Help - Search - Members - Calendar
Full Version: Ual93 Fdr Decoder Program
Pilots For 9/11 Truth Forum > Flight Number > United 93
wstutt
I have created a program to decode UAL93's raw FDR file that I am making available for download along with it's source code.

It allows you to decode all the subframes other than just the final flight.

It decodes the "PRES POSN" (Latitude and Longitude) and "RADIO HEIGHT" parameters which are not in the CSV files from the NTSB.

It can decode incomplete subframes.

You can read more about it here
rob balsamo
Thank you Warren,

Excellent work. If you like, send us the new CSV files you create with your decoder and i'll put them up for download.

Thanks again....

Rob
dMz
Hi again Warren,

Very nice work. The sourcecode looks quite a bit like Java2. The open source idea should scare away a few of the "debunkers" hopefully... rolleyes.gif

Would it be possible to modify your code slightly to have it spit out all of the UA93 FDR data in Excel-sized chunks? (Less than 256 column "IV"- probably about 225 columns "wide" to leave room for formulas, and less than 65500 rows in "length.") I think I got 7 CSV files for AA77 when splitting the 2nd "aa77_complete.csv" source file columns manually as I recall. Also, Excel doesn't like more than 30,000 "rows" when graphing XY graph types.

Thanks again for your excellent work here. cheers.gif
d
wstutt
QUOTE (rob balsamo @ Feb 12 2008, 04:28 AM)
Thank you Warren,

Excellent work. If you like, send us the new CSV files you create with your decoder and i'll put them up for download.

Thanks again....

Rob

Thanks Rob,

People can run the program to generate CSV files for whatever parameters and periods of time they want themselves, however if there are any that people are particularly interested in, I could generate and upload them.

Warren.
wstutt
Hi dMole,

QUOTE (dMole @ Feb 12 2008, 04:49 AM)
Hi again Warren,

Very nice work.  The sourcecode looks quite a bit like Java2.  The open source idea should scare away a few of the "debunkers" hopefully... rolleyes.gif


Thanks dMole. My intention in making the source code available is that if anything interesting is found in the data, that it could not be simply excused away on the program making up the data. smile.gif

QUOTE (dMole @ Feb 12 2008, 04:49 AM)
Would it be possible to modify your code slightly to have it spit out all of the UA93 FDR data in Excel-sized chunks?  (Less than 256 column "IV"- probably about 225 columns "wide" to leave room for formulas, and less than 65500 rows in "length.")  I think I got 7 CSV files for AA77 when splitting the 2nd "aa77_complete.csv" source file columns manually as I recall.  Also, Excel doesn't like more than 30,000 "rows" when graphing XY graph types.

Thanks again for your excellent work here. cheers.gif
d


At this point the program only generates a maximum of 188 columns when all fields are included so I believe that would not cause a problem with Excel.

I could add an option for splitting up the file to limit the number of rows to 30000. Some programs like Microsoft Access are not limited to 65500 rows, so I would make it an option.

I could either create sets of columns for each group of 30000 rows, or I could create files with sequential numbers inserted in their names so that no file has more than 30000 rows. Which option do you think would be best?

In the meantime, the program has the option of selecting which subframes are to be included, so by choosing the ranges of subframes to number no more than 30000, files with no more than 30000 rows could be generated individually.
dMz
QUOTE (wstutt @ Feb 12 2008, 12:20 PM)
I could either create sets of columns for each group of 30000 rows, or I could create files with sequential numbers inserted in their names so that no file has more than 30000 rows. Which option do you think would be best?

Hi Warren,

I personally like the batch processing/sequential numbers approach. That way, people could poke around in whichever data file is applicable. I also like having some empty columns to paste data from other CSV files and formulas into.

Thanks again for your expertise,
d

EDIT: Re-reading your reply, so are there only 188 columns of data for the alleged UA93 FDR then?
wstutt
Thanks dMole,

QUOTE (dMole @ Feb 12 2008, 05:43 PM)
I personally like the batch processing/sequential numbers approach.  That way, people could poke around in whichever data file is applicable.  I also like having some empty columns to paste data from other CSV files and formulas into.


