I upgraded to 2.03b two month ago and my first thought was:
"That sounds bad."

I spent this weekend switching back and forth between 2.xb and 1.x and 2.x really sounds worse.

But the bigger problem is:
2.x won't play some songs at all.

I transfered some of the songs back (via netcat) and they play perfectly on all my other players. (xmms, mpg321, NexII, DVD-Player).

Did anyone else notice these problems ?

And why did they change the decoder at all ?!

Are you sure your testing in 2.0b3 didn't come by the known-bug with normalizing? (Search the BBS - it will sound like crackling, popping or hissing, etc. It can sound one way in one file and much worse in another) And you're sure you've tried b7?

Other than that, start getting more specific about your problem files. 2.0b7 is no worse "sounding" than 1.03 was for me and it supports my VBR files properly (ok, reverse seek farther than the start of a song still skips a track).

Bruno

I'm talking about not playing at all.
Try this http://www.webhome.de/anti/files/av-fy.mp3, first 200k of "Forever Young", "low" quality encoded.
I have around one dozend of those (scattered all over the disk.) and if the player shuffles into one of them while playing at 0db you are in serious trouble.

And even more are "unlistenable" (try "Suicide Commando", but that's hard stuff for most decoders).

I would switch back to 1.03, but I really like/want/need the 2.0bx features.

The decoder on the player should play back the files perfectly, except for some known bugs in 2.0 beta 3 where some songs would be played over-clipped and sound like they had static. This is a known bug that was fixed in 2.0beta7. Please do not report or complain about bugs in 2.0beta3.

Either upgrade to 2.0 beta 7 or downgrade to 1.03. After you have done that, see if the problem is still there. If it is still there, please check the file itself by playing it back on the PC. If it sounds good on the PC but bad on the player, please describe in detail exactly what is wrong with the sound and we'll see if we can figure it out.

File decoding will depend on how the files were encoded to begin with.

As I mentioned before, give some details of the encoder and the resulting MP3 file.

Bruno

FWIW, both playback bugs I complained about in this thread are alive and well in 2.0b7...

That Alphaville track plays back fine on my player running 2.0b7. Obviously it cuts short, but that's because you've truncated it at 200KB.

Bruno

I don't consider the inability to play a mangled or corrupt song a bug. Sure, the decoder may not be able to elegantly skip through garbage MP3s, but when you have a properly encoded piece, it seems to play fine.

The second example you gave of the mp3.com cd sampler is a more interesting one. Were you ever able to break down what elements of the song were the cause of the problem? You may want to scan the file with the newest versions of the tag and repair tools that are out there.

I'm also assuming that you wrote to the bugs email address at empeg and offered them the files in addition to posting to the BBS, right?

Bruno

I thought the first example (the one which switches to warp 9) would be interesting, though the file *is* mangled. The second example passed all tests by various utilities I tried on it two or three months ago. It is just one of several files from mp3.com that make decoder choke in the same way.

Hm, I think I e-mailed to bugs address, too. Anyway, I can do that now...

Well...

1. I wasn't complaining. I just want to know if someone had similar experiences (with mp3s not playing at all).
2. The Alphaville song is not playing on the empeg.
3. It's playing on all other mp3 players (I even tested winamp now)

The encoder was lame (3.91) with abr set to 128 (low quality).

Question:
Since it (still) playes on 1.03 and not on 2.0bx (x=3 or 7),
I assume the decoder was switched.

All I want to know:
Why?
And will this be fixed ?

ps:
Don't get me wrong, I still love my empeg and think the support is great!!!
But if nobody knows it's broken, you can't fix it.

pps:
If the (above) song is broken, let me know. All tools say it's OK.

ppps:
Maybe a broken mp3 should just be stopped instead of playing
very loud mixed frequency noise.

(I even tested winamp now)

Yeah, but Winamp will play ANYTHING! I noticed that a file would simply NOT play on MusicMatch but MusicMatch would allow me to edit the id3 tag.... But Winamp WOULD play the song... ends up, the file was an mp3 with a .wav file extension. Simply renaming the file (a sound effect I downloaded in this case) took care of it. But I found it funny that Winamp will play just about anything. In early versions of Winamp, you could even drag .txt files and hear what they sounded like! (kinda neat!)

Sounds like my standard method for determining that my sound device is working properly: ``cat /bin/sh > /dev/audio'' while positioning my hand over Ctrl-C.

You know you're truly a programming god when you can pick out the bugs by ear.

Calvin

Actually, that reminds me. There's a network sniffer that comes with the standard Solaris installation called ``snoop''. Snoop supports the `-a' option, which, according to its man page, does the following:

Listen to packets on /dev/audio (warning: can be noisy).

I've never been quite sure of the utility of this option, but there you go.

