Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[RandyR]

I am having problems getting MC11 (build 1.68) to update the "Bitrate" and "Duration" tags in my MP3 files.  I have checked the appropriate boxes in Tools-->Options-->Library-->approriate tag fields, but it doesn't change them.

The program that originally placed the info in the file was incorrect as it did not deal with VBR MP3 files correctly.  The correct info shows in the MC libary view, but it isn't getting stored in the tag.

Help please.

Thanks.

[KingSparta]

MC Does Not Store The Bit Rate Or Duration In The Tags

And It Does Not Make Any Sence To Do So.

[RandyR]

King,

Given that incorrect information is in there at this point, how would you suggest I correct the info?

Bitrate is really for display purpose only, but I use some software to create mixes which relies on the duration numbers in the id3 tag to determine how many songs to put in the playlist.

Randy

BTW, if I rip using MC, the infomation is stored in the tags.  If MC doesn't do it, how does it get there?

Also, these are both editable tag fields in the library.  If it makes no sense to be able to save them, why can they be changed?

[KingSparta]

Here is a Listing of all the ID3v2 Frames

http://www.id3.org/frames.html

BitRate And Duration Is Not Part If Id3v2 Tags

That information is calculated as it is played by Analyzing the Mpeg Frames When A MP3 Starts To Play Or Is Imported Into A Taging Program.

Also There Is A Header (Not Really Part Of The Id3v2 Tags) Where This Info is Stored, And There Are Some Programs That Can Fix The VBR-headers Here Is One Of Them, And This One Works Good.

http://www.magnusbrading.com/mp3ts/

Also, these are both editable tag fields in the library.  If it makes no sense to be able to save them, why can they be changed?

That Is New To MC11 and That May Be A Bug, That Has Not Been Reported Yet. MC11 Should Not Allow You To Change This Data Or Even Attempt To Edit It.

[RandyR]

King,

Thanks for the reference to the program...I will check it out.

However, in the interest of closure, I want to point out the following:

1.  The field that MC refers to as "duration" is called "length" in the url you point me too...and even MC uses "length" as a synonym in searching.

2.  The field called "bitrate" is reported in tag&rename, Winamp, WMP, and several other programs after I rip something with MC...at the exact same exact value...and without any analysis at all by these reporting programs.  If MC is not storing this field...as you say...and these programs are not doing any analysis on the file themselves, how do they all get the same bitrate to report?

I don't mean to come off as argumentative or disrespective, but I am really not getting what appears to be an objection on principal on your part to what I am asking about.

All I want to be able to do is save the same information that MC appears (to me) to save when I use it to rip a CD - the same information that it reports to me in its library view.

Please help me understand why you seem to feel so strongly that this is unreasonable.

[GHammer]

Bitrate is not a thing to be stored in a tag.
If I have a 320 k MP3 and I transcode it to a 128 k MP3 for my portable why would I want to have it tagged as 320 k?

If you have broken rates in this or that program, use something that will fix them.
KingSparta gave you a link to one.

Depending on how large the files are you could use the free version of mp3Trim for this.

[RandyR]

Thanks Ghammer.

I'm really not that concerned about bitrate...its more that I want the lenght/duration correct so that programs that rely on this data to create playlists can use it.

Given that duration/length is reported and editable in MC, I was hoping to be able to store it back in the file without getting yet another piece of software.

But if that's objectionable...I'll just do something else.

[GHammer]

Programs do not look for a tag to determine length/duration.
That is a function of the bitrate and size of the file. Those two are readily determined and then a simple math function gets the length.

If the bitrate is off for some reason, then you get bad or no duration calculated.
Mainly happens with VBR files in my experience.

From a CBR File:
bitrate = 320
codec = MP3
channels = 2
samplerate = 44100
10327151 samples @ 44100Hz
File size: 9 373 547 bytes

From a VBR File:
enc_delay = 576
enc_padding = 1680
mp3_accurate_length = yes
bitrate = 237
codec = MP3
channels = 2
samplerate = 44100
16710960 samples @ 44100Hz
File size: 11 235 223 bytes

And another VBR:
enc_delay = 576
enc_padding = 1236
mp3_accurate_length = yes
bitrate = 252
codec = MP3
channels = 2
samplerate = 44100
13091820 samples @ 44100Hz
File size: 9 360 936 bytes

See the difference in the padding of the two? And the average bitrate? If a file does not have that info, apps will not be able to accurately determine duration.

