MediaInfo(Lib) 0.7 - Reading information about media files

[vlada]

Zenitram> I fixed the image, but Kurtnoise already perfectly explained, what I meant.

Kurtnoise> How do I get the track number from MediaInfo? I can only see StreamKindID and StreamKindPos. This helps me to identify order of tracks of certain type (usually audio tracks) but I still don't know, which track is first (audio or video). In the VOB (DVD-Video) example I posted the video track has ID higher the all audio tracks but still ffmpeg sees it as first track (#0.0). I tried to search the ffmpeg documentation but I couldn't find anything that would help me to understand it.

[Zenitram]

Originally Posted by vlada

Kurtnoise> How do I get the track number from MediaInfo? I can only see StreamKindID and StreamKindPos.

?
As you can see in the MediaInfo output, there is an ID field too. ID is the number in brackets in FFMpeg (in blue in Kurtnoise example).

[vlada]

Originally Posted by Zenitram

?
As you can see in the MediaInfo output, there is an ID field too. ID is the number in brackets in FFMpeg (in blue in Kurtnoise example).

I know this. But to demux or transcode certain audio track using ffmpeg, I have to use -map switch. This switch accepts the red number. So I'm wondering if there is any possibility to get this number from MediaInfo.

[Zenitram]

Originally Posted by vlada

This switch accepts the red number. So I'm wondering if there is any possibility to get this number from MediaInfo.

The red number seems to be the order of appearence.
There is no such field in MediaInfo.
I don't plan to integrate such a field myself, this is possible, but I don't want to spend time on it, not my priority.

[vlada]

Zenitram> OK, understand. I will code a workaround for this, it shouldn't be that difficult.

Btw. some time ago we talked about IFO parsing. Has this moved a little higher on your to-do list? Basically all I need to know is correct movie length, audio/subtitle languages and chapters. Do you plan to support it?

Thanks for all your work and perfect support!

[Zenitram]

Originally Posted by vlada

Btw. some time ago we talked about IFO parsing. Has this moved a little higher on your to-do list? Basically all I need to know is correct movie length, audio/subtitle languages and chapters. Do you plan to support it?

No, I currently have more "professional" and paid feature requests, I focus on it (I try to live from MediaInfo), so IFO, not very professional ;-), is even lower in my to-do list.

[LoRd_MuldeR]

I understand why you focus on paid feature requests, sure. But how is IFO parsing not professional? IFO files are part of the official Video DVD specs, as used on millions of DVD discs.

[Zenitram]

Originally Posted by LoRd_MuldeR

But how is IFO parsing not professional? IFO files are part of the official Video DVD specs, as used on millions of DVD releases.

IFO is created by professionals, but the usage of MediaInfo for IFO (or Blu-ray) is not very professional, I know a bit where is used MediaInfo for theses formats ;-).
Professionals using my software are currently interested more in digital preservation --> DV, MPEG-TS (Satellitte), and now MXF , I currently focus on them, but I try to not forget my base, and I will try to have a good parser for all formats... But days are only 24 hour long and bug tracker is not empty .

And unfortunatly, even if MediaInfo si open-source software, nobody is currently helping me, I hope to be able next year to pay someone in order to fix all issues like your IFO problem.
(if you like formats, if you like C++, don't hesitate to contact me :-D )

[LoRd_MuldeR]

Sorry, but I don't have time to work on yet another project. I already work on several personal/hobby projects in addition to my "real life" job

[zn]

Resolution : 16 bits

This is irrelevant description for MP3, MP2, OGG and AAC audio, because they use float values inside.
(decoders mpg123, faad, oggdec supports float or 16/24/32 int output)

[Zenitram]

Originally Posted by zn

This is irrelevant description for MP3, MP2, OGG and AAC audio, because they use float values inside.
(decoders mpg123, faad, oggdec supports float or 16/24/32 int output)

OK.
I removed "Resolution" field for MPEG Audio, Vorbis (OGG is not an audio format ;-) ), ADTS, ADIF and MPEG-4 AudioSpecificConfig (the last 3 are the different forms of AAC I have), they actually were "forced" in my code, this was not a good idea.

[Keiyakusha]

Ohh, I always forget to ask... mediainfo shows for mkv files something like this:
Container profile=Unknown@4.1
what this means? 4.1 looks like AVC level, there should be something like "High" instead of unknown then?

Also does some infotip configuring planned?

[MrC]

Hi Zenitram,

first of all I would like to thank you for your hard and great work on MediaInfo. I have been using it since ages in AVStoDVD and I must say it is a very reliable and useful tool.

I have a reading issue to report: this 5.1 DTS file (4 MB) should have a bitrate of 1536 kbps and that bitrate was correctly read by MediaInfo until release 0.7.10. Latest releases read 1509 kbps, actually saying that that file is not DVD compliant, while it should be.

Thanks in advance for your reply



Bye

[Zenitram]

Originally Posted by Keiyakusha

mediainfo shows for mkv files something like this:
Container profile=Unknown@4.1
what this means?

This is something I didn't yet understand.
Matroska trace example:

(...)
0000115A    CodecPrivate (41 bytes)
0000115A     Header (3 bytes)
0000115A      Name:                                 23A2 (9122)
0000115C      Size:                                 26 (38)
0000115D     Specific (38 bytes)
0000115D      Header (0 bytes)
0000115D      Profile:                                  1 (1)
0000115E      Compatible profile:                       64 (100)
0000115F      Reserved:                                 0 (0)
00001160      Level:                                    29 (41)
00001161      Reserved:                                 3F (63)
00001161      Size of NALU length minus 1:              3 (3)
00001162      Reserved:                                 7 (7)
00001162      seq_parameter_set count:                  1 (1)
00001163      seq_parameter_set (25 bytes)
00001163       Size:                                    17 (23)
00001165       nal_ref_idc:                             3 (3)
00001165       nal_unit_type:                           7 (7)
00001166       profile_idc:                             64 (100)
(...)

The same thing in a MP4:

(...)
00000643        AVC decode (33 bytes)
00000643         Header (8 bytes)
00000643          Size:                             21 (33)
00000647          Name:                             avcC
0000064B         Version:                           1 (1)
0000064C         Specific (24 bytes)
0000064C          Header (0 bytes)
0000064C          Profile:                                  4D (77)
0000064D          Compatible profile:                       40 (64)
0000064E          Level:                                    C (12)
0000064F          Reserved:                                 3F (63)
0000064F          Size of NALU length minus 1:              3 (3)
00000650          Reserved:                                 7 (7)
00000650          seq_parameter_set count:                  1 (1)
00000651          seq_parameter_set (12 bytes)
00000651           Size:                                    A (10)
00000653           nal_ref_idc:                             1 (1)
00000653           nal_unit_type:                           7 (7)
00000654           profile_idc:                             4D (77)
(...)

I see this in some MKV and in some FLV.
Theses files are reported to be NOT playable on some HW players, so users want to see if the "Profile" field is the same as in the seq_parameter_set. if they are different, I display "Container profile" (this value is not used by any player I know), and 1 is un unknown profile from specs.

I am interested if you have any explaination about this...

[Zenitram]

Originally Posted by MrC

this 5.1 DTS file (4 MB) should have a bitrate of 1536 kbps

No, see DTS specifications:

Due to the limitations of the transmission medium the actual bit rate may be slightly different from the targeted bit rate,
as listed in table 5.8 for the two types of applications

Table 5.8: Targeted and actual bit-rate for the CD and DVD-Video applications
Targeted Bit Rate --> Actual Bit Rate on DVD-Video Discs
768 --> 754.50 (DVD-Video Discs)
1 411,2 --> 1 234,8 (DTS CDs)
1536 --> 1509.75 (DVD-Video Discs)

MediaInfo tries to display the real bitrate on the medium, so from v0.7.15 I correct bitrates to mach with actual bitrate and not targeted bitrate.

[MrC]

Thanks for the crystal clear explanation, Zenitram.

May I ask you another question? When reading chapters position in matroska containers, the visualization is quite different from other media parameters:

Menu
00:00:00.000                     : :Chapter 1
00:02:11.200                     : :Chapter 2
00:05:10.240                     : :Chapter 3
00:09:09.240                     : :Chapter 4
00:13:44.040                     : :Chapter 5
00:16:07.120                     : :Chapter 6
00:21:51.480                     : :Chapter 7
00:24:50.360                     : :Chapter 8
00:31:43.480                     : :Chapter 9
00:39:12.400                     : :Chapter 10
00:41:17.520                     : :Chapter 11
00:46:40.360                     : :Chapter 12
00:52:08.720                     : :Chapter 13
00:58:27.840                     : :Chapter 14
01:02:42.600                     : :Chapter 15
01:06:37.440                     : :Chapter 16
01:11:50.200                     : :Chapter 17
01:15:31.080                     : :Chapter 18
01:20:52.400                     : :Chapter 19
01:23:23.880                     : :Chapter 20

It is quite difficult to read infos formatted in that way using mediainfo.dll reading function (MediaInfo_Get), because MediaInfo_Info_Text or MediaInfo_Info_Name are not univocal, like i.e. FrameRate or ScanType.

Is there any workaround I can use to automate chapters reading thru mediainfo.dll?

Thanks in advance



Bye

[Zenitram]

Originally Posted by MrC

It is quite difficult to read infos formatted in that way using mediainfo.dll reading function (MediaInfo_Get)

Yes.
I unfortunalely didn't find a good way to display chapters like the other fields.

Originally Posted by MrC

Is there any workaround I can use to automate chapters reading thru mediainfo.dll?

Pseudo code:

int Chapters_Pos_Begin=MediaInfo_Get("Chapters_Pos_Begin"); //You must transform the string in int
int Chapters_Pos_End=MediaInfo_Get("Chapters_Pos_End");//You must transform the string in int
for (int Pos=Chapters_Pos_Begin; Pos<Chapters_Pos_End; Pos++)
{
    string Time=MediaInfo_Get(Pos, Info_Name);
    string Text=MediaInfo_Get(Pos);
}

[MrC]

That's just what I was looking for

Thanks!



Bye

[Zenitram]

Shark007 and SeeMoreDigital:

Originally Posted by Shark007

For your consideration; This 7.1 TrueHD audio sample is detected as only having 6 channels. Current FFDshow (rev3094) releases detect all 8 channels.

Originally Posted by SeeMoreDigital

If MediaInfo could identify the properties of the "core" and "HD" parts separately, yes this would be very useful.

I am working on this issue! I tried to make a understable display about Core and HD parts, next version would have for this audio stream:

Audio
ID                               : 4352 (0x1100)
Menu ID                          : 1 (0x1)
Format                           : AC-3
Format/Info                      : Audio Coding 3
Format profile                   : TrueHD / Core
Muxing mode                      : Stream extension
Duration                         : 12s 10ms
Bit rate mode                    : Variable / Constant
Bit rate                         : Variable / 640 Kbps
Channel(s)                       : 8 channels / 6 channels
Channel positions                : Front: L C R, Surround: L R, rs: L R, LFE / Front: L C R, Surround: L R, LFE
Sampling rate                    : 96.0 KHz / 48.0 KHz

I hope you like it!

Edit: I inverted TrueHD and Core data for compatibility with third-party software which uses "text to integer" values and want the highest result.

[Shark007]

Very nice solution Zenitram. good work.