Paul Taylor
2009-08-20 13:21:05 UTC
Hi
"the four bytes of decompressed size will precede the encryption method
byte" is certainly wrong as it doesn't exist. I think after rereading
this section half a dozen time that you right and I will modify
jaudiotagger accordinlgy. However jaudiotagger only reads these values
it doesn't write them so :
Do we know of any other application that write the information
misunderstood this
Can we publish 2.4.1 on the website with this amendment (and any agreed
others) so others don't repeat the mistake
It could be good practise to write the read code to check both Data
Length/Encryption and Encryption/Data Length order when both Compression
and Encryption are set to cope with incorrectly written files, but not
sure if safe algorithm is possible.
Anyone know of any application that compresses and encrypts frame in 2.4.
Paul
"the four bytes of decompressed size will precede the encryption method
byte" is certainly wrong as it doesn't exist. I think after rereading
this section half a dozen time that you right and I will modify
jaudiotagger accordinlgy. However jaudiotagger only reads these values
it doesn't write them so :
Do we know of any other application that write the information
misunderstood this
Can we publish 2.4.1 on the website with this amendment (and any agreed
others) so others don't repeat the mistake
It could be good practise to write the read code to check both Data
Length/Encryption and Encryption/Data Length order when both Compression
and Encryption are set to cope with incorrectly written files, but not
sure if safe algorithm is possible.
Anyone know of any application that compresses and encrypts frame in 2.4.
Paul
Hi,
I think that the statement "the four bytes of decompressed size will
precede the encryption method byte" within ID3v2.4 simply has been
overlooked when adopting the text from the ID3v2.3 specification (a
classical copy & paste error).
I therefore assume that this statement is *not* correct within ID3v2.4
- this is the only thing that makes sense. I think that the ID3v2.4
statement "this information is added [...] in the same order as the
flags that indicates them" should be considered as the "stronger" rule.
I'd therefore say you should write the additional information bytes in
the order as the flags indicate them. If you've found an ID3v2
implementation which does it the other way round, you could contact
them and ask them to fix it.
The statement in question should be removed (or adopted) within the
ID3v2.4 specification as it causes confusion and is definitely
incorrect. (Who is responsible for that?)
Furthermore, there seems to be a typo within the statement "A 'Data
Length Indicator' byte MUST be included in the frame." from section
4.1.2 "k - Compression". It should either be "A 'Data Length
Indicator' bit MUST be included in the frame." or "A 'Data Length
Indicator' MUST be included in the frame.". A data length indicator
*byte* however makes no sense to me and also causes confusion - it
should be the data length indicator *bit* in my opinion.
Regards,
Mathias K.
------------------------------------------------------------------------
*Gesendet:* Freitag, den 7. August 2009, 21:21:41 Uhr
*Betreff:* [ID3 Dev] Order of additional information fields (ID3v2.4)
Hi,
I'd like to rephrase a question that has been asked on this list before,
but there has never been a complete answer.
Both ID3v2.3 and ID3v2.4 support additional information fields of frames
Some flags indicates that the frame header is extended with
additional information. This information will be added to the frame
header in the same order as the flags indicating the additions. I.e.
the four bytes of decompressed size will precede the encryption
method byte. These additions to the frame header, while not included
in the frame header size but are included in the 'frame size' field,
are not subject to encryption or compression.
Some frame format flags indicate that additional information fields
are added to the frame. This information is added after the frame
header and before the frame data in the same order as the flags that
indicates them. I.e. the four bytes of decompressed size will precede
the encryption method byte. These additions affects the 'frame size'
field, but are not subject to encryption or compression.
While the order for ID3v2.3 is clear
Data Length indicator $xx xx xx xx
Encryption method $xx
Grouping indentificator $xx
it appears to be ambiguous for ID3v2.4. For ID3v2.4 order of flags is
1) Grouping identity requires a single unsigned byte
2) Compression requires the presence of a data-length-indicator flag,
but it does not add any information itself (unlike in ID3v2.3)
3) Encryption requires a single unsigned byte as the encryption method,
and might optionally use a data-length-indicator flag
4) Unsynchronisation might optionally use a data-length-indicator flag
5) Data-length-indicator flag is an sync-safe 32-bit value
This means that both the compression and the unsynchronisation flags will
not add extra data and the resulting order of fields is
Grouping indentificator $xx
Encryption method $xx
Data Length indicator 4* %0xxxxxxx
But the ID3v2.4 standard still includes this statement from ID3v2.3
I.e. the four bytes of decompressed size will precede the encryption
method byte.
which makes no sense since there is no additional data related to
compression.