[RandyR]

GHammer,

I do see.  But I'm not sure what to do about it.

I have approx 10,500 mp3 files which were encoded with VBR 320 by a lansoic unit using the Xing encoder.  I used this units because it controlls a 400 CD carousel and I was able to load up CDs and go away.  However, years later, I have dicsovered that it put 32 as the birate in virtually all the files. 

MC seems to be able to tell what birate and duration "should" be and displays it properly (I think).  I guess I am wondering why King seems to wants so much to prevent it from fixing the tag as well.

I any event, assuming he/JRiver don't relent on their position, short of re-ripping (which I really can't do for several reasons), how can I get the birate/duration fixed - preferably without buying even more software.

[KingSparta]

I guess I am wondering why King seems to wants so much to prevent it from fixing the tag as well.

MC does not save it as a tag it saves this info in the database

once you fix the vbr header the MC database may need to be updated and i am not sure if mc can do that correctly without deleting it from the library and re-importing.

MC will not fix VBR headers, that program i pointed you to a few days ago will. If you download it and use it it just may fix the vbr headers and then it will also correct the data in MC. Again the VBR headers have nothing to do with ID3v2 tags.

"Update tags from Library" will not do it, but "Update library from tags" may (i am not sure if MC will re-analize the file) to get the correct bit rate and length\duration.

[RandyR]

King,

I will try the program.

Not quite sure I understand your point though about re-analysis.  Even if MC were to re-analyze and store the correct info in its database (which it somhow appears to have NOW as it IS showing the correct length and bitrate), what good would the re-analysis do me in my situation, if it did not update the mp3 file as well as the MC database?

Remember, I am having problems with this information in other software that relies on information stored somewhere in the mp3 file.  These programs are reporting songs as 25 minutes when the are 5ish.

Randy

BTW...thanks to you and GHammer for your insights.

[KingSparta]

Remember, I am having problems with this information in other software that relies on information stored somewhere in the mp3 file.

VBR header is bad

once you fix the VBR headers re-import them into MC again.

some programs will use the VBR headers and not analize the mp3 (since it is quicker)

so if the vbr headers are wrong it will always be wrong untill you use a program to fix the headers.

MC does not fix VBR headers.

[RandyR]

Ok King.

Thanks for your explanations and patience.

I just downloaded and tried correcting a sample file (Lord Grenville by Al Stewart).  However, results are not what I expected.

1.  Originally, bitrate was reported (by Tag&Rename and others) as 32 and duration as 31:29. MC reported bitrate as 182 and duration as 5:32.


2.  After repair by MP3TagStudio, bitrate is reported as 106 and duration as 9:28.  MC picks up the repair and reports same information.

3.  Watching the song play in MC, bitrates varies between 160 and 260 and appears to average around 210+.  Actual duration is 4:57...a little more than half of what is reported.

Not sure where to go from here though.

Any ideas?

[GHammer]

GHammer,

I do see.  But I'm not sure what to do about it.

I have approx 10,500 mp3 files which were encoded with VBR 320 by a lansoic unit using the Xing encoder.  I used this units because it controlls a 400 CD carousel and I was able to load up CDs and go away.  However, years later, I have dicsovered that it put 32 as the birate in virtually all the files. 

MC seems to be able to tell what birate and duration "should" be and displays it properly (I think).  I guess I am wondering why King seems to wants so much to prevent it from fixing the tag as well.

I any event, assuming he/JRiver don't relent on their position, short of re-ripping (which I really can't do for several reasons), how can I get the birate/duration fixed - preferably without buying even more software.

You are looking too much at a display in a particular program.

What we are both saying is that the MP3 headers are broken in most or all of your MP3s.

We are saying that you need to repair those files if you want most players to know the duration of the songs.

KingSparta explained that MC does not repair MP3 headers. It does determine the length in another way, but for me I find that it is more accurate when the headers are correct.

After you repair the headers with another app (we have suggested several), I'd either select them all in MC and use "Update from files" or simply do an import and let MC see that the files have been changed by an external app.

Yeah, you have a lot of files to repair. I'd certainly look at something that can batch them. I know mp3Trim can do that, don't know about other apps.

[GHammer]

Ok King.

Thanks for your explanations and patience.

