Board index FlightGear Development Aircraft

Looking for aircraft with 'paper cut' bugs

Questions and discussion about creating aircraft. Flight dynamics, 3d models, cockpits, systems, animation, textures.

Re: Looking for aircraft with 'paper cut' bugs

Postby bugman » Thu May 07, 2015 11:13 pm

IAHM-COL wrote in Thu May 07, 2015 10:48 pm:well the point being, I believe the ratings are going to be needed for the QT5 launcher.
So that's yet another corner of paper to cut.

Getting a template with a mock value (zero seems a reasonable binary of "not rate yet added") seems the way to go to the wealth of non=rated aircraft.


I know that James will be able to initialize the ratings variables to zero, and leave that as the default if nothing can be found in the *-set.xml file. This is a small programatic fix that requires no fleet-wide modifications. It is a million times simpler to fix there, so if James leaves me anything to fix, which I strongly doubt, then I'd fix it there.

Regards,

Edward
bugman
Moderator
 
Posts: 1808
Joined: Thu Mar 19, 2015 10:01 am
Version: next

Re: Looking for aircraft with 'paper cut' bugs

Postby IAHM-COL » Thu May 07, 2015 11:21 pm

OK
The other paper cut is the variant of tag.
If we gave everybody in the World free software today, but we failed to teach them about the four freedoms, five years from now, would they still have it? Probably not, because if they don’t recognise their freedoms, they’ll let their freedoms fall
User avatar
IAHM-COL
Retired
 
Posts: 4057
Joined: Wed Aug 08, 2012 6:40 pm
Location: Homey, NV (KXTA) - U.S.A
Callsign: HK-424D or ICAO4243
Version: 3.7-git
OS: Linux

Re: Looking for aircraft with 'paper cut' bugs

Postby Thorsten » Fri May 08, 2015 8:26 am

I think a "vote" system may allow for ponderation or averaging of opinions. Rating could even be continuous scale as oppose to a discrete scale (like 3.25, for example), if several opinions are averaged.


The problem is that some votes matter way more than others. Even if there are hundred votes stating an aircraft is good, but there's one dissenting vote from an aerospace engineer who is familiar with the plane, I'd discard the hundred votes.

FDM quality is not about opinions, likes and dislikes it's objectively definable and measurable.

Except, we can't seem to agree who is able to understand enough aerodynamics to judge planes. Case in point - observe Simon (Bomber) consistently arguing that the relevant measure is time spent with JSBSim rather than understanding the math for instance.

I readily recognize Vincent's points as entirely valid and reasonable, and I suspect so would anyone who knows aerodynamics, but the people who create unrealistic FDMs and overrate them would not recognize the same points as reasonable (or they would make different planes in the first place), nor recognize why general physics understanding is really helpful in judging what a plane should do - so we end up with the classic situation that proving expertise to an expert is not needed because an expert recognizes it anyway, but proving expertise to a non-expert is close to impossible because the non-expert does not know enough to recognize an expert.

So, an open vote would lead to a meaningless 'like/dislike' statement, whereas a vote among an expert panel would leave the question of who qualifies to be part of the panel.
Thorsten
 
Posts: 12490
Joined: Mon Nov 02, 2009 9:33 am

Re: Looking for aircraft with 'paper cut' bugs

Postby Johan G » Fri May 08, 2015 2:25 pm

Thorsten wrote in Fri May 08, 2015 8:26 am:
IAHM-COL wrote in Thu May 07, 2015 6:27 pm:I think a "vote" system may allow for ponderation or averaging of opinions. Rating could even be continuous scale as oppose to a discrete scale (like 3.25, for example), if several opinions are averaged.


The problem is that some votes matter way more than others. Even if there are hundred votes stating an aircraft is good, but there's one dissenting vote from an aerospace engineer who is familiar with the plane, I'd discard the hundred votes.

FDM quality is not about opinions, likes and dislikes it's objectively definable and measurable.

I can only agree. For an extreme example, consider hvengels North American F-51D Mustang. He has literally spent hundreds on hours spread over more than a year perfecting the FDM so that it performs very close to the flight manual.

As it performs by the flight manual and published data, that also means that it is very tricky for an unexperienced pilot both to start, take off and land. Thus I have no trouble whatsoever imagining that a vote on the FDM rating would be low. It is simply too hard to fly for most users.
Low-level flying — It's all fun and games till someone looses an engine. (Paraphrased from a YouTube video)
Improving the Dassault Mirage F1 (Wiki, Forum, GitLab. Work in slow progress)
Some YouTube videos
Johan G
Moderator
 
Posts: 6629
Joined: Fri Aug 06, 2010 6:33 pm
Location: Sweden
Callsign: SE-JG
IRC name: Johan_G
Version: 2020.3.4
OS: Windows 10, 64 bit

Re: Looking for aircraft with 'paper cut' bugs

Postby IAHM-COL » Fri May 08, 2015 4:24 pm

IAHM-COL wrote in Thu May 07, 2015 7:08 pm:makes sense Vincent.