I've included the option you suggested in the program and uploaded it to my web site as version 1.1

If you have already installed the program, you can just install it again to get the updated version. There is no need to uninstall it first.

QUOTE (dMole @ Feb 12 2008, 05:43 PM)
EDIT: Re-reading your reply, so are there only 188 columns of data for the alleged UA93 FDR then?


The NTSB CSV files have more than 400 columns of data and there are yet more parameters according to the NTSB's United Airlines Flight 93 FDR Report. My program does not handle them all.

The columns in the NTSB CSV files that my program does not handle all contain values that did not change during the final flight and except for about 3 columns are single bit columns that can only have two values (i.e. the ON/OFF type of columns). e.g. the status of the lavatory smoke detector did not change during the final flight smile.gif

If there are any parameters that you would like to see handled, let me know.

Warren.
wstutt
I've uploaded a copy of the raw FDR file so that you can download it without having to download a whole CDROM ISO image. You can access it from the help page.

I've also added some comments to the source code and removed a few lines of code that were commented out.

Warren.
UnderTow
QUOTE (wstutt @ Feb 24 2008, 10:04 PM) *
I've uploaded a copy of the raw FDR file so that you can download it without having to download a whole CDROM ISO image. You can access it from the help page.

I've also added some comments to the source code and removed a few lines of code that were commented out.

Warren.


WoW Warren.
Has anyone gathered any new information from these read outs?

I'll have to look at the C. Do you think it would work on the other file as well?
wstutt
Thanks Undertow,

QUOTE (UnderTow @ Mar 10 2008, 05:50 AM) *
Has anyone gathered any new information from these read outs?

I am not aware of anybody else gathering any new information by using the program.

I did however check a few things myself.

The data goes back to a flight in progress on 2nd September 2001 GMT.

I checked the airports that the plane had landed and taken off from and found nothing that surprised me. I can post or upload more details if anyone is interested.

I checked the final location (latitude/longitude) and found that it agrees with the official story.

I expected the final radio height values to decrease to zero but they didn't. I consider the last values to be invalid. I suspect this is due to the radar not working properly since the plane was pitched at such a large angle, however I am certainly no expert on this.

I could run the program and generate a file of these final values and upload it if anyone is interested.


QUOTE (UnderTow @ Mar 10 2008, 05:50 AM) *
Do you think it would work on the other file as well?

Unfortunately not as it is.

The program is specifically written to decode flight data that uses the data frame layout used by UAL93. AAL77 uses a different data frame layout. If I had an uncompressed AAL77 FDR file, I could reverse engineer it like I did for the UAL93 FDR file (which was not compressed) and write a similar program to decode the uncompressed AAL77 FDR file.

At this point, I only have the (mostly) compressed AAL77 FDR file and I have not yet been able to uncompress it. I do not know if I will eventually be able to uncompress it or not. I did however find some short (up to 4 seconds in length) segments within the file that are uncompressed. I believe these segments are generated when the FDR is first powered up when the engines are started. The segments appear to be in the format that is outlined in the file 573b.Short.txt on your web site.

If I had an uncompressed AAL77 FDR file, this would probably help me to be able to uncompress the compressed AAL77 FDR file. In other words, I could probably work out how the uncompressed file was generated from the compressed file. That would then give me enough information to write a program that decodes the compressed file that does not need to use the uncompressed file.

Warren.
rob balsamo
The last values for RadAlt should read 2500 if it is to support the govt story. The RadAlt was beyond its pitch/bank limits for recording any value due to aircraft attitude, according to the govt story, therefore the RadAlt should show its max altitude limit which is 2500.

Most of the information contained in the csv file and animation for UA93 does support the govt story. A bit too much if you ask me... almost as if they learned their lesson with AA77 files and went in to make UA93 files more accurate. Such as true altitude.. etc. UA93 files actually debunk the "duh-bunkers" who try to use "altimeter lag" as an excuse for AA77.

However, UA93 files do not support observed events as described in our documentary and press release.
UnderTow
Thanks Warren.
I'm gonna check a few things and see what I can do.
wstutt
Hi Rob,

QUOTE (rob balsamo @ Mar 10 2008, 12:21 PM) *
The last values for RadAlt should read 2500 if it is to support the govt story. The RadAlt was beyond its pitch/bank limits for recording any value due to aircraft attitude, according to the govt story, therefore the RadAlt should show its max altitude limit which is 2500.

