of a second. We can only hear the difference in lower quality MP3's because the 16/24bit value is limited by the xxxxkpbs, so the encoder compresses as best it can whilst maintaining as much audio quality. Noticed how VBR MP3's bitrate drops up and down, on 'simple' segments of the track, the bitrate falls to meet the V0, size, etc, contstraints.
 
Anyway...
 
Spend the money on better amplifing the headphone, speaker, etc. The better you can reproduce the reformed MP3 the better it will sound, regardless of the bitrate. Think of listening to FLAC through Apple crappy buds, vs. a 128k MP3 through Grado's..... an extreme situation yes, but at least proves where the easiest and most dramatic improvement can be made.
 
OP, in a word, YES, upgrade up to your point of diminishing returns!

Click to expand...

 

Quote:

eric0531 said:

The inability of a lot of players to deal with this is a real pet peeve of mine.  My first DAP was a Rio Karma which did support gapless MP3 playback, something that was rare at the time.  I would never, ever even consider buying a DAP that could not play MP3 as gapless (I don't do a lot of classical, but live concerts are a music staple of mine and nothing ruins the mood more than a clunky gap either between songs or at a transition point where songs should flow seamlessly from one to another).  
 

Click to expand...

hmm. with itunes you can just tell it to fade from one track to another....
 

Quote:

yepimonfire said:

i made a thread about this awhile back showing a frequency analysis graph which proved this point. MP3's at 256kbps are reconstructed perfect with the only difference showing up on the graph that anything above 20khz is deleted, which makes no difference because we cannot hear higher then 20khz. at 320kbps, the information above 20khz is still intact.


 

Click to expand...

What you are describing is lossless encoding. Mp3, pretty much by definition, does not do this.
 
Do some research on audio encoding, as well as compression technology. I mean REAL research. there is a reason that the maximum compression you can get out of FLAC is around 50% under most circumstances.
 
and as for killer samples that no bitrate MP3 can restore properly? a quick google search gook me to: http://www.hydrogenaudio.org/forums/index.php?showtopic=69454 which even the newest (at the time of posting) LAME 3.98 at 320kbps was still easily ABX'd (I believe the OP got 7/8).
 
MP3 can not restore full audio fidelity no matter what bitrate you use. Please stop advocating that it can, because you are completely missing its true worth, which is that while it can not restore true audio fidelity, it can do an amazingly good job of tricking the brain into thinking it can the vast majority of the time.
 
Mp3 compression, hell all lossy compression, DISCARDS DATA. this data is gone. you aren't getting it back. what you are getting back is a reconstruction that may, if the encoder is well designed, convincingly simulate the original data. lossless compression is the only method by which you can recreate the original data because the data was never discarded, only compressed.

oh really? do some real research? do you even know WHO you are saying that too????

Quote:

yepimonfire said:

oh really? do some real research? do you even know WHO you are saying that too????

Click to expand...

 

Well considering that you were comparing frequency response graphs as proof of identical reproduction (which is laughable), claiming that the difference between 256kbps and 320kbps mp3 was the lack of post-20khz signals in the former (when highpass filter is a variable you can set on any mp3 bitrate), and claiming that nobody can hear the difference between 320kbps mp3 and lossless/source, when that is demonstratably untrue (if quite rare)... I'm pretty confidant that it doesn't matter who I'm talking to. The fact remains that you have not done your homework and it shows.

http://www.mediafire.com/?16r1f7zj3mtej35
 
what this is is a series of sine waves at various frequencies from 100hz to 20000hz. all at the same exact volume level. the inversion test file is and mp3 file and a wave file loaded into a sound editor, one is inverted so that any similarites are cancelled out. what remains is the differences between the files. the wav noise file is the noise floor of the two waves inverted to completely cancel eachother out and the volume of it was boosted about 150dB to be audible. the noise floor of the mp3 file is the same thing but the wav and mp3 inversion rather then wav on wav inversion also amplified, about 100 dB. as you can see the only difference between these files is the noise floor. other then that they both faithfully reproduce the original sine waves.
 
 
i will do some other test tonight, including one using actual music.

first off you are completely incorrect here because i did not compare 320 to 256 i compared 256 against a wav.   laugh at yourself.
 
Quote:

skarecrow77 said:

 

Well considering that you were comparing frequency response graphs as proof of identical reproduction (which is laughable), claiming that the difference between 256kbps and 320kbps mp3 was the lack of post-20khz signals in the former (when highpass filter is a variable you can set on any mp3 bitrate), and claiming that nobody can hear the difference between 320kbps mp3 and lossless/source, when that is demonstratably untrue (if quite rare)... I'm pretty confidant that it doesn't matter who I'm talking to. The fact remains that you have not done your homework and it shows.

Click to expand...

 

my head hurts.