rcCommand is after rc expo and also features like roll yaw uptilt compensation. Would be nice to be able to log actual stick positions for training, as well as configure the viewer to show either stick positions or rcCommand (or both at the same time, for comparison?).

I currently struggle to open a blackbox logfile within the viewer. I am able to extract it using the tools but there seems to be no interaction with the viewer once I try to open it. Please see below the decoding (it might be possible that entry 2 is causing this) and the attached logfile

blackbox_decode LOG00134.TXT
Decoding log 'LOG00134.TXT' to 'LOG00134.01.csv'...

Log 1 of 4, start 00:10.546, end 01:25.804, duration 01:15.257

Statistics
Looptime 1882 avg 458.0 std dev (24.3%)
I frames 2312 38.1 bytes avg 88195 bytes total
P frames 34570 24.5 bytes avg 846409 bytes total
E frames 1 12.0 bytes avg 12 bytes total
S frames 46 4.0 bytes avg 184 bytes total
Frames 36882 25.3 bytes avg 934604 bytes total
Data rate 490Hz 12467 bytes/s 124700 baud

17 frames failed to decode, rendering 115 loop iterations unreadable. 38360 iterations are missing in total (38367ms, 50.98%)

Decoding log 'LOG00134.TXT' to 'LOG00134.02.csv'...

Log 2 of 4Statistics
E frames 1 12.0 bytes avg 12 bytes total
Frames 0
Data rate: Unknown, no timing information available.

1 iterations are missing in total (0ms, 100.00%)

Decoding log 'LOG00134.TXT' to 'LOG00134.03.csv'...

Log 3 of 4, start 02:25.534, end 02:29.398, duration 00:03.864

Statistics
Looptime 999 avg 23.2 std dev (2.3%)
I frames 121 33.7 bytes avg 4074 bytes total
P frames 1812 21.8 bytes avg 39527 bytes total
E frames 2 9.0 bytes avg 18 bytes total
S frames 1 4.0 bytes avg 4 bytes total
Frames 1933 22.6 bytes avg 43601 bytes total
Data rate 500Hz 11579 bytes/s 115800 baud

1932 loop iterations weren't logged because of your blackbox_rate settings (1931ms, 49.99%)

Decoding log 'LOG00134.TXT' to 'LOG00134.04.csv'...

Log 4 of 4, start 03:16.295, end 03:27.295, duration 00:11.000

Statistics
Looptime 999 avg 23.5 std dev (2.4%)
I frames 338 34.4 bytes avg 11642 bytes total
P frames 5043 22.6 bytes avg 113946 bytes total
E frames 2 9.0 bytes avg 18 bytes total
S frames 5 4.0 bytes avg 20 bytes total
Frames 5381 23.3 bytes avg 125588 bytes total
Data rate 489Hz 11548 bytes/s 115500 baud

3 frames failed to decode, rendering 14 loop iterations unreadable. 120 iterations are missing in total (119ms, 1.09%)
5500 loop iterations weren't logged because of your blackbox_rate settings (5499ms, 50.00%)

Similar to the way quicktime player works (or any video player) it would be immensely helpful to be able to use the left / right arrow keys to go frame by frame. Shift arrow could go 10 frames.

when I run my log file the icon on the upper left display of the quad, jerks madly in all directions, the graphs display ok. Is this a icon rendering error or is it actually what is recorded incorrectly ?

After removing all the fields in the graph setup and restarting the browser the buttons to hide/show the legend don't work. It's working as long as you don't close the browser. The bug appears in Chrome/Firefox/Safari on Mac OS 10.10.3.

Hi there, a long term follower of cleanflgiht blackbox log analysis here. I'm wondering if folks are fine to sparate the parsing code in flightlog_parser.js out into a reusable library that can be used by third-party apps? The reason for the ask is, with the current logging logic in the firmware, especially with all the field encodings and frame format, it's really hard to code something from ground-up (nor does it make much economic sense) just to work with the logging code that was written over 5 years ago. Hopefully with a separate library, it'll make things easier and encourage a richer ecosystem around PID tuning and issue diagnosis, as other developers can focus on high-level data representations instead of worrying about all the nitty-gritty encoding/frame format issues.

if the log data contain heading we can estimate the G vector, and subract it to the acceeleration vector, so we get the real acceleration along the 3 axis.

Integrating those acceleration also give a rough estimation of the speed, that would be nice to have.