<network engineer hat>
I think a better adaptation of the concept would be to categorize the type of packets, and associate them with certain sounds. That way, you can *listen* for http requests/response packets, you can *listen* for ftp and you can *listen* for database activity. For example. Why would you want to do that? Well, networks being crowded as they are, sometimes it is sufficient to know the *amount* of activity. Traditional sniffers have you looking and poring over packet activity logs which is well, plain boring and time consuming.
</network engineer hat>

Calvin

2. The Alphaville song is not playing on the empeg.

I thought I already told you plainly that the Alphaville song you posted had no problems playing under 2.0b7 on Mk2a hardware?

I think at this point you should start suspecting yout specific player for problems - or the data that is on your player. Using emplode 2.0b7's ability to DOWNLOAD, pull the offending tracks back to your local hard drive and then play them back. Also look at them with MP3 Tag Studio. Next, put them up for others to try.

Bruno

I don't know about other operating systems, but Skwonk! does this for Mac OS X. Pretty nifty!

Ah...you want the Peep Network Auralizer. This really can give you that Enterprise bridge feeling.
And it's useful, or something
This is only useful for Unix junkies, but then again, who else would want it?

Sorry for answering this late, but I was very busy...

In reply to:

---

I thought I already told you plainly that the Alphaville song you posted had no problems playing under 2.0b7 on Mk2a hardware?

---

And I told you that they didn't play on my model with 2.0b7.
Strange.

In reply to:

---

Using emplode 2.0b7's ability to DOWNLOAD, pull the offending tracks back to your local hard drive and then play them back.

---

I used netcat to pull the file of the player.
The checksum was still the same as the original (I have all of them on DVD.)
All player played them.

In reply to:

---

Also look at them with MP3 Tag Studio.

---

I used different mp3-check tools and none complained...

In reply to:

---

Next, put them up for others to try.

---

That's why I put up that 200k of Forever Young.
And it worked on others system.

However.
I just installed 2.0b11 and it fixed all my problems.
Well, I still "feel" that the quality is a little bit degraded, compared to 1.03, but that's only in the lab and unnoticeable in the car.

Problem solved.
And to tell you the truth,
I don't really care what fixed it.
As long as it stays fixed.

BTW:
In the whole process I never touched that "broken" Alphaville song,
it's still the same I uploaded 14month ago.

anti

ps:
I'm sorry about the harsh tone I hit in my earlier posts;
And I apologize for this.

The tracks problems still have me perplexed. No problem about the tone. It just felt like you were ignoring the information others were posting trying to help.

I ran 1.03 for a while and didn't notice any degradation of sound/musical quality when moving to 2.0b3 (and subsequent builds). If you look through the change logs for 2.0b11 you'll see there's nothing mentioned that would explain why the quality has now improved for you either.

Definitely keep an ear open for any strange qualities while running b11 and when updating to the next beta. Then post an update in here.

Bruno

Honestly, I jumped directly from 1.03 to 2.0b7 and I noticed a quality difference as well. I think that it might have been a difference in the EQ. Mine were all flat at the time, and maybe the response was somehow different. But I definitely noticed it.

Ditto here, though I've never gone back to double-check it. v2betaXX does sound as good as I remember v1.03 sounding. Probably just the alzheimers kicking in, but..

You guys are obviously not playing your music loud enough.

The decoding did change, but with my files (lame 3.90 r3mix-like encodes of perfect rips) I coulnd't tell the difference between player software versions during casual listening at home or in the car.

I'm tempted to flash everything back down to 1.03 just to test this out... But not that tempted.

And.... (Starting sentence in this way just for Bitt) one of Anti's big problems was apparently one of gross-error on decode of certain tracks. I've never had this happen. And his partial song clip played fine in B7 for me (as mentioned way back when).

Bruno

My observations were totally subjective, but I don't think I would just be making them up. I didn't expect any change at all, and I never revved back to check, but I could swear that there was a quality change. (Not quality necessarily as in good or bad, but more as in timbre. I thought it sounded worse at the time, but that might just be due to unexpected change.)

And not only did you start a sentence with a conjunction, but the whole paragraph. (I actually don't have a big problem with that one.) I've never seen any egregious errors like Anti describes, either.

I'm reasonably convinced that there is a problem in the players eq code where channel lock appears to set the 2 channels differently. I want to sit on the fence until I can get a white noise input and a scope on the output, but looking at the numbers that the player sends for the frequency and q factor suggests that there is a bug, and someone posted a rather bizarre graph recently that would appear to suggest the same.

If you keep to flat eqs then there wont be any problem as the gain bits are kept in sync.

I'll keep y'all posted as to anything I find out.

This is an interesting piece of info.

ANTI, hopefully you're still following this thread. Did you have any custom EQ settings? (Can't go back to check the original message right now - mostly out of laziness )

Bruno