It will set - just like encryption and unsynchronisation might -
simply the
data-length-indicator flag. However it seems some people interpret this
statement in the way that the order should be
Grouping indentificator $xx
Data Length indicator 4* %0xxxxxxx
Encryption method $xx
I think the first order is correct (and it is the only one that make
sense to
me). Jess Harpur has also voted for this order on this mailing list
http://article.gmane.org/gmane.comp.multimedia.id3v2/610
However, the only implementation I was able to find that supports
these fields
is jaudiotagger v1.0.9 which uses the second order.
So, right now its 2:1. And that's not good.
While its easy to identify the fields while reading (encryption method
symbol
is in the range 0x80,...,0xf0, and data-length-indicator bytes are
%0xxxxxxx),
it is still not clear to me how to write them in a standard compliant way.
I would really appreaciate your opinion about this.
Thanks,
Martin
---------------------------------------------------------------------
I think that the statement "the four bytes of decompressed size will
precede the encryption method byte" within ID3v2.4 simply has been
overlooked when adopting the text from the ID3v2.3 specification (a
classical copy & paste error).
I therefore assume that this statement is *not* correct within ID3v2.4
- this is the only thing that makes sense. I think that the ID3v2.4
statement "this information is added [...] in the same order as the
flags that indicates them" should be considered as the "stronger" rule.
I'd therefore say you should write the additional information bytes in
the order as the flags indicate them. If you've found an ID3v2
implementation which does it the other way round, you could contact
them and ask them to fix it.
The statement in question should be removed (or adopted) within the
ID3v2.4 specification as it causes confusion and is definitely
incorrect. (Who is responsible for that?)
Furthermore, there seems to be a typo within the statement "A 'Data
Length Indicator' byte MUST be included in the frame." from section
4.1.2 "k - Compression". It should either be "A 'Data Length
Indicator' bit MUST be included in the frame." or "A 'Data Length
Indicator' MUST be included in the frame.". A data length indicator
*byte* however makes no sense to me and also causes confusion - it
should be the data length indicator *bit* in my opinion.
Regards,
Mathias K.
------------------------------------------------------------------------
*Gesendet:* Freitag, den 7. August 2009, 21:21:41 Uhr
*Betreff:* [ID3 Dev] Order of additional information fields (ID3v2.4)
Hi,
I'd like to rephrase a question that has been asked on this list before,
but there has never been a complete answer.
Both ID3v2.3 and ID3v2.4 support additional information fields of frames
Some flags indicates that the frame header is extended with
additional information. This information will be added to the frame
header in the same order as the flags indicating the additions. I.e.
the four bytes of decompressed size will precede the encryption
method byte. These additions to the frame header, while not included
in the frame header size but are included in the 'frame size' field,
are not subject to encryption or compression.
Some frame format flags indicate that additional information fields
are added to the frame. This information is added after the frame
header and before the frame data in the same order as the flags that
indicates them. I.e. the four bytes of decompressed size will precede
the encryption method byte. These additions affects the 'frame size'
field, but are not subject to encryption or compression.
While the order for ID3v2.3 is clear
Data Length indicator $xx xx xx xx
Encryption method $xx
Grouping indentificator $xx
it appears to be ambiguous for ID3v2.4. For ID3v2.4 order of flags is
1) Grouping identity requires a single unsigned byte
2) Compression requires the presence of a data-length-indicator flag,
but it does not add any information itself (unlike in ID3v2.3)
3) Encryption requires a single unsigned byte as the encryption method,
and might optionally use a data-length-indicator flag
4) Unsynchronisation might optionally use a data-length-indicator flag
5) Data-length-indicator flag is an sync-safe 32-bit value
This means that both the compression and the unsynchronisation flags will
not add extra data and the resulting order of fields is
Grouping indentificator $xx
Encryption method $xx
Data Length indicator 4* %0xxxxxxx
But the ID3v2.4 standard still includes this statement from ID3v2.3
I.e. the four bytes of decompressed size will precede the encryption
method byte.
which makes no sense since there is no additional data related to
compression.
It will set - just like encryption and unsynchronisation might -
simply the
data-length-indicator flag. However it seems some people interpret this
statement in the way that the order should be
Grouping indentificator $xx
Data Length indicator 4* %0xxxxxxx
Encryption method $xx
I think the first order is correct (and it is the only one that make
sense to
me). Jess Harpur has also voted for this order on this mailing list
http://article.gmane.org/gmane.comp.multimedia.id3v2/610
However, the only implementation I was able to find that supports
these fields
is jaudiotagger v1.0.9 which uses the second order.
So, right now its 2:1. And that's not good.
While its easy to identify the fields while reading (encryption method
symbol
is in the range 0x80,...,0xf0, and data-length-indicator bytes are
%0xxxxxxx),
it is still not clear to me how to write them in a standard compliant way.
I would really appreaciate your opinion about this.
Thanks,
Martin
---------------------------------------------------------------------