With the laptop in the field on a sunny day i find it very hard to see the lines. Thicker lines would help.

Title says it all. Would be a killer feature!

When log file is opened after video file, no graph will be rendered. The other way around works just fine.

betaflight/betaflight#2093

Any chance of an option to display the Tx sticks in mode 1?

feature request:
add the ability to change the axis orientation for the 3D quad so that it can be seen from sides and front&rear, versus only being able to view it 'from above' (as it is now). This would help in diagnosing issues with roll & pitch movement (oscillations, bounce-back, etc.) by being able to better view small changes in the quad's attitude.

I prefer my graph colors to be consistent, so that I can learn to intuitively read the graphs. For example, I always set my graphs up with P, then I, then D so that the colors are consistent. If I have P and D turned on but I turned off, I set the traces up as 1:P, 2:D, 3:D, so that the 3rd trace overlays the 2nd trace and maintains the same color as if the 2nd, I trace, was on. This is kind of cumbersome. Ideally, I'd like to be able to assign a color to each of the traces manually and have that setting remembered. For example, in a master configuration, assign PID_P[pitch] = cyan across all graphs; PID_I[pitch]=magenta; and so forth.

I recently got some logs tearing around in my Mitsuko and found there is no way to change the visualization from a X quad to a + quad. This would be helpful in visualizing how much the rear motor is working to providing forward thrust.

It's clear the rear motor (1 on the + quad) corresponds to the back left motor (1 on the X quad), of course, but it would be nice if the visualization matched the data.

Improve the automatic color assignment for graph lines so that similar traces across multiple graphs will receive similar coloring (e.g. if you have P, I, D plotted for pitch and Gyro, P, I, D plotted for roll, the PID terms should receive the same coloring even though they don't appear in the same order on the graph legends).

As I use mode 1 the stick display is incorrect as it displays mode 2, can this be a feature to change

Gyro data is no longer rendered. The new gyroADC data when rendered are basically strait lines and now useless.

Also the quad model is missing.
1.9.0 pre release was good.
This started with 1.9.0 stable
Image is at 1/1 sample rate

Reference cleanflight PR#998
cleanflight/cleanflight#998

Under "Introduction to Blackbox" , "Tuning your carft", and "Other Tools" Tabs, all links are broken.

Thanks

Unable to type I or O characters in graph axis labels. It instead assigns the in and out points behind the graph config modal.

I have several common setups that I use for analysis. It'd be nice to be able to switch between them without having to manually set them up every time.

Cannot load extension with file or directory name Thumbs.db. The filename is illegal.

Wondering if it would be feasible to utilize the blackbox data to assist with video stabilization within the viewer, or produce some kind of transform/tracking information that can then be used in post, such as After Effects or Final Cut Pro X, to save time or produce better results. I see some information at http://www.google.com/patents/US8488010, so people have done similar research prior to this idea.

I using FC KAKUTE V2 with BT v3.3.0. Flash size for LOG data is 16MB. When I try opening this file, Chrome Crasching or MSG BOX Page Unresponsive.

blackbox-tools also can't read - blackbox-decode

How cut this file for few smallest?

Currently, I'm new to Cleanflight and I've been having some trouble.
I flew my tricopter and recorded about 8:00 of blackbox data, and then tried to overlay it with matching Mobius footage as usual. The rendering in Blackbox Explorer goes smoothly, but then stops at 52.8% every time.

I'm running OS X, both videos were done at 30FPS 720P w/ 0% dimming.
This issue seems to be specific to these files, as those from a later flight render just fine.

Hi !

I have no issue to read my log files with blackbox_render tool.
But when i try to load them into blackbox explorer then it says : "Error : Log is truncated, contains no data"

I record with those settings :
115200,26,3,0,1,1,0
baud,escape,esc#,mode,verb,echo,ignoreRX

Blackbox is triggered from a switch.

And I import my log files directly from the SD card (i cannot plug openlog TX to my FC to import from cleanflight).

May you please tell me what is going wrong?
i've attached a log as example.
LOG00016.zip

Hello guys, is it possible to add a feature showing the total reported current draw for a flight/log?

taking gp4 file and rendering it to 966x540 60fps, that 5 min file comes out to 137mb

import it into bblv and render at 720 and the file is 1.03 gig

any hints on how to reduce exported file size?

Hi,

I'm using this tool and works very well but when i try to render a logfile without video, i have the screen that say calculating and it freeze at this point. I can move the windows, the app is working but the render won't start.