@Thorsten and @Johan

Yes. I also agree Vincent brings a very valid point about a voting system in a complex scale that requires experience in judgement: call it FDM

I also see how such measure will largely benefit by an "expert" panel assigning scores. The problem being, if the same panel does not rank all the aircraft, then some aircraft --ranked by the panel-- will be subjected to a different ruler than all other aircraft. Different rulers make for biased rates. I imagine the expert panel may be more though than the regular public, and in other situations (like p51d example by Johan) more lenient.
If we gave everybody in the World free software today, but we failed to teach them about the four freedoms, five years from now, would they still have it? Probably not, because if they don’t recognise their freedoms, they’ll let their freedoms fall
User avatar
IAHM-COL
Retired
 
Posts: 4057
Joined: Wed Aug 08, 2012 6:40 pm
Location: Homey, NV (KXTA) - U.S.A
Callsign: HK-424D or ICAO4243
Version: 3.7-git
OS: Linux

Re: Looking for aircraft with 'paper cut' bugs

Postby IAHM-COL » Sat May 09, 2015 3:13 am

bugman wrote in Thu May 07, 2015 7:06 pm:Continuing with the an2 aircraft example



Dear Edward

I think this is how I will play
(understand play here, as in play guitar, not as in play football-soccer, since we are referring to audio files)


Right now, all stereo audio files had been corrected in every single iteration of FGMEMBERS.
There were about 266 stereo files in 84 aircrafts. Not all of them belonging to FGAddon.

I merged the channels, and created a mono-aural that represents the stereo rather well, but may loose depth in a few specific cases as you pointed out.

If you ever decide to bring a fix over the FGAddon, I will merge your more fancy technique downstream, but in the interim, the aircrafts in the FGMEMBERs come with the full power of absolute "mono" audio immersion :D

Looking forward to see your fixes

Best,
IH-COL

https://github.com/FGMEMBERS/an2/commit ... d16df6cd34
If we gave everybody in the World free software today, but we failed to teach them about the four freedoms, five years from now, would they still have it? Probably not, because if they don’t recognise their freedoms, they’ll let their freedoms fall
User avatar
IAHM-COL
Retired
 
Posts: 4057
Joined: Wed Aug 08, 2012 6:40 pm
Location: Homey, NV (KXTA) - U.S.A
Callsign: HK-424D or ICAO4243
Version: 3.7-git
OS: Linux

Re: Looking for aircraft with 'paper cut' bugs

Postby Bomber » Sat May 09, 2015 11:55 am

Having a PhD and being condescending doesn't make one an expert on fdm development that's my argument.... Not the amount of time spent.

'magic numbers' are rife in fdm development here, with the aim of making the plane hit the numbers.... Not fly correctly.

My opInion is that looking at the code that makes up the flight model against published data, this data being distributed with the fdm is the start of understanding the quality of a flight model.

The highest rating for any fdm here is a 3...
Last edited by Bomber on Sat May 09, 2015 1:17 pm, edited 1 time in total.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1933
Joined: Fri Dec 14, 2007 8:06 pm
OS: Windows XP and 10

Re: Looking for aircraft with 'paper cut' bugs

Postby KL-666 » Sat May 09, 2015 12:14 pm

Hello Bomber,

I was actually just thinking about requiring the developer to include the data he used for the fdm. It should not be too hard to put the pdf's of the aircraft data and the pilots manual in a documents folder of the aircraft. Then there can not be any discussion about: "The document i use says 140 kts" versus "No you are wrong, mine says 150 kts".

The penalty for failing to add these documents should probably be: No points.

Kind regards, Vincent
KL-666
 
Posts: 781
Joined: Sat Jan 19, 2013 2:32 pm

Re: Looking for aircraft with 'paper cut' bugs

Postby Bomber » Sat May 09, 2015 12:47 pm

exactly Vincent..

I followed flug's sopwith camel data folder in this way and include ALL the data I used, and any spreadsheets and drawings I created along the way..

I think its almost impossible to have a conversation about a flight model if you've no clear idea as to what the author was intending to simulate...

I personally laugh when a person says that a prop plane does or doesn't hit the numbers, as they never use the sentence.... with a climb prop on it doesn't reach it's top speed, or with a cruise prop its climb is incorrect... or with a speed prop.....

Simon
Last edited by Bomber on Sat May 09, 2015 1:19 pm, edited 1 time in total.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1933
Joined: Fri Dec 14, 2007 8:06 pm
OS: Windows XP and 10

Re: Looking for aircraft with 'paper cut' bugs

Postby Bomber » Sat May 09, 2015 1:04 pm

Also a POH is about playing it safe and not necessarily the planes max performance.
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1933
Joined: Fri Dec 14, 2007 8:06 pm
OS: Windows XP and 10

Re: Looking for aircraft with 'paper cut' bugs

Postby Hooray » Sat May 09, 2015 1:18 pm

KL-666 wrote in Sat May 09, 2015 12:14 pm:I was actually just thinking about requiring the developer to include the data he used for the fdm. It should not be too hard to put the pdf's of the aircraft data and the pilots manual in a documents folder of the aircraft. Then there can not be any discussion about: "The document i use says 140 kts" versus "No you are wrong, mine says 150 kts".


