phoboslab/pl_mpeg

Single file C library for decoding MPEG1 Video and MP2 Audio

pl_mpeg's Issues

Can it be added to ReactOS and WINE's quartz.dll?

subject

Corrupt decoding :(

Hi,
I was looking forward to using this in a m.a.m.e. project but for some reason it doesn't correctly decode the mpg files I've got.
The files are OK, they play with vlc and all other players I've tried including a hardware decoder :(
The decoded data is wrong from frame one :(
I've absolutely no idea how to fault finding on this :(
I've many files all with similar problems. This is the first frame from a file here.
https://drive.google.com/file/d/1_wrjTUzgTOf9hCEqX-G6tlmaoBSp1I2_/view?usp=sharing

Anyone any suggestions or ideas ? Any help is really appreciated.
Thanks
Paul

Segfault found by AFL

For fun I ran american fuzzy lop against this. Here's an MPEG that segfaults when pl_mpeg_extract_frames is run on it. Looks like a wild pointer dereference.

bad.mpg.gz

Corrupted video on GBA

Hello, great library. I am attempting to play a video on the GBA the library compiles fine with devkitpro. However everything decoded is corrupted. For example, my first frame should be completely black. But when grabbing a frame with the library on the gameboy advance it has randomish characters like 7e 6e 0f 0f 00 44 44 FF FF FF instead of being 0's. I can send the project to an email or upload here if interested. It would be awesome if you were. I did verify the same code on windows works. So I know it's GBA specific just not sure what.

Thanks!

Using mini_audio to play the audio?

I noticed that the example uses SDL2, which lets you "push" audio, I don't think miniaudio supports this - any way to allow miniaudio to pull the required pcm frames from the audio as it needs?

Streaming from the net introduces latency or requires separate thread

When feeding a plm_buffer() from the network (or slow media) there's no way to tell the buffer that no data is available at this time, but may be available later.

This forces you to either only decode a frame whenever you are sure that the data for the whole frame is available (as the video decoder can't pause in the middle of a frame) or to run the decoder in a separate thread and busy wait in the plm_buffer_callback until more data is available.

Making sure that enough data is available for decoding a full frame is not straight forward, because we don't know the size of a frame until it has been fully decoded, or we find the PICTURE_START code of the next frame. This introduces unnecessary latency for streaming.

The problem is described in more detail in this blog post towards the end.

This issue is meant for discussion of the problem and possible solutions.

sfml + pl_mpeg

Does anyone know of any examples of using pl_mpeg with SFML?

This library is awesome!

I featured this project in a talk I gave this weekend at Hackaday Supercon and used it as part of my badge hack to play video on a Raspberry Pi Pico. I've got a short demo video at https://nycr.social/@unwiredben/111356286237886104 and my badge hack is visible on the Supercon livestream at https://www.youtube.com/live/11Js0cOif4c?si=qjjBUfFO0E1LFGzw&t=2036. Thanks so much for sharing this code!

Corrupted video

When playing video sometimes I get corrupt blocks on the right hand side.
When played in VLC media player it is not corrupted.

See example here:

Source file here:
https://www.mdawson.net/files/termcrop.mpg

Frame number 980
I wonder if it's something to do with certain motion blocks on the right hand side?

(Minor suggestion) Compiler warnings

Lines 1187 and 1872 involve an assignment within a conditional. Some compilers complain about this. Easy fix: explicitly adding "!= 0" in each conditional appears to work equivalently and silences the warnings.

Small encoder lib suggestions?

Thanks a lot for doing this!! This lets me use video in my apps without the hassle of gigantic libs dependencies!

Now the only open question is... how can I encode too, without depending on FFmpeg? I searched and found some possibilities, like ezMPEG, or jo_mpeg, but, it seems neither of them support P frames, so they encode every frame as an independent image (I frame), and besides, they don't encode audio either...

Do you have any recommendation for a small (yet with reasonable quality) MPEG1 encoder with audio that would work with your decoder?

Another segfault found by AFL

Run through pl_mpeg_extract_frames and observe memory corruption.

bad2.mpg.gz

What a fantastic library

It took me 30 minutes to integrate into my project https://github.com/QuantScientist/TorchRayLib
Once I am done, I shall credit you of course.

System and VideoCD streams

Does it support them?

Windows Builds of the PL_MPEG Player

It would be nice if 32 and 64-bit builds of the player were made available for downloading.

