MP3 meta data over http response - php

This is going to sound like an odd request.
I have a PHP script pulling a mp3 stream from SoundCloud and repeating the stream with the correct headers to allow WinAmp to play the file. But it only shows the local url I have the script running from. Before anyone asks, I am injecting ID3v1 into the file before echoing it.
Is there any way to provide WinAmp with the meta data from php?

Just to clarify, you are effectively proxying an MP3 file from SoundCloud, and you want to embed metadata into it?
Winamp will pick up ID3 tags in an HTTP-served MP3 file. However, if you are using ID3v1, those tags don't exist until the very end of the file. If you want the file to be identified without having to download the whole file, you must use ID3v2 which are typically located at the beginning of the file. (I actually recommend using both ID3v1 and ID3v2 for broader player compatibility, but almost everything supports ID3v2, so it is your choice.)
Now, there is another method but if you use this method the metadata won't be saved in the file when downloaded. You can use SHOUTcast-style metadata. Basically, Winamp and other clients (like VLC) send a request header, Icy-MetaData: 1. This tells the server that it supports SHOUTcast-style metadata. In your server response, you would insert metadata every 8KB or so. Basically, you want the reverse of what I have detailed here: https://stackoverflow.com/a/4914538/362536
In the end, simply adding ID3v2 tags will solve your problem in the best way, but I wanted to mention the alternative option in case you needed it for something else.

Related

Can I serve media to web browser from MongoDB

I will try to keep this simple without lots of extra information. I have been investigating MongoDB and I believe it will work well for my next project. There is one thing I am fuzzy about though: storing and retrieving files (chunks) from GridFS.
Lets take a CMS for example. If I wanted to display (output via the browser) an image, I would go my MySQL database, find the key and pull the metadata that would include the file source on the File System and then display an image tag with that href. I know that I can store image/video/etc files in Mongo beautifully and I can retrieve binary, but if I wanted to display that file (push it to the browser) would I have to write the contents to a temporary file and then echo my img tag with the href? That can't be more efficient.
I feel like I'm missing something. For this circumