clever FFmpeg-GUI, a lightweight and portable GUI for FFmpeg

[ProWo]

Newest beta (v3.1.6.01), online now.

Added:
Can handle files without duration info (automatic fix).
Mux pcm_alaw in 3gp, avi, mkv container.
Select and order files in join grid.
Automatic generation and insertion of chapters at merging points (join grid).

Fixed:
Bug in muxing code.
Bug in video analysis.

[ProWo]

----------------------------------------------------------------------------------
November 3, 2023 release version 3.1.7.
----------------------------------------------------------------------------------

New:
Can handle files without duration info (automatic fix).
Mux pcm_alaw in 3gp, avi, mkv container.
Select and order files in join grid.
Automatic generation and insertion of chapters at merging points (join grid).

Fixed:
Bug in muxing code.
Bug in video analysis.
----------------------------------------------------------------------------------

[mannequin80]

3.1.7 (and v3.1.6) crashes on sub extraction.

[ProWo]

3.1.7 (and v3.1.6) crashes on sub extraction.

What sub type?

[mannequin80]

3.1.7 (and v3.1.6) crashes on sub extraction.

What sub type?

subrip srt.

also, need to double check but truehd 7.1 to eac3 5.1 (downmix) results in a 5.0 file (haven't experienced that prior to 3.1.7).

checked: 3.1.5 produces a proper 5.1 file.

[ProWo]

3.1.7 (and v3.1.6) crashes on sub extraction.

What sub type?

subrip srt.

also, need to double check but truehd 7.1 to eac3 5.1 (downmix) results in a 5.0 file (haven't experienced that prior to 3.1.7).

checked: 3.1.5 produces a proper 5.1 file.

For testing I downloaded clever FFmpeg-GUI v3.1.7 from videohelp.com and tested both mentioned problems, but could not reproduce them. Everything works as expected, subrip subtitles successfully extracted, audio conversion from truehd 7.1 to eac3 5.1 successfully performed (6 channels.)

Please test it again and report back here.

[mannequin80]

Please test it again and report back here.

had to delete user.configs and start from scratch but all is good now, working as it should.

[NewTwoVideo]

Hello Everyone,

I have been using Clever FFmpeg GUI to convert 1080/59.94p (PSF) AC-3 mts files from my Panasonic camcorder to 1080/29.97p AAC-LC mp4 files.
My videos are just family memories and I am just trying to make them easier to view using a PC instead of connecting my camcorder to the TV.

Clever FFmpeg GUI works great and is easy to use once you get the workflow memorized.
I deeply appreciate ProWo's efforts to create this program.


I was wondering, since Clever FFmpeg GUI has so many options for output,
WHAT IS THE BEST STORAGE CONTAINER / CODEC / FILE FORMAT / OPTIONS / PARAMETERS etc. to save / archive the mts files in a way that takes up less disk space?
As you know mts files are pretty big, around 150 ~ 200 MB per minute.
Most alternatives are smaller. But what combination of settings will preserve the most information and quality of the original?

I know that mts files have a timestamp in each frame. Preserving that would only be needed if I was documenting something for a situation like an insurance claim for storm damage, so in that instance I think I would save the original mts file.

So if that is not a factor, what are the best settings to save space but still retain the quality of the video and audio information?


TIA

[ProWo]

WHAT IS THE BEST STORAGE CONTAINER / CODEC / FILE FORMAT / OPTIONS / PARAMETERS etc. to save / archive the mts files in a way that takes up less disk space?
As you know mts files are pretty big, around 150 ~ 200 MB per minute.
Most alternatives are smaller. But what combination of settings will preserve the most information and quality of the original?

I know that mts files have a timestamp in each frame. Preserving that would only be needed if I was documenting something for a situation like an insurance claim for storm damage, so in that instance I think I would save the original mts file.

So if that is not a factor, what are the best settings to save space but still retain the quality of the video and audio information?


TIA

Every video has a timestamp in every frame, otherwise it could not be played.
The timestamp has nothing to do with quality.

Likewise, the different containers (mp4, mkv, mts, m2ts, ts, vob, mpeg, mpg, avi, flv, webm, wmv, ogg, mov and others) have nothing to do with quality, they are just containers for the video, audio and subtitle tracks that contain the information.

The quality is determined by the bitrate, the higher the better.
But not unlimited.

For example, a bitrate of 200mb/s for 1920x1080 video is certainly too high for normal use.

For normal use (TV, projector), 3-5mb/s is more than sufficient with modern codecs for video such as h264(avc), h265(hevc).
For audio, aac, ac3 64kb/s per channel is sufficient.

So you have to decide what you want to use your videos for and choose accordingly.

[NewTwoVideo]

Thank you for your reply.

Does Slow vs Fast add anything to the quality of the converted file in this circumstance ?
What CRF would you recommend ?

[ProWo]

Thank you for your reply.

Does Slow vs Fast add anything to the quality of the converted file in this circumstance ?
What CRF would you recommend ?

Fast results in slightly larger files than slow (but the difference is not really big).
Take CRF 18-20 for x264 and 23-25 for hevc, with very good quality.
x264 is universally applicable and is accepted by all devices, which is less the case with x265.

[NewTwoVideo]

Thank you for your reply.

Does Slow vs Fast add anything to the quality of the converted file in this circumstance ?
What CRF would you recommend ?

Fast results in slightly larger files than slow (but the difference is not really big).
Take CRF 18-20 for x264 and 23-25 for hevc, with very good quality.
x264 is universally applicable and is accepted by all devices, which is less the case with x265.

Great ! Thank you.

[jraju]

Hi, prowo,
could you add features, in such a way, that the same videos of two resolutions , converted or resolution changed video, or any of the two same videos and audio is analyzed between the two and give the best among the two. For eg, if a downloaded video is in lesser resolution, user tend to increase the resolution to fit his media, but with some losses. Some videos may still look good in increased resolution without much loss and some with much loss. kind of comparing two videos . separately or side by side, or one after another play. it is only a suggestion. using your software.

[ProWo]

Hi, prowo,
could you add features, in such a way, that the same videos of two resolutions , converted or resolution changed video, or any of the two same videos and audio is analyzed between the two and give the best among the two. For eg, if a downloaded video is in lesser resolution, user tend to increase the resolution to fit his media, but with some losses. Some videos may still look good in increased resolution without much loss and some with much loss. kind of comparing two videos . separately or side by side, or one after another play. it is only a suggestion. using your software.

I have no idea how this could be achieved.

[ProWo]

Newest beta (v3.1.7.01), online now.

Added:
Framerate change without recode (video stream only).
Mux grid separate default and forced checkbox.
Muxgrid default checked for all streams.
Edit chapters in join grid.

Fixed:
Bug in mux grid sorting.
Bug in audio delay analysis (mux grid).
Bug in chapters handling (join grid).

Improved:
raw h264 and hevc stream handling.

[Ronstang]

I like the program but I need it mainly to re-encode the audio and then remux it....works great but why can't I change the bitrate of the chosen audio encoder? There is a white box that has a bitrate in it, but no option to choose a different bitrate and if I just delete that bitrate and type in a supported one, which it allows me to do, nothing happens and the audio is always encoded at what appears to be a default bitrate. Can you please change this? I want to use a different bitrate and other programs allow the bitrate to be changed but they don't do the remux too....so your program would be perfect for my needs if I could chose the bitrate.....well and batch processing LOL.

[ProWo]

I like the program but I need it mainly to re-encode the audio and then remux it....works great but why can't I change the bitrate of the chosen audio encoder? There is a white box that has a bitrate in it, but no option to choose a different bitrate and if I just delete that bitrate and type in a supported one, which it allows me to do, nothing happens and the audio is always encoded at what appears to be a default bitrate. Can you please change this? I want to use a different bitrate and other programs allow the bitrate to be changed but they don't do the remux too....so your program would be perfect for my needs if I could chose the bitrate.....well and batch processing LOL.

The minimum recommended bit rate for the selected configuration is displayed in the text box.
If you enter a higher value here, this value will be used for encoding.
You can easily check this in the output file with Mediainfo.
No bitrate is displayed for aac, as aac is known to do variable encoding. However, you can see from the size of the file that the higher value has been used.

As far as batch is concerned, this is not an option for me, as I think that each video needs different interventions and mass processing does not lead to an optimal result.

[Ronstang]

As far as batch is concerned, this is not an option for me, as I think that each video needs different interventions and mass processing does not lead to an optimal result.

It could be an option if you set it up like MKVToolNix where you make all the individual choices per file and then add them to the job queue that you execute after your'e done queueing up jobs. It takes a while to encode and if I have two dozen encodes I need to do the audio on I don't want to sit here and wait for each to be done before doing the next. I'd rather choose all the indiviidual settings and queue them up and leave them be until they are done....or have them run while I sleep if there are a lot.

[Ennio]

Hi ProWo,

Thanks for adding framerate change without recoding, but I'm a bit confused. Do I correctly understand that 'Various' two functions "Fix Timestamps" and "Change Framerate" are basically the same? I.e both based on rewriting timestamps?

When I import a 23.976fps elementary h264 stream and go "Change Framerate" to 24fps, I notice something weird when checking with MediaInfo.

The intermediate ***_23976_TSF.mp4 file - properly - shows one line: Frame rate: 23.976 (24000/1001) FPS

But the resulting ***_TSF_24FPS.mp4 file shows two lines:

Frame rate: 24.000 FPS
Original frame rate: 12.0000 FPS

In my experience - please correct or add - MediaInfo does this where videotiming information on container level differ from that of elementary stream.
Can you please look at it?

- -

When doing a 23.976 --> 24 fps framerate change with an elementary hevc stream, it seems that only the (correct) intermediate ***_23976_TSF.mp4 file is created. But the 24fps mp4 endresult seems missing. GUI does confirm "Task succesfully completed". Is it possible the outcome is written somehere unknown to me? I also tried after removing user.config file.

I also want to ask if it's possible to implement choosing between mp4 and elementary stream output for the functions under 'Various'.

Another thing: can you show me a working example of a commandline for ffpmeg.exe to rewrite timestamps of an elementary AVC stream?

[Ronstang]

The minimum recommended bit rate for the selected configuration is displayed in the text box.
If you enter a higher value here, this value will be used for encoding.
You can easily check this in the output file with Mediainfo.
No bitrate is displayed for aac, as aac is known to do variable encoding. However, you can see from the size of the file that the higher value has been used.

OK I got that to work and I was able to encode the DTS track to 640 kb/s but it took 9 minutes and 40 seconds just to create the track whereas Avidemux does the same thing in 2 minutes 15 without the need to multiplex the audio back in....but it destroys the subs so your program is more desirable but I don't understand the HUGE time difference since both programs are using ffmpeg. Any idea? Almost 10 minutes to convert audio without batch ability is painful. I can't sit here all day to convert 20 files.

Also, once I multiplex the file and the audio is in the container if I extract just the audio it shows up as a size of 0??

[ProWo]

Hi ProWo,

Thanks for adding framerate change without recoding, but I'm a bit confused. Do I correctly understand that 'Various' two functions "Fix Timestamps" and "Change Framerate" are basically the same? I.e both based on rewriting timestamps?

No, the two are not the same thing.
Fix Timestamps is for h264/hevc videos that show 3 framerate infos in mediainfo: framerate, min_framerate, max_framerate. This state is corrected to a constant framerate, whereby the whole file is corrected, including audio and subtitles.

When I import a 23.976fps elementary h264 stream and go "Change Framerate" to 24fps, I notice something weird when checking with MediaInfo.
The intermediate ***_23976_TSF.mp4 file - properly - shows one line: Frame rate: 23.976 (24000/1001) FPS
But the resulting ***_TSF_24FPS.mp4 file shows two lines:
Frame rate: 24.000 FPS
Original frame rate: 12.0000 FPS

This happens when you read in raw h264/hevc.
The file analysis reads r_frame_rate and avg_frame_rate from the video stream and both these values are incorrect.
In the window where you are asked to enter the FPS, you cannot leave the default value (r_frame_rate), but you have to enter the correct frame rate, in your case 23.976.
Then the rest will work as expected.
In the next beta I will add a raw test, which will leave the FPS field empty.

When doing a 23.976 --> 24 fps framerate change with an elementary hevc stream, it seems that only the (correct) intermediate ***_23976_TSF.mp4 file is created. But the 24fps mp4 endresult seems missing.

This is a bug. Will be corrected in the next beta.

I also want to ask if it's possible to implement choosing between mp4 and elementary stream output for the functions under 'Various'.

This should be possible. Will be implemented in the next beta.

Another thing: can you show me a working example of a commandline for ffpmeg.exe to rewrite timestamps of an elementary AVC stream?

Here you go:

FFmpeg.exe -i input.h264 -c copy  -bsf:v h264_metadata=tick_rate=25000/1000 output.h264

For raw streams only (example any to 25 fps).

[ProWo]

Newest beta (v3.1.7.02), online now.

Added:
Video stream raw export for h264 and hevc.
Audio encoding coced bitrate range info.
Extracted files are available in the muxgrid.

Fixed:
Bug in muxstreams handling.

Improved:
Stream info handling.
Many code improvements.

[Ennio]

Thanks for your explanations and update, ProWo.

[ProWo]

----------------------------------------------------------------------------------
November 17, 2023 release version 3.1.8.
----------------------------------------------------------------------------------

New:
Batch processing.