Memory corruption found by AFL

Running pl_mpeg_extract_frames causes memory corruption. AFL found this one too.

bad3.mpg.gz

Custom malloc/realloc/free and disable stdlib

Sometimes we don't have access to the standard library, and/or we want to use custom allocation functions. Thus, it would make sense to add in some defines that can be set by the user of the library before the header is included. These would define memory allocation functions. There will also be a define that controls whether stdio is used, and also remove the *_create_with_file functions to prevent compilation errors.
stb_image is an example implementation of this.
Of course, it would be trivial for the user to manually modify the source code, and the license is permissive enough. However, it is better to have this be done in a well-defined manner, signed off by someone who has a more intimate knowledge of the code, and therefore is less likely to make mistakes than some impatient newbie running %s/malloc/my_malloc.

Divide-by-zero if macroblock width is zero

AFL found this one. Here's the test case it found.

bad4.mpg.gz

Any chance of a seek API?

I'm using this library with raylib to play MPEG files, it works flawlessly! In under 150 lines of C I got a nice video player!

Just wondering if there is any plan to add a frame/sample seek API. It would be really great.

(I can keep decoding frames/samples to implement a seek mechanism but it seems a bit inefficient)

EDIT: Just created a function to get video length but it's extremely inefficient, it takes up to 10 seconds on my system. I can probably execute it in a second thread while starting playing the video in main thread... but... any better idea to get video length?

int plm_get_total_frames(const char *filename)
{
    int total_frames = 0;
    
    plm_t *plm = plm_create_with_filename(filename);

    if (plm != NULL)
    {
        plm_frame_t *frame = plm_decode_video(plm);
        while (frame != NULL)
        {
            total_frames++;
            frame = plm_decode_video(plm);
        }

        plm_destroy(plm);
    }

    return total_frames;
}

Modify public struct typedefs to enable fwd declaration in C++

It would be helpful to change

typedef struct {
...
} plm_frame_t;

into

typedef struct plm_frame_t {
...
} plm_frame_t;

This would make it easier to forward-declare the structs in a C++ header without including pl_mpeg.h from there, exposing it to the rest of the program.

Example my.hpp:

struct plm_frame_t; // trying to not include pl_mpeg.h here

class MyClass {
public:
    void render( plm_frame_t *frame );
};

Example my.cpp:

#include "my.hpp"
#define PL_MPEG_IMPLEMENTATION
#include "pl_mpeg.h"

MyClass::render( plm_frame_t *frame ) {
...
}

Duplicate lines in plm_buffer_has method

There seem to be some duplicate lines in plm_buffer_has method:

int plm_buffer_has(plm_buffer_t *self, size_t count) {
	if (((self->length << 3) - self->bit_index) >= count) {
		return TRUE;
	}

	if (self->load_callback) {
		self->load_callback(self, self->load_callback_user_data);
	}

	if (((self->length << 3) - self->bit_index) >= count) {
		return TRUE;
	}
	
	if (self->total_size != 0 && self->length == self->total_size) {
		self->has_ended = TRUE;
	}
	return FALSE;
}

Green rectangles when decoding

I'm porting pl_mpeg_player.c to GLFW and it decodes smoothly and accurately, except for these rectangles which appear in the areas of the video which are updating. (The colors are accurate where there is no movement.)

I'm hoping that someone knows at a glance what I'm doing wrong. Is there a step that, when skipped, results in these artifacts? The same artifacting appears regardless of whether I'm using the three-texture YCRCB approach or the one-texture RGB approach; they work equally well and give the same result.

I'd like to thank the creator for his hard work; this is a fantastic library with a ton of utility for creators such as myself.

pl_mpeg + physfs = possible?

Hi, I'm looking around for a video playback solution and considering pl_mpeg. Can it be made to work with physfs?? I'm currently using physfs to load resources from a zip archive, so I need to be able to also play the videos from this archive as well. At present I don't see a way to easily do it. Maybe I've overlooked something, please advise.

Thanks

Jarrod

Decode slice macroblock loop condition may be incorrect

When looping to decode macroblocks you call plm_buffer_no_start_code, which is an aligned check for no start code.
I think this can lose undecoded bits in the stream. In the spec it seems they do an unaligned bitcheck for 23 0's.
I believe if there were no more macroblocks and the stream wasn't aligned it would be padded with 0's.
What do you think?