Database Info

Camera BMP bug and fix

I was testing the quality of the camera after the QTek 106 ROM upgrade and noticed using the BMP format at XL size gave edges in images really bad jaggies.
Scaling by 50% gave a good image and looking closely at the image I suspected an interlacing fault. Turns out the odd and even scanlines have been swapped. A stupid mistake and one that HTC can very easily fix.
In the meantime I've made and attached a Photoshop action that will fix these jaggy images. JPG images don't have this problem, only BMP ones.
Original image:
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Fixed image:
(in this case both images were rotated 90 degrees from landscape to portrait after fixing)
I hope this fix will work for others, let me know if it does.
BTW: The upgrade makes the video recording of the camera work absolutely great, no more split images.
I'm not sure attachments are working, here's a download link: click

can somebody please extract the cam soft from 1.06 rom?
thanks in advance

dutty said:
can somebody please extract the cam soft from 1.06 rom?
thanks in advance
Click to expand...
Click to collapse
i need it,too .Thanks

Hm... :? do not understand how to upgrade with downloaded file from this post? :?:

The attached file is not an update for your phone, it's an action for photoshop. You copy it to your Photoshop directory in "\presets\photoshop actions\".
Then in Photoshop select the menu item Window->actions (or alt-F9) and in the actions list you should see the "fix MDA Compact BMP" item.
After loading a BMP image from the camera you can select that action item, then press the 'play' button at the bottom of the actions window to activate the action.
This will fix the BMP image for you, after which you can save it.
There's no fix for this bug on the phone itself, that's why I made this Photoshop action to fix BMP images until such a firmware update is released.

thnx for the fix!
but why would you want to save as bmp anyways? does it yield better quality or retain more details? i think most ppl just save as jpg; they're probably pictures not worth quality anyways if they decided to take it on this phone

Thank you very much for your comments It really works!
Hey, do you now why .BMP photo doubles .JPG weight? Which one to prefer for best pic quality?

BMP is an uncompressed format, JPG has lossy compression, which means you lose image quality. Some noise and unsharpness is added.
Try taking the same picture in JPG and BMP format and compare them, the BMP image has noticable better quality.
Also, if you plan on editing an image, BMP is better. Otherwise you use JPG, edit and save in JPG again and get twice the quality loss.
With the bad quality of the camera any improvement in sharpness and noise is welcome

Yeah!:idea:
I`ve made some pics with my cam in JPG and BMP and now I can see difference between this two formats
Speaking the truth, I never had an intention to use BMP before...
You right - BMP is much more better.

Good to hear that you like the results.
When I have enough space on my memory card and want the best image quality I always use the BMP format. Drawback is that it's a bit slower to process on the phone.

hi I tried your photoshop fix but I got the ff errors.
warning no pixels selected
the command move is not currently available
warning no pixels selected
Could not fill because no pattern has been defined.
I think mine got stuck in the middle of your set of actions...
I am left with four hidden layers (in this order)
pattern (filled black)
odd
even
background (with lock)
I am using Windows XP SP2 and Adobe Photoshop 7

That is strange. I'm using Photoshop 8 and Windows XP SP2.
Sounds like the creation of the pattern went wrong, if it turned out all black.
I'll have a look at the script to see where the problem could be.

[PC-APP] Tweaking your X10 Equalizer

{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Android allows for customization of 8 bands of sound, each with individual values for gain, frequency range, type and qf. The EQ settings is disabled per default in our X10s, and this and more information is stored in a file called AudioFilter.csv (/system/etc/).
I wrote a handy little JAR-application that helps you edit this file and save a new copy on your PC. With our new found root privileges, this can then be pushed to device and overwrite the existing settings, modifying your X10's sound levels as you see fit.
As the /system/ parition is mounted as read-only filesystem, we can't simply push files straight to /system/etc. Therefore we will push the file to the /sdcard/ and, remount the partition as read-write, and then use the superuser privileges to copy the file over and replace the existing one.
Start by editing the AudioFilter.csv. Here is a copy of the original X10 settings file (Download: MegaUpload, MediaFire). Open it in the app, and edit it to your liking. Then save it as a new copy and follow the instructions below on how to replace the current settings. To enable the EQ, set "Enabled" value to 'FFFF', and then play with the bands' individual values. When you have the EQ set, move on to the step by step instructions:
1. PC: "adb push AudioFilter.csv /sdcard/
2. PC: "adb shell mount -o remount,rw -t yaffs2 /dev/block/mtdblock2 /system"
3. X10: "su"
4. X10: "cp /sdcard/AudioFilter.csv /system/etc/"
5. X10: "reboot"
And there you go, your settings should be changed after boot. I am new to Android, so please correct me if I am completely lost on the matter.
This is not yet working, so no need to download yet - if you don't want to help out of course!
Download: MegaUpload, MediaFire
Please do not redistribute or mirror this application without my concent.

Reserved for future usage.

Looks like good work.

You know Java, we have root, why not making an android app to do just that?
What's the point of running commands when you can do it directly from an app?

oh, it's geat idea to put EQ into X10 !! But has anyone tested?? Plz do android app and clear tutorial how to .....

tuxStyle said:
You know Java, we have root, why not making an android app to do just that?
What's the point of running commands when you can do it directly from an app?
Click to expand...
Click to collapse
Yes, that would naturally be the next step. I just wanted to see if people found this useful at all first..?

If there were an app for this, i would be the first to DL it

Has anyone tried this yet. I'd love to try it, but not sure how the values work, within the program :/

It maybe worth listing what each of the bands do and also what codes to use to increase the volume or decrease the volume.
This may also be useful
http://forum.xda-developers.com/showthread.php?t=517745

This is F****in pro!! Thank you , and this is very useful, an app would be even better but I'm happy as is
Edit: Do you have a link to explain the freq codes?

I don't understand instruction how to apply the equalizer, I havent also my x10 rooted but I will do.... could anyone do instructions how to instal the eq, step by step?????
Sent from my X10i using XDA App

@karooluss
Download both links, Mr Rusch's csv, and x10 EQ. Open eq, hit open file, find csv you downloaded. To apply follow the script on the first page.

AndyD777 said:
It maybe worth listing what each of the bands do and also what codes to use to increase the volume or decrease the volume.
This may also be useful
http://forum.xda-developers.com/showthread.php?t=517745
Click to expand...
Click to collapse
If I may assume this works like any other normal EQ: The bands themselves are not controlling different things as someone said here before, but rather different frequency ranges. So basically, per the default settings - band1 are the very lowest of frequencies (most probably rumbling "bass" sounds), and band8 are the highest (sounds we can probably not even hear). I will try to explain it part by part..
Gain: This is volume increase/decrease, normally measured in dB.
Freq. (Frequency Range): This is the peak of the frequency range, which together with Type and QF determines the effect on the sound curve.
Type: This is the type of filter used on this frequency range. I believe 1 represents a lowpass filter, 2 a highpass filter, and 3 a notch filter. There may be more types, I don't know.
QF (Quality Factor): This is, simply put, the width of the curve in this frequency range. A low QF usually affects a wide frequency range, and a high QF usually a narrow one.
As for the values used, these are all half a byte of hexadecimal. 0 is zero, or null. FFFF is max, and would in decimal represent 65535. So, the frequency values used by standard (from band1 through 8) are 400, 500, 2000, 4000, 5000, 8000, 9000, 18000 in decimal, and probably actually a Hz value.
So, the values you enter have to be in HEX. Use a decimal to hexadecimal converter if HEX values are completely foreign to you.

app please! great job man

Have you tested this on your own handset?

AndyD777 said:
Have you tested this on your own handset?
Click to expand...
Click to collapse
Of course, wouldn't post it here otherwise. Haven't had time to find good levels yet tho. Maybe you guys can help out.

MrRusch said:
Yes, that would naturally be the next step. I just wanted to see if people found this useful at all first..?
Click to expand...
Click to collapse
You are kidding?
How many audio players out there have an EQ?
This one would help a lot.

@mrrusch
Thanks for the update on how to use but one more question (it may be dumb) does changing the number in enable have any effect? I also turned the qf to 500. It seems like mp3s are a bit clearer but i cant hear a difference in wav files. Im using westone 3s no amp
Sent from my X10a using XDA App

gavriel18 said:
@mrrusch
Thanks for the update on how to use but one more question (it may be dumb) does changing the number in enable have any effect? I also turned the qf to 500. It seems like mp3s are a bit clearer but i cant hear a difference in wav files. Im using westone 3s no amp
Sent from my X10a using XDA App
Click to expand...
Click to collapse
Enable needs to be set to either 0 or FFFF (0 = off, FFFF = on). This toggles whether to use the EQ or not.

When opening the original enable is set to 0. Therefore its turrned off. If we turn it on and save it. Would we ntice any difference?

Alcatel X'pop (5035D) Camera raw files (and some questions)

Here is a small tutorial in order to shoot raw files on an alcatel 5035D phone and how to use those files. At the moment I've only tested it on a Mac.
Here is the difference between a standard jpeg file and the one coming from a the raw file shot with engineer mode.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Here is the big problem I couldn't manage to solve.
As you can see the developed raw image has worse resolution (as I couldn't solve pixel-level demosaicing problem) but a much better dynamic range and color.
HOW TO SHOOT RAW FILES
To shoot RAW you need to use mediatek phones EngineerMode, this can be in 2 different ways:
1) opening "Phone" and writing this with the keypad: *#*#3646633#*#*
2) installing SvsEngRunner app (you can find it HERE)
Once you entered EngineerMode:
1) swipe right until you reach the "Hardware Testing" tab
2) press on "Camera"
3) press on "Set Raw Capture Mode" and select "Capture Mode" then press "OK"
4) press on "Raw Type" and select "Pure Raw"
5) if you want you can also turn off lights flickering effect selecting 50Hz flicker in Europe and 60Hz one in US
6) you can also select ISO or let the phone auto-select the right one.
7) now press on "Start Preview" frame your subject and press "Capture" to shoot
In /storage/sdcard0/DCIM/CameraEM you'll find a .raw file and a similarly named JPEG.
HOW TO EDIT THOSE RAW FILES
Once you've copied those raw files on your PC if you have Photoshop already installed you'll notice you'll be able to open those .raw files with it. Don't do it. Well, if you want to experiment with Ps raw import tool good luck, I couldn't manage to get anything usable with it.
Download and install GraphicConverter from Lemke Software (HERE is the trial download link).
Once you've downloaded and installed it comes the difficult part:
1) open the .raw file with GraphicConverter
2) "Photo RAW Import" appears with a blank square in it. Just press the "OK" button on bottom right
3) another window will tell you that the image you tried to open is broken, now press on the "Try RAW import..." button
4) now copy the values you need to enter from the following image and then press "OK":
5) save the image in TIFF format and open it with any image editor.
I suggest opening it in Photoshop and apply de-interlace both horizontally and vertically (you can remove vertical interlacing by rotating by 90° the image and applying again the de-interlace filter).
Now here is a small question: does anyone know how to open a CameraEM Alcatel Mediatek MTK Raw File in any other way? Can he/she share his method or a possible solution to the problem? I've been able to access the greyscale raw image but I couldn't demosaic it, can anyone use the MATLAB demosaic tool on it and check if the conversion is better? Is there anyone willing to write something like raw2dng for this particular RAW image file format? Thanks in advance and sorry for my very bad english :crying:.
Giova

Experiments
Same photo.
Left: JPEG, Right: edited RAW file

Comparison
Top left ->> jpeg shot together with the raw file from engineermode
top right ->> "developed" raw file
bottom left ->> jpeg from standard camera.
The comparison images are zipped and attached to this post
Giova

gio91ber, could you upload somewhere RAW from your 5035D? I'm not sure whether it is because of you, or because of someone else, I've found support for Alcatel 5035D in small RAW converter called "dcraw". I need the RAW to check whether it is really can be processed by dcraw, and if so, since I also was able to process it with GraphicConverter, there may be a chance that the author of dcraw could add support for my platform/camera as well (MT6589).
Oh, and thanks for the tip with GraphicConverter, although I must admit that it works only in some cases: sometimes it decodes pictures in a wrong way and won't show window saying that it failed, i.e. "Try RAW import..." button. Moreover there's no equivalent command in main menu...

When open the camera, showing a connection error, what is the reason?

8day said:
gio91ber, could you upload somewhere RAW from your 5035D? I'm not sure whether it is because of you, or because of someone else, I've found support for Alcatel 5035D in small RAW converter called "dcraw". I need the RAW to check whether it is really can be processed by dcraw, and if so, since I also was able to process it with GraphicConverter, there may be a chance that the author of dcraw could add support for my platform/camera as well (MT6589).
Oh, and thanks for the tip with GraphicConverter, although I must admit that it works only in some cases: sometimes it decodes pictures in a wrong way and won't show window saying that it failed, i.e. "Try RAW import..." button. Moreover there's no equivalent command in main menu...
Click to expand...
Click to collapse
That's probably because of my, as I sent Dave Coffin a raw files a few months ago. It actually works with my version of dcraw. Anyway there's also another possible way to open 5035d raw files with Irfanview using these parameters.
you then have to save the image in pgm format and use this demosaicer http://www.fastcompression.com/solutions/demosaicing/demosaicing.htm to convert it from not demosaiced image to a demosaiced ppm format color image.
Anyway, I attach to this post a raw file coming from my phone. I won't be able to send any new one since I sold the phone but anyway here it is.

gio91ber said:
That's probably because of my, as I sent Dave Coffin a raw files a few months ago. It actually works with my version of dcraw. Anyway there's also another possible way to open 5035d raw files with Irfanview using these parameters.
Click to expand...
Click to collapse
Thank you! It seems I finally made it to work. BTW! There's a cool and free software for working with raw pictures called RawTherapee. If you'll have another camera/smartphone with ability to take raw pictures definitely check it out.
If someone wonders how to set up conversion of raw pictures from Lenovo A820, here's details:
DL IrfanView;
DL its "formats.dll";
drag&drop your raw into IrfanView;
in opened window set these:
width 3200;
height 2400;
BPP 10, Not normalized;
Vertical flip;
Bayer pattern used BG.

Custom Animated Boot Screens

I am close to getting custom animations working.
So far i think i got it but it's not working.
I know setup of x264 profiles pretty good and i have dumped the video animation
from the Moto-E rev.2 (2015 version) onto my PC and i have checked it out carefully.
What i did was create an x264 profile with Avidemux that copy cat's what Moto did.
The x264 encoder command line options are identical to what i got going !
I have created and tested some video's but so far none have been showing up at boot.
I don't know why..
The last thing i am checking is the frame rate (which i doubt matters)
My test video is running 24fps and the Moto animation is 30fps
But my container is fine and i have no audio on my file and i modded the x264 profile
so it has a higher CRF (constant rate factor)
In order to keep the file size exactly like the original.
Chances are there is a hard coded buffer.. so your videos would prob have to be
the same size as the original roughly (1mb)
The Moto-E 2015 bootanimation is packed in a zip file and contains
a text file with a config and then two video files.. one 10 seconds long and the other 1 sec long.
The OS loads the first one then switches to continually loop the 1 sec clip i think.
If i load up the OS with nothing it freezes on the bootloader picture (shown before animation)
and stays on that until the OS finishes background services loading etc.
The phone STILL starts up just fine !
But i seen some weirdness !
I just tried loading the files extracted in the system folder
and once again the video did not load.. no surprise..
BUT, this time a weird pulsing logo popped up that just said "android"
it was using the newer Roboto font and it had a weird glowing / pulsing effect on it.
And it's weird because there is no other files in that directory..
So i have no idea where that came from or why a bad video would trigger it vs nothing.
Note if i remove my bad video it doesn't show it !
SO..
I wanted to report in with you guys and see if anyone got a custom one running ok.
My test screen has a clip from the show Aqua Teen Hunger force.
from the season intro where they are robbing a bank.
I clipped out the band playing parts and merged the heist parts (same length as original moto vid)
so wish me luck !

Ok i figured out what the problem is. (.zip files must have NO compression)
My phone now has a custom boot-logo AND custom animated Bootscreen i made myself
I am uploading the zip file for you all to try out if you want as an attachment here.
It's about 1mb and is a dark colored Aqua Teen Hunger Force (TV show) video clip.
It's part of the intro from season 10 where they rob a bank.
i just picked a random video to make a test video-clip with.
To install it you need root of course.. then use something like RootExplorer.
First off get the zip file i made then put it on your phone like on the sdcard etc.
Open up RootExplorer and navigate to,
root/system/media
Back up your original animated boot screen file called "bootanimation.zip"
i copied it to my PC for safe keeping and left the original where it is but renamed it.
Then just copy my file into that same spot.. reboot ..enjoy
Lastly if people want i will add the Avidemux video editor x264 profile here.
It mimics the exact precise encoder settings Motorola used.
But.. i don't know how large a custom file can be.
So what i did was load my x264 profile then i adjusted the main setting (CRF)
I kept increasing it higher and re-saving my video until it was 1mb or smaller.
I'm a proud Peacock
Here is a preview.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}

Here is the Avidemux config i mentioned..
Install Avidemux (example for windows but it's available for Linux too)
Run it once and turn it off so you have the data directories created.
Put my config in this folder (on Windows)
C:\Users\<name>\AppData\Roaming\avidemux\pluginSettings\x264\3
Run Avidemux and open a video..
Select x264 then open encoder options and pick Moto-E 2015 config.
Then select mp4 container.
Open audio menu options and pick select track and remove the check box beside audio track (if it exists)
(That will get rid sound on your saved video)
Then make sure the video is the right dimensions.. 540x960
You may want to crop etc but the filter "swResize" is what you need probably..
(i like resize method Bicubic it is more accurate / detailed)
At this point you just need to set the video CRF setting in the options.
that will control the quality and in turn the file size.
and lastly you need to create 10 second clip and then a 1 second clip for looping.
call the 10 second one "01_MainPart.mp4"
call the 1 second clip "02_EndLoop.mp4"
then you create a text file called "videodesc.txt"
and put this inside it.. with the last line blank
Code:
1 0
0 1
Save your video and use 7zip or WinRar to pack the 3 files in zip format with Storage setting.
That is no compression.. otherwise the video won't work !
I tried both zippers mentioned above and both worked.
I also tried zip'ing the files on the phone and it did NOT work.. do this on your PC !
I will attach the config file itself.. but also paste it in here as a comment too below.
Code:
{
"useAdvancedConfiguration" : true,
"general" : {
"params" : "AQ=20",
"threads" : 0,
"preset" : "ultrafast",
"tuning" : "film",
"profile" : "baseline",
"fast_decode" : false,
"zero_latency" : false,
"fast_first_pass" : true,
"blueray_compatibility" : false,
"fake_interlaced" : false
},
"level" : 31,
"vui" : {
"sar_height" : 1,
"sar_width" : 1
},
"MaxRefFrames" : 1,
"MinIdr" : 30,
"MaxIdr" : 300,
"i_scenecut_threshold" : 40,
"intra_refresh" : false,
"MaxBFrame" : 3,
"i_bframe_adaptive" : 1,
"i_bframe_bias" : 0,
"i_bframe_pyramid" : 2,
"b_deblocking_filter" : true,
"i_deblocking_filter_alphac0" : 0,
"i_deblocking_filter_beta" : 0,
"cabac" : true,
"interlaced" : false,
"constrained_intra" : false,
"tff" : true,
"fake_interlaced" : false,
"analyze" : {
"b_8x8" : false,
"b_i4x4" : false,
"b_i8x8" : false,
"b_p8x8" : false,
"b_p16x16" : false,
"b_b16x16" : false,
"weighted_pred" : 1,
"weighted_bipred" : true,
"direct_mv_pred" : 1,
"chroma_offset" : 0,
"me_method" : 1,
"me_range" : 16,
"mv_range" : 32,
"mv_range_thread" : -1,
"subpel_refine" : 2,
"chroma_me" : true,
"mixed_references" : false,
"trellis" : 0,
"psy_rd" : 1,
"psy_trellis" : 0,
"fast_pskip" : true,
"dct_decimate" : true,
"noise_reduction" : 0,
"psy" : true,
"intra_luma" : 11,
"inter_luma" : 21
},
"ratecontrol" : {
"rc_method" : 0,
"qp_constant" : 0,
"qp_min" : 0,
"qp_max" : 69,
"qp_step" : 4,
"bitrate" : 0,
"rate_tolerance" : 1,
"vbv_max_bitrate" : 20000,
"vbv_buffer_size" : 25000,
"vbv_buffer_init" : 0,
"ip_factor" : 1.4,
"pb_factor" : 1.3,
"aq_mode" : 1,
"aq_strength" : 1,
"mb_tree" : true,
"lookahead" : 10
}
}
Lastly this is just an example.. use what ever programs / tools you all want.
For example you can use ffmpeg.exe to compile a folder full of jpeg's into an .mp4 file.
Moto used this to make the default boot screen..
Yamb 2.1.0.0 http://yamb.unite-video.com
It's like Mp4box i think..
have fun video editor's and ...Share your work

hmm i guess nobody cares
So ya guys know i just learned about the Moto seasonal Logo's
It turns out when Moto wants to push a new boot animation to us they use an alternate directory.
that the moto boot services app loads..
this behavior was confirmed on the Moto-X
and i bet the same is happening on this phone.
We had a user in this sub-forum complain his boot animation / logo was changed on him over night.
i would bet this is hat happened.. an OTA pushed a temp logo out to select devices.
i seen on the web a soccer themed variant for example in celebration of some soccer event.
so.. what this means for us is we can add custom logo's to this new location and leave the original one untouched.
Simply download a compatible bootanimation.zip and put it in this folder on your phone.
/data/local/moodle
Credits to user named, mikeoswego for posting this info on the Moto-X sub-forum
And yeah i tested it myself too.
I restored the original bootanimation.zip in system/media/data
then i copied my custom Aqua Teen animation .zip to the moodle folder and rebooted.
It showed my custom .zip no the stock default .zip

Thanks for your knowledge. I was looking for this., I'll try on a weekend. Thanks.

NICEEEEE.... any anonymous ones please?

Software Development FYT DVR video files speed up and merge program

Hi People,
I wanted to merge the DVR footage from long trips and get a video file that is significantly sped up, and here's the crude result. Works for my T'eyes CC3, but should work for most FYT units.
Youtube video of resultant file
DashCam_Stitcher.zip <- Contains all executable and the Python script.
FFMPEG & FFPROBE need to be in your PATH for this to work
The program is a command line program, used in the format-> DashCam_Stitcher F 50
Where the source drive from the DVR is at F (in my case)
And the number of times the video footage is sped up is 50 (You can use 1, 2, 10, 100, or whatever)
Only for windows, and only for MP4. Hack my source if you want to change this.
That's the short version if you want to avoid my waffle below.
I driven my wife nuts trying to get this working, and so even though you might think you should thank me, if you do use this please comment with "Thank You Benja"
NOTE: This program will not automatically erase or fix any of your files. If that is needed, the program will prompt you with some helpful hints to fix yourself. I used someone’s demo program and was lucky I had a backup as it wiped my tf card.
Summary
I’ve used Python to script, FFPROBE to find all the corrupt files using a two-pass system, and you will be prompted to use FFMPEG to fix a few of those corrupt files (Or delete). And then automated FFMPEG creating sped up versions of each file in date time stamp order and FFMPEG merged them all into one big video file.
Waffle
If they don’t exist this program will create the following directories
C:\Users\[currentuser]\Videos\DashCam – Where the final merged file appear.
C:\Users\[currentuser]\Videos\DashCam\SpedUp – Where all the individually sped up versions of each file on your TF card will appear.
A sorted list of only the MP4 files is created
A first quick corrupt file check is done on each file using FFPROBE -hide_banner -v error eachfile.mp4. If bad MP4’s are found the program will stop, and ask you to fix them using FFMPEG -i xxxxx.mp4 -c copy xxxxxa.mp4 (or any method you can think of, including deleting the file).
Run dashcam_stitcher again to continue. If no bad files are found, a second much more thorough CPU intensive check is done. The second pass is because I have found files that slip though the first pass, and then kill the merge process at the end.
This second pass uses FFPROBE -show_entries stream=r_frame_rate,nb_read_frames,duration -select_streams v -count_frames -of compact=p=1:nk=1 -threads 3 -v 0 video.mp4
Which on a good file will show stream|25/1|299.000000|7475, but I have found some that cannot be repaired that read as stream|12000/1|131.160000|N/A or similar. If so, program stops again asking you to resolve as before.
Run dashcam_stitcher again to continue. If the video files get through both these checks, each file is then processed and put into the SpedUp folder. Then the program concatenates all the files together in order, and places a uniquely named (based on date and time) video file in the Dashcam folder.
Checking, encoding, and stitching the files together using this program will take a while, but much quicker than doing it by hand.
Again, if you can please comment with "Thank You Benja"

Very nice script.
However, I would use "-movflags faststart" to enable streaming (relocate the moov atom from the end of the file to the beginning of the file).
like
Code:
cmd = "ffmpeg -f concat -i "+concattextfile+" -movflags faststart -c copy "+outputmoviefilename
Of course: Youtube will do this automatically for you, but on your computer or from a server it will start streaming without the player first having to load the entire file
Edit: Sorry.
"Thank You Benja"

surfer63 said:
Very nice script.
However, I would use "-movflags faststart" to enable streaming (relocate the moov atom from the end of the file to the beginning of the file).
like
Code:
cmd = "ffmpeg -f concat -i "+concattextfile+" -movflags faststart -c copy "+outputmoviefilename
Of course: Youtube will do this automatically for you, but on your computer or from a server it will start streaming without the player first having to load the entire file
Edit: Sorry.
"Thank You Benja"
Click to expand...
Click to collapse
Thanks mate.
Links above have been updated to version 0.2. Your suggestion, and a minor change to file extension filtering.