I'm under windows 10, the app is installed by chrome.
thank you.

Taranis/OpenTX collects much of the same data as the Blackbox, so it'd be useful to be able to read Taranis telemetry logs in the viewer.

When attempting to click the graph below the main window in Firefox (Windows 7, 37.0.1), most of the time, playback stops as if it was reset to the beginning and paused, the red bar disappears, and playback can not be resumed. Seemingly at random, it does work to skip to the spot in the graph, but only rarely and not the same spot twice. Also, the index timer in the bottom right displays NaN:NaN:NaN when this happens.

None of this occurs with Chrome, using the exact same log file and exact same log viewer.

See https://youtu.be/dnqv-sP-AtE. What is the scoop? Why is the latest version of Chrome (2.5.13) have less feature, more missed header file reads, and the scaling of RC command screwed up and 2.5.11, which you can download here, seem better and RC command scaling is right?

Using the Chrome App, the video is jumpy. An example can be seen here:

https://www.youtube.com/watch?v=sbt2P54ZjRQ&feature=youtu.be

The video feed comes from the Fatshark V2 on board DVR. The format was not supported, so it was converted to H264, H264 forced 30FPS and MP4 (using handbrake).

MP4 did not work, and both H264 formats jumped like in the Youtube video.

When viewed in the chrome app, everything looks fine though.

OS: Mac OS 10.10.4
Chrome: 44.0.2403.155 (64-bit)

This is a new feature request.

I use the latest blackbox explorer (9333f41).

I open a log file and it can be seen, however the "View/hide analyser display" doesn't seem to work at all. When clicked, the window is not shown and the icon doesn't turn yellow.

Did it change? It used to work for me some time ago...
How to debug it further?

It would be nice to be able to change reporing time interval to see more clearly differences between loop to loop and smaller ripples or even to zoom out to see the bigger picture.

Firefox seems not to be able to run the blackbox-log-viewer as standalone app.
Steps to reproduce:

1. Install firefox quantum 68.0.1 (64-bit) on Ubuntu
2. Download full zip of blackbox-log-viewer from github and unzip it to a folder
3. Run firefox and open the index.html page
4. See debug console:
   20:06:57.789 require does not exist, maybe inside chrome index.js:84:17 20:06:57.788 ReferenceError: chrome is not defined main.js:1892:9 20:06:57.792 ReferenceError: chrome is not defined release_checker.js:16:5 20:06:57.809 ReferenceError: chrome is not definedlocalization.js:9:9
   Try to open any BBL file, nothing will happen.

overlay % does not seems to work, using 80% or 20% seems to give the same video result.

also give possibility to specify desired overlay, i've got some dark video where a 10% or even 0% should be sufficient.

It would be nice to be able to save different graph setups so you don't have to reconfigure each time when looking to different things.

The specific thing that got me thinking about this was trying to analyze vbat. It jumps around so much that it's basically useless to look at the trace.

Hi, on both blackbox_decode and the blackbox-log-viewer, i am getting the following message when i open one of the .TXT files from my Blackbox SD card:

Error opening log: No logs in this file could be parsed successfully

I can't seem to get the tool to accept any of the files, yet they are a few hundred kilobytes big. My looptime is 1600, is that too low? i've seen a proper looking logfile and mine looks totally garbled.
Thanks.

I tried on Debian Jessie Mate, Ubuntu Mate 16.04 (now fresh install) & Chrome 53.0.2785.116 (64-bit) at all times and nothing happens, always Calculating... the file size, and no rendering.

Here is the DeveloperTools console output:
http://imgur.com/a/eypT0

...

Hi,

I want to look my log in viewer but I can't play it in my Chrome. Sometimes shows WebGL error. I can load the log into viewer but the viewer shows me blank, the graph don't exist.
I'm using Windows 8.1 32bit, Ati Radeon Xpress 1100.

I Googled for WebGL error and give me result that is from VGA driver. I have updated it but still no luck. Any how to fix this?

Thanks

Hello!
I want to add the ability to use blackbox-log-viewer with MAVLink .bin (log files)
Currently i think about some kind of log file converter from MAVLink .bin files to blackbox .txt
Is there some relevant projects/papers/blog post that i should read?
I found good description of Blackbox protocol. Should I learn it or it will be enough to use included libraries?
Can you suggest better approach? Any help will be welcomed! Thx in advance!