The last values for what I believe to be the different radio altimeters and captains display are 4830, 4590, 4710 and 4817 feet. These values appear 2 or 3 times in a row in the final 11 seconds (one of the four values is recorded every second) which is why I believe they are invalid. You have confirmed my suspicion as to why this would be. thumbsup.gif

The radio height values for the final flight increase to about 5480 feet as the plane is taking off and climbing, before repeating values of 5480 or 5500 feet which would appear to be the maximum radio height that can be detected.

QUOTE (rob balsamo @ Mar 10 2008, 12:21 PM) *
However, UA93 files do not support observed events as described in our documentary and press release.

I have seen the documentary. I agree that at least to my inexperienced eyes, that the impact site does not appear to be consistent with a plane at a 40 degree pitch. I also couldn't see nearly as much metal from the plane as I would have expected to see (as pointed out in Loose Change 2).

Warren.
UnderTow
Warren,
Can you add an output option to only output the sync word row? This would greatly shorten the number of rows and make some other things easier when working with the csv.

In other words, for each subFrame there are 8 rows, I just want the 1st row of each subFrame.

Thanks,
UT
rob balsamo
Thats interesting Warren. Can you put together a small csv file with the RadAlt next to pressure altitude with time stamp?
wstutt
Hi UnderTow,

QUOTE (UnderTow @ Mar 10 2008, 09:47 PM) *
Warren,
Can you add an output option to only output the sync word row? This would greatly shorten the number of rows and make some other things easier when working with the csv.

In other words, for each subFrame there are 8 rows, I just want the 1st row of each subFrame.

Thanks,
UT

I presume you selected vertical acceleration as one of the fields to include in the output. That would be why there are 8 rows per subframe.

I could add an option to output only the first row of each subframe as you suggest, however for fields like vertical acceleration (there are also a few others), you would then only get one of the vertical acceleration values from each subframe instead of all 8 of them.

Would that be OK?

Warren.
UnderTow
Hm. I hadn't checked, but if I just deselect those columns (like VertAcc), will it auto-shrink the result?

If so then I can do that and you don't have to modify anything.
For most basic analysis, only the whole second data points are needed.

Thanks again,
UT
wstutt
Hi Undertow,

QUOTE (UnderTow @ Mar 11 2008, 04:03 AM) *
Hm. I hadn't checked, but if I just deselect those columns (like VertAcc), will it auto-shrink the result?

Yes it will auto-shrink the result. The fields you need to make sure are deselected are:
CONTROL COLUMN POSN-CAPT
CONTROL WHEEL POSN-CAPT
INDICATED AOA
LATERAL ACCELERATION
LONGITUDINAL ACCEL
RUDDER PEDAL POSN
RUDDER POSITION
VERTICAL ACCELERATION

QUOTE (UnderTow @ Mar 11 2008, 04:03 AM) *
If so then I can do that and you don't have to modify anything.
For most basic analysis, only the whole second data points are needed.

I'll hold off on making the modification to the program at this point until either you decide you would like it or I am making other modifications.

Warren.
dMz
Just wanted to say hi Warren.

It looks like you have been busy with UA93 "SSFDR" data. I got re-directed back toward AA77 and a few other personal projects. Carry on and good work! cheers.gif
wstutt
I've just discovered today that the install process from the web page for the program is not working (at least for me anyway).

I'm working to fix this. I believe the problem is with the configuration of my web hosting provider's web server software.

Warren.
wstutt
I've switched my web hosting plan from one running Apache under Linux to one running Microsoft IIS under Windows and the install process is working for me now. If anyone is still having problems with it let me know.

Warren.
wstutt
I've uploaded version 1.2 which adds the option of generating no more than one line per subframe as suggested by UnderTow earlier in this topic . I've also added an output file with the RADIO HEIGHT, ALTITUDE (1013.25mB) and GMT parameters for the final flight like what Rob asked me to email him earlier in this topic. I've also changed references to fields to refer to parameters.

Warren.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2017 Invision Power Services, Inc.