I just downloaded and tried correcting a sample file (Lord Grenville by Al Stewart).  However, results are not what I expected.

1.  Originally, bitrate was reported (by Tag&Rename and others) as 32 and duration as 31:29. MC reported bitrate as 182 and duration as 5:32.


2.  After repair by MP3TagStudio, bitrate is reported as 106 and duration as 9:28.  MC picks up the repair and reports same information.

3.  Watching the song play in MC, bitrates varies between 160 and 260 and appears to average around 210+.  Actual duration is 4:57...a little more than half of what is reported.

Not sure where to go from here though.

Any ideas?

Continue with your repairs.

That file seems to be a VBR.
It is normal for MC to show the instant bitrate while playing and the average bitrate in the library listing.

[RandyR]

GHammer,

I appologize.  I think I wasn't clear.

What I am saying is that AFTER repair using Mp3TagStudio, the NEW information (bitrate of 106 and duration of 9:28) is still wrong.

The VBR shown by MC as the song plays is correct and NEVER below 160 and the "true" time is 4:57 - no question about that.  The "repaired" values are nowhere near correct.  The values estimated by MC from the original data file (bitrate=182 and duration =5:36) are much closer than the repaired values of bitrate=106 and duration=9:28.

These changed but still incorrect values are not of any more use to me than the original data from the corrupted header which showed 32K as bitrate and 31+ minutes as duration.

I really don't know what to do from here.

[GHammer]

What I am saying is that AFTER repair using Mp3TagStudio, the NEW information (bitrate of 106 and duration of 9:28) is still wrong.

The VBR shown by MC as the song plays is correct and NEVER below 160 and the "true" time is 4:57 - no question about that.  The "repaired" values are nowhere near correct.  The values estimated by MC from the original data file (bitrate=182 and duration =5:36) are much closer than the repaired values of bitrate=106 and duration=9:28.

These changed but still incorrect values are not of any more use to me than the original data from the corrupted header which showed 32K as bitrate and 31+ minutes as duration.

I really don't know what to do from here.

Go get two tools. We're gonna take a look at two things and I like to get confirmation from two sources.

Get mp3Trim. There is a free version which will work on files of a reasonable length. I don't know if the free version does batch processing, but we'll find out.

http://www.mptrim.com/index.html
Nevermind, the free version will not repair VBR info. The useful paid version is $31.95

Get foobar2000 Special Instaler. It comes with the plugins you'll need packaged.

http://foobar.morbo.org/download.html

Let me know when you have them. I'll be around for an hour or so.

[RandyR]

GHammer,

OK.  Got Foobar installed.

Also, I had previously download MP3Trim as you suggested, and you are correct that VBR repair requires either the PE or Pro versions - which are both pay versions.

BTW, I really appreciate you taking the time to help me with this.

Randy

[GHammer]

Ok, run foobar.
Add one file that you know to be bad to the playlist.
Right click on the selection.
Choose Fix MP3 Header
For now, we're gonna go with the default values that will be in the window that will pop up.
Click OK

Tell me what you get.

[RandyR]

GHammer,

Here are the results:

Originally....

bitrate = 32
codec = MP3
channels = 2
samplerate = 44100
mp3_stereo_mode = joint stereo
mp3gain_minmax = 000,191
mp3gain_album_minmax = 000,192
mp3gain_undo = -002,-002,N
replaygain_track_gain = -1.560000 dB
replaygain_track_peak = 0.887324
replaygain_album_gain = -0.500000 dB
replaygain_album_peak = 0.939113
    --------------------
82915823 samples @ 44100Hz
(rounded samples : 82915644)

Length = 31:20

_________________________________


After repair...

enc_delay = 576
enc_padding = 0
mp3_accurate_length = yes
bitrate = 202
codec = MP3
channels = 2
samplerate = 44100
extrainfo = VBR
mp3_stereo_mode = joint stereo
mp3gain_minmax = 000,191
mp3gain_album_minmax = 000,192
mp3gain_undo = -002,-002,N
replaygain_track_gain = -1.560000 dB
replaygain_track_peak = 0.887324
replaygain_album_gain = -0.500000 dB
replaygain_album_peak = 0.939113
    --------------------
13215791 samples @ 44100Hz
(rounded samples : 13215300)

Length is 4:57...which is correct!!