Fixed:
DTS encoding bitrate range.

Improved:
Mux grid streams handling.
Many important code improvements.
----------------------------------------------------------------------------------

[ProWo]

Newest beta (v3.1.8.01), online now.

New:
Automatic definition of the output file in the mux grid if the muxed streams do not belong to the source file.

Fixed:
Aborted trigger bug
Mux container now corresponds to the source file.

Improved:
Preprocessed and batch streams handling in mux grid.

[FTG]

Newest beta (v3.1.8.01), ...

Hello, would it be possible to make the program doesn't alter the "original"
source parameters from any given audio track, while trying to encode it? I mean,
I usually use the program to change the framerate from different audio tracks,
but I will always have to modify the bitrate, frequency and encoder values, to match
those from the source. By the way, in case of the "AC-3" encoder, I think the "DRC"
value should be set to "off" by default, since that's just an "optional" parameter.

Is there some known limitation to deal with "DTS-HD MA" elementary audio tracks?
Since I'm trying to extract the "core" from one of them, but I'm getting a "No Duration found!"
error message, and the programs ask me about its "FPS" value and to encode it using a "MP4" container.

Thanks.

[ProWo]

Hello, would it be possible to make the program doesn't alter the "original"
source parameters from any given audio track, while trying to encode it? I mean,
I usually use the program to change the framerate from different audio tracks,
but I will always have to modify the bitrate, frequency and encoder values, to match
those from the source..

This has already been implemented. Codec, channels and bitrate are taken from the source. If the bitrate is variable, a default bitrate is suggested. The frequency is deliberately not taken, as in many videos the audio tracks have 44100Hz, although the standard for videos is 48000Hz. With DTS the codec is not adopted, this is a bug and will be corrected.

By the way, in case of the "AC-3" encoder, I think the "DRC"
value should be set to "off" by default, since that's just an "optional" parameter.

FFmpeg default setting for AC3 DRC is on.

Is there some known limitation to deal with "DTS-HD MA" elementary audio tracks?
Since I'm trying to extract the "core" from one of them, but I'm getting a "No Duration found!"
error message, and the programs ask me about its "FPS" value and to encode it using a "MP4" container.

Since the last release, elementary audio streams are no longer queried for duration.
However, if the dts stream is muxed with a hevc video stream, for example, and no duration information is included, then the query for the FPS is made. However, nothing is encoded, but the video is only muxed into an mp4 container and then read in again.
During this muxing to mp4, the missing duration information is rewritten.

[FTG]

I'm having a weird problem with an audio track, basically, I'm trying to convert an "AC3" file,
keeping the same quality parameters as in the source (192 kbps, 44 kHz, no DRC), and
the resultant files' synchronization will be way different not only if the "FPS" values are
changed (which is expected), but even when I change the "frequency" values, which makes
no sense at all to me. When I leave the program's default value of "48 kHz", the synchro will
have according to what I'm looking for; but if I change the kHz value to match the original file's
(44 kHz), the synchro will be altered by just that mere change.

[ProWo]

I'm having a weird problem with an audio track, basically, I'm trying to convert an "AC3" file,
keeping the same quality parameters as in the source (192 kbps, 44 kHz, no DRC), and
the resultant files' synchronization will be way different not only if the "FPS" values are
changed (which is expected), but even when I change the "frequency" values, which makes
no sense at all to me. When I leave the program's default value of "48 kHz", the synchro will
have according to what I'm looking for; but if I change the kHz value to match the original file's
(44 kHz), the synchro will be altered by just that mere change.

Will check this and come back to you.

[FTG]

[QUOTE=ProWo;2713378]

This has already been implemented. Codec, channels and bitrate are taken from the source. If the bitrate is variable, a default bitrate is suggested. The frequency is deliberately not taken, as in many videos the audio tracks have 44100Hz, although the standard for videos is 48000Hz. With DTS the codec is not adopted, this is a bug and will be corrected.

OK, I'm using v3.1.5, and in the case of the specific file I'm dealing with now, it is an "AC-3" stereo, 44.1 kHz, 192 kbps (constant),
which when is processed by the program, is set to use "48 kHz", and "160 kbps" by default. Same encoder value is respected.

[QUOTE=ProWo;2713378]

Since the last release, elementary audio streams are no longer queried for duration.
However, if the dts stream is muxed with a hevc video stream, for example, and no duration information is included, then the query for the FPS is made. However, nothing is encoded, but the video is only muxed into an mp4 container and then read in again.
During this muxing to mp4, the missing duration information is rewritten.

OK, I will try with the last beta version.