That would be awesome, and has been suggested a number of times previously - unfortunately, it isn't exactly feasible usually due to legalities - i.e. depending on the license of the aircraft you're distributing, you can rarely -if ever- include proprietary documents (think PDF files), unless they happen to have compatible licensing (e.g. public domain). Otherwise, you will open up a whole can of worms - and given how even perfectly legal aircraft have been removed over the years due to legal concerns, it's a fine line to walk.

As a compromise, my suggestion would be to "delegate" storage of the corresponding files to another website: usually, you can include URLs to the corresponding files - and you can also upload the files elsewhere to ensure that they remain available (if in doubt, use a torrent) - that will keep the repository itself "clean", because it will only ever contain URLs to files hosted elsewhere.
Please don't send support requests by PM, instead post your questions on the forum so that all users can contribute and benefit
Thanks & all the best,
Hooray
Help write next month's newsletter !
pui2canvas | MapStructure | Canvas Development | Programming resources
Hooray
 
Posts: 12707
Joined: Tue Mar 25, 2008 9:40 am
Pronouns: THOU

Re: Looking for aircraft with 'paper cut' bugs

Postby Bomber » Sat May 09, 2015 1:49 pm

Whatever the method, the data should be made available
"If anyone ever tells you anything about an aeroplane which is so bloody complicated you can't understand it, take it from me - it's all balls" - R J Mitchel
Bomber
 
Posts: 1933
Joined: Fri Dec 14, 2007 8:06 pm
OS: Windows XP and 10

Re: Looking for aircraft with 'paper cut' bugs

Postby bugman » Sun May 10, 2015 6:34 pm

As I earlier said I would, I have now brought the stereo sound file issue up on the development mailing list:


My strategy is rather more involved than averaging the file into a mono source. I have instead performed a full audit of each file individually. Importantly, the proposed changes should result in no stereo information lost! I have gone through every file one by one. I first checked if the file was even used (i.e. they were in an XML tag anywhere in that aircraft directory). I then used the following Python script to listen to the difference between the two channels, using the command line audio utility sox:

Code: Select all
#! /usr/bin/env python

# Python script for testing stereo audio files.

# Python module imports.
from os import system
from shutil import rmtree
import sys
from tempfile import mkdtemp


def run(cmd):
    print(cmd)
    system(cmd)

# Check.
if len(sys.argv) != 2:
    print("Usage:\n    stereo_test.py audio_file")
    sys.exit()

# The file name.
file = sys.argv[1]

# Create a temporary directory for storing intermediate audio files.
dir = mkdtemp()

# Execute everything safely.
try:
    # Split off the first channel.
    run("sox %s %s/left.wav remix 1" % (file, dir))

    # Split off the second channel, inverted.
    run("sox -v -1 %s %s/right.wav remix 2" % (file, dir))

    # Combine.
    run("sox --combine mix %s/left.wav %s/right.wav %s/comb.wav" % (dir, dir, dir))

    # Play.
    run("play %s/comb.wav" % dir)


# Final cleanup.
finally:
    rmtree(dir)


And finally I used Audacity to look at the waveforms, and then take the difference and look at that (see the devel list message for how to do this). As you'll see in the archived message, from all this information I have made a file-by-file proposal for how to resolve this issue once and for all.

Regards,

Edward
Last edited by bugman on Sun May 10, 2015 6:59 pm, edited 1 time in total.
bugman
Moderator
 
Posts: 1808
Joined: Thu Mar 19, 2015 10:01 am
Version: next

Re: Looking for aircraft with 'paper cut' bugs

Postby IAHM-COL » Sun May 10, 2015 6:38 pm

It "Sounds" great to me :D
I already got rid of stereo files in FGMEMBERS. Then I am merging back whenever you get this more technical solution over.
In the interim, FGMEMBERS aircraft have the mono sound.
Good work Ed.
If we gave everybody in the World free software today, but we failed to teach them about the four freedoms, five years from now, would they still have it? Probably not, because if they don’t recognise their freedoms, they’ll let their freedoms fall
User avatar
IAHM-COL
Retired
 
Posts: 4057
Joined: Wed Aug 08, 2012 6:40 pm
Location: Homey, NV (KXTA) - U.S.A
Callsign: HK-424D or ICAO4243
Version: 3.7-git
OS: Linux

Re: Looking for aircraft with 'paper cut' bugs

Postby Necolatis » Sun May 10, 2015 7:01 pm

Remember if the files are used as <type>avionics</type> then you cannot split them into 2 sources positioned differently, as positions is ignored for this type, then you should downmix.
"Airplane travel is nature's way of making you look like your passport photo."
— Al Gore
User avatar
Necolatis
 
Posts: 2233
Joined: Mon Oct 29, 2012 1:40 am
Location: EKOD
Callsign: Leto
IRC name: Neco
Version: 2020.3.19
OS: Windows 10

PreviousNext

Return to Aircraft

Who is online

Users browsing this forum: No registered users and 16 guests