Now, what?

[GHammer]

Now back to MC.
Do a Update Library from tags on that file.

The bitrate and duration should update to the new/correct value.

What after that? Well, you have LOTS of broken MP3s. You are going to have to repair them. One by one looks no fun to me.

That $32.00 might be worth it to be able to batch repair using mp3Trim.

Those are the only two tools I've had good luck with. Maybe someone else knows a free tool that will repair bad headers.

At least you know what the problem is now.

[RandyR]

GHammer,

Again, you are correct.  MC now has the correct values.

In terms of bath repair, are you saing FooBar does not permit this?

Also, I'm not unwilling to shell out the $32 for MP3Trim - if that will fix the problem...but how do I know - before buying the PE or Pro version - that it will produce the same results as FooBar?  After all, MP3TagStudio (suggested by KingSparta) did NOT fix the problem properly.

Randy


An idea:  Could I e-mail you (off list) the sample file we have been using.  As you seem to have MP3Trim, perhaps you would be willing to run my file and see what it does with it.  Let me know if this would be OK.

Thanks again,
Randy

[GHammer]

Because I said so. 

The thing is, if you buy it, take the values that come up in the registration page.
And then pay attention to what your settings are. You'll want to uncheck "Remove digital silence" on the 'Batch' tab. All else should be ok at defaults.

If you have a small file I'd be happy to test it for you before you buy.

Sorry for being slow. I was wasting time looking for another header repair tool.
Didn't find any...

[RandyR]

The file is 7.39M.  Could I e-mail it to you?  I would feel much better know it produces the desired results.

Randy

[GHammer]

The file is 7.39M.  Could I e-mail it to you?  I would feel much better know it produces the desired results.

Randy

Check for a PM from me.

[GHammer]

You may want to give this a try before anything else.
I have no broken files to try it on.

http://www.willwap.co.uk/Programs/vbrfix.php

Get the MFC version.

[KingSparta]

After all, MP3TagStudio (suggested by KingSparta) did NOT fix the problem properly.

no, it is not perfict

you could report this to him and give him a copy of the files it does not fix so he can fix his program, it is howerver free.

[RandyR]

GHammer,

I download the "vbrfix" you referred to.  It comes close (4:59 vs "true" 4"57) and would be fine - except I can't figure out how to batch run files in the actual program - although some of the screen shots seem to indicate that it can.


Any ideas??

King,

I did post the discrepency on the Mp3/Tag Studio forum.

[GHammer]

GHammer,

I download the "vbrfix" you referred to.  It comes close (4:59 vs "true" 4"57) and would be fine - except I can't figure out how to batch run files in the actual program - although some of the screen shots seem to indicate that it can.

If it's close enough for you cool. It's free.

Under 'File' Select Output Directory
Under Options "Output Method" Choose Output To Selected Output Directory.
You can choose Overwrite here if you're feeling lucky...
You have lots of files, I'd do 50 or so and see how it goes. Then maybe I'd choose overwrite. Because you're gonna need lots of room to duplicate the files.

Ok, after you decide on that, File "Fix Directory of MP3s" Or "Fix Directories recursively".

All done.

Let us know how it goes, it is not an uncommon problem.

[RandyR]

GHammer,

I'm with you most of the way.

However, the version of the program I am running (1.0, extracted from 0.2r5 zip file) doesn't have "Fix Directory of MP3s" Or "Fix Directories recursively" under the File menu...it only has "Fix MP3(s)" and "Select Output Directory"...and only accept a single MP3 file at a time.

What am I missing?

[GHammer]

GHammer,

I'm with you most of the way.

However, the version of the program I am running (1.0, extracted from 0.2r5 zip file) doesn't have "Fix Directory of MP3s" Or "Fix Directories recursively" under the File menu...it only has "Fix MP3(s)" and "Select Output Directory"...and only accept a single MP3 file at a time.

What am I missing?

vbrfixwin-0_25.tar.gz

http://www.willwap.co.uk/Downloads/vbrfixwin-0_25.tar.gz

[RandyR]

GHammer,

I did download the .gz file and found a program to unzip it (XP doesn't have one natively).

Well...because nothing is easy...now it accepts the directory as input but doesn't store anything in the output directory.  I switched to the same directory (input and output) using the option to add a "vbrfix" notation as part of the title in the revised file and, watching the log activity, it seems to only create a file for the last entry in the directory.

Were you able to successfully process an entire directory?  If so, any idea what I might be doing wrong?  I feel like an idiot.

[GHammer]

Go back to the version you downloaded first.
Set your options.
Drag and drop files onto it. I just tried and it worked fine with 10 files.
Newer version overwrites and only outputs the last file in the list.

[RandyR]

Ok.

Making progress...got that to work.  And the good news is the ouput matches Foobar...at least for several albums I've tested.

The next problem...I can't drop a directory on it...it seems to only accept .mp3 files in batch.  I have each album (987 of them) in a separate directory folder. 

Do you think I am going to have to copy them all to a single directory to batch process them and then refile them in separate directories when complete?

The "recursive" directory function in the .gz version doesn't seem to be implement here in the .zip version.

You've been so great so far...any ideas here???

[GHammer]

I'd dump them all in a single directory.

IF

They are all properly tagged.
Then after you repair them you can use MC to move them to a proper folder order.

[RandyR]

Ok.

Was hoping to avoid that...but...

Perhaps, I'll post a quesiton on the vbrfix forum and see what he says first.  Seems kind of silly to require this.

Thanks again, GHammer.  You've been great!

[JimH]

Anyone using this program...
Mp3/Tag Studio 3.5, beta 8

Please read this post by Alex B:
http://yabb.jriver.com/interact/index.php?board=3;topic=25711.0#msg178385

[RandyR]

Will repaired his "VBRFix" program so that it now works with recursive directories (folders of files within folders).  For those not follwoing this entire thread, this is a free utility that will repair the VBR header on mp3 files so that bitrate and duration/length are correct.

There are still some problems when outputting to a different directory, but if rewriting files, it seems to work fine.

Latest version for Windows can be downloaded from...

http://www.willwap.co.uk/Downloads/vbrfixwin-0_26.zip

Will suggests you run a virus checker on it before installing.

[KingSparta]

I do think a bigger issue is how did they all get corrupt (if you ripped them your self).

If they were downloaded I can understand why

[RandyR]

King,

Not sure.  As I noted above, I used a Lansonic DAS and its built in MP3 decoder (which according to info in the file is Xing) to rip them originally at VBR 320.

The original file info is also pasted in a reply above so you may be able to glean something from that.  I really don't know how to read these things, but it looks like the file was not tagged as VBR...at least some of the fields report by MP3/Tag Edit Studio are not only different in value, but different fields comparing before and after a fix.

I guess I'll leave this to the experts (e.g you)...I'm just glad you and GHammer helped me find a way to fix it.

Thanks again.

Randy

BTW, I did communicate with the guy who wrote MP3/Tag Edit Studio and gave him a sample file that wasn't working in his program.  Also, told him about Will's program (which is open source) so he can compare his algorhythm to one that seems to get correct results.

[JimH]

There isn't any VBR 320.  It's either 320 or VBR.

[RandyR]

Jim,

That's true in for the LAME encoder I know, but the implementation built into the Lansonic hardware uses "VBR 320" as the highest quality VBR menu setting.  Actual ABRs of the results are mostly around 200 judging from the actual corrected header results.

[GHammer]

All's well that ends well. I'm glad you were able to get Will to get you a fix so quickly.

A good reason to use a reasonably new version of LAME and to use "accepted" settings for it. Like in MC. I wonder if you could rip from your changer using MC?

I understand that you had to use their tool their way. I'd be asking if there is an update or if they could allow an external encoder though. What a pain.

[RandyR]

I actually think it can be done with MC using slink-e, which allows a PC to control a sony CD multi-changer.  But, when I did this several years ago, that wasn't really an option.

Now, since I'm only ripping a few CDs at a time, I just run MC on several laptops concurrently with 8x rip speeds and everything goes pretty fast, but for a large collection the MC slink-e route is well worth investigating.

[hit_ny]

i would recommend that RandyR run Mp3Utility on his files after fixing them.

I use this program a lot to check for broken stuff, it has a recurse option and is free.

It wont fix anything, tells you where in the file there are problems if any. Might help to narrow down which files to fix

[RandyR]

Thanks.  I'll check it out.

[KingSparta]

Thanks.  I'll check it out.

it is very good

I had him add a few options a few years ago to it but have not use it much since.

I also have it on my ftp site