If you can't get to the files. Big oops on my part.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Your file violates Google Drive's Terms of Service​‘DashCam_Stitcher.zip’ contains content that violates Google Drive's Malware and Similar Malicious Content policy. Some features related to this file may have been restricted. If you think this is an error and would like the Trust & Safety team to review this file, request a review below.
Restricted file

Google claims it found a virus. But my system seems clean. Windows defender is not having problems??

mjcummins said:
View attachment 5718109 Google claims it found a virus. But my system seems clean. Windows defender is not having problems??
Click to expand...
Click to collapse
I don't find viruses in your files. However, you use portable executables (pe32+). They are a known source of malware. That is also why ffprobe.exe and ffmpeg.exe are both 117MB. Maybe you get false positives on these exe files as they do contain all libs and necessary files inside and Google can't access them.

Well, I did put in the "Review Request" to almighty Google.
Are the files currently accessible? I own them, so my testing access doesn't really count I think.

mjcummins said:
Well, I did put in the "Review Request" to almighty Google.
Are the files currently accessible? I own them, so my testing access doesn't really count I think.
Click to expand...
Click to collapse
I use Linux laptops and Chromebooks. I can download them on my systems. Of course I can't use the windows exes and I need to modify your python script to make it linux compatible (or more universal).

surfer63 said:
I use Linux laptops and Chromebooks. I can download them on my systems. Of course I can't use the windows exes and I need to modify your python script to make it linux compatible (or more universal).
Click to expand...
Click to collapse
More universal would be good. Bit more code required to be able to use multiple systems. I have seen some examples, but not a good example yet. I've still got a lot of relearning and learning to get back up to speed.

If your process uses Python & FFMpeg, etc. than simply create scripts for Windoze, Linux, etc. making sure any folders created are relative to the users starting location.
Then people can simply install the official ffmpeg for their OS, and use your scripts. No need to create non-portable executables.

pwood999 said:
If your process uses Python & FFMpeg, etc. than simply create scripts for Windoze, Linux, etc. making sure any folders created are relative to the users starting location.
Then people can simply install the official ffmpeg for their OS, and use your scripts. No need to create non-portable executables.
Click to expand...
Click to collapse
In line 37 to 40 you could use:
Code:
userDir = os.path.expanduser("~")
userVideoDir = os.path.join(userDir, r"Videos")
dashcamdir = os.path.join(userVideoDir, r"DashCam")
spedupdir = os.path.join(dashcamdir, r"SpedUp")
This would immediately make it compatible on every by python supported platform, no issues with forward/backward path separators: python will do it for you. (And I like the fact that you do use raw strings, but in this case it is all ansi characters and nothing weird in there)
The source dir (sys.argv[1]), and thereby source_folder, would be an issue as Windows is the only system in the world using drive letters.
But in that case source-dir should be omitted and "source_path" should be used which could be "D:\DCIM" or "<sdcard name>/DCIM" Or something else if some exotic dashcam breaks all ISO-standards and uses something else than DCIM.
Code:
if platform.system() == 'Windows':
do something windows
else
do something all other OSes

surfer63 said:
Code:
userDir = os.path.expanduser("~")
userVideoDir = os.path.join(userDir, r"Videos")
dashcamdir = os.path.join(userVideoDir, r"DashCam")
spedupdir = os.path.join(dashcamdir, r"SpedUp")
This would immediately make it compatible on every by python supported platform, no issues with forward/backward path separators: python will do it for you. (And I like the fact that you do use raw strings, but in this case it is all ansi characters and nothing weird in there)
Click to expand...
Click to collapse
Or slightly different to show the "automatic power" of python.
Code:
userDir = os.path.expanduser("~")
userVideoDir = os.path.join(userDir, r"Videos")
dashcamdir = os.path.join(userDir, r"Videos", r"DashCam")
spedupdir = os.path.join(userDir, r"Videos", r"DashCam", r"SpedUp")
But maybe the first code block is probably more clear to any reader what is mentioned.
And then finally:
Code:
spedupdir = os.path.join(os.path.expanduser("~"), r"Videos", r"DashCam", r"SpedUp")
But this makes it really untransparent for the unexperienced reader/programmer "where we came from".
I prefer the extra steps, also for myself not to get lost and make errors.