StreamParser 2.0 Beta Available

Version 2.01 of StreamParser has been posted. More cameras added. Timing calculations are now being done a different way. Before, timings were calculated based on transport stream timers. Now they are based on display timestamps. This should result in time calculations that more closely match what editors show and correct timing stats for other cameras. You can get it here: http://www.dvxuser.com/V6/showthrea...ta-Available&p=2113515&viewfull=1#post2113515.

There have been some requests for editing capabilities. This is not as simple to do as one might think, so I have no plans to add such features - there are plenty of good editors out there. Although StreamParser can read MTS files from other sources (i.e. not Panasonic cameras), the stats could come out completely wrong because StreamParser relies on private header data produced by Panasonic cameras.

Chris
 
Hello Chris,
many thanks for the update and the answers. :)

I wonder if I'm the only one to really need a simple 100% lossless MTS cutter (maybe I-frame based) for GH13/GF13 footage:
I will search/ask in another more appropriate post some info for this purpose,
(any suggestion is still welcome...).

Bye :)
 
alphaomega said:
I wonder if I'm the only one to really need a simple 100% lossless MTS cutter (maybe I-frame based) for GH13/GF13 footage:
I will search/ask in another more appropriate post some info for this purpose,
(any suggestion is still welcome...).

Bye :)
Click to expand...

I would like to be able to cut, copy and paste native .MTS files as well. I do not see why I should be forced to convert to another format just to cut some parts out or merge/join files.
 
Hello Chris!

Great tool, thank you very much.

I think the AVCHD Lite files (i.e. 25p stream in a 50p wrapper) are incorrectly indicated in the information panel. I have not tested it with other AVCHD Lite files, m but at least with my Panasonic DMC-TZ10, this is what the information panel shows

Stream: Max Speed = 18 382 980 bps
Mode = 720/50p
Clip Time = 0:00:04,94

Mode and Clip Time are indicated incorrectly.

Actually it is 720 /25p (wrapped insinde 50p)
and the correct Clip Time is 2x longer (around :09,8)
 
Yeah, I found that bug and I am fixing it. I should be able to post the fix in a couple of days. I don't actually own these other cameras so testing is hard. It turns out that the 50p and 60p clips are actually 25p and 30p - and not really in a wrapper. They report at 50p or 60p but every second frame is missing altogether insread of being empty. It seems pretty odd to me - but that's how it works with some of these other cameras. Thanks for the bug report!

Chris
 
The AVCHD-Lite files are using a bitstream flag to indicate duplicated frames rather than actually encoding them. The problem is that many AVCHD editors and players don't recognize the flag, in particular Adobe Premiere Elements.
 
I find it quite ironic that the first semi-official explanations of GH2 shortcomings (no 25p, interlaced modes in 1080/50 or 60 mode) mentioned the absolute need to stick to the Holy AVCHD standard. And then we have "AVCHD-lite".
 
cbrandin said:
Yeah, I found that bug and I am fixing it. I should be able to post the fix in a couple of days. I don't actually own these other cameras so testing is hard. It turns out that the 50p and 60p clips are actually 25p and 30p - and not really in a wrapper. They report at 50p or 60p but every second frame is missing altogether insread of being empty. It seems pretty odd to me - but that's how it works with some of these other cameras. Thanks for the bug report!

Chris
Click to expand...

If you need some sample MTS footage from my DMC-TZ10, I can upload it for you. Also there is a sample DMC-TZ10 file for download here:
http://www.photographyblog.com/reviews/panasonic_lumix_dmc_tz10_review/sample_images/
See bottom of the page.
 
StreamParser 2.02 now available here: http://www.dvxuser.com/V6/showthrea...ta-Available&p=2113515&viewfull=1#post2113515.

There are two major changes:

1) A "Quick Test" menu item has been added under the "File" menu. It opens only the first 10 seconds, or so, of a stream file. This is so you can do a quick analysis of the first 10 seconds without waiting to process an entire file.

2) Bitrates are now calculated two different ways. In frames mode they are calculated according to frame presentation time (like an editor would use). In time mode bitrates are calculated according to TS stream time stamps (i.e. when data is written versus when it should be displayed). Therefore, the bitrates for the two modes will differ somewhat.

I also fixed various reported bugs. I believe the times and bitrates for some of the other cameras (the ones with a pseudo 50p or 60p framerate that is actually half that) will come out right now.

Chris
 
cbrandin said:
StreamParser 2.02 now available here: http://www.dvxuser.com/V6/showthrea...ta-Available&p=2113515&viewfull=1#post2113515.

I also fixed various reported bugs. I believe the times and bitrates for some of the other cameras (the ones with a pseudo 50p or 60p framerate that is actually half that) will come out right now.
Click to expand...

I can confirm that this is indeed fixed. Thank you. My test file now shows correct information like this.

Stream: Max Speed = 18 382 980 bps
Mode = 720/50p (actual frame rate is half)
PCR Timing Interval = 0,095 seconds
Clip Time = 0:00:09,88

I have a further query regarding this issue. Do you think it would be possible to develop a utility that removes the offending frame repeat flag and changes the framerate indication to the correct 25p or 30p. In practical terms it would be like a utility to "patch" existing "double framerate" MTS footage. As described in lpowell's post above, this double framerate thing still causes problems for a number of NLEs and othe utilities.
I am the first to admit that i know absolutely nothing about these video streams and that it may well be more complex and practically impossible to make. This is just an inquiry.
 
Unfortunately that would require rebuilding the video stream data. StreamParser operates at the TS stream level, so it doesn't have a CODEC in it that I can change. Sorry, you'll have to find another way to do it. I think most editors can be set up to handle this.

Chris
 
I recently tried Stream Parser on an MPEG-4 format AVCHD video encoded by MainConcept. (When it didn't recognize the .mpg extension, I renamed it with a .mts extension.) Stream Parser found something it didn't like at the beginning of the file and rejected it. Would it be a lot work to give Stream Parser the ability to read AVC streams from MPEG-4 files?
 
I suspect if you re-wrapped the video into a transport stream (e.g. tsmuxer) , that the utility would work on your video
 
It would probably be difficult. StreamParser just looks at TS stream headers, not the AVCHD video data underneath. Also, some of the TS header data is specific to Panasonic cameras. I'm not sure MainConcept even produces AVCHD data contained in a TS stream.

Chris
 
lpowell said:
I recently tried Stream Parser on an MPEG-4 format AVCHD video encoded by MainConcept. (When it didn't recognize the .mpg extension, I renamed it with a .mts extension.) Stream Parser found something it didn't like at the beginning of the file and rejected it. Would it be a lot work to give Stream Parser the ability to read AVC streams from MPEG-4 files?
Click to expand...

You can use this to read MPEG format: http://www.winhoros.de/docs/bitrate-viewer/
 
hi help me out here please I am trying to determine if the files i have recorded have the native hack applied as i ticked in my hack. hack applied properly. i cant find a way to work out if the hack is applied and what I am recording is 24 (or 25 in my caseP) All software reports the files before and after the hack as being the same. Stream parser only reports what you say in the dialog box that comes up when you load a file asking if it is native or not. How can i tell?
 
You must log in or register to reply here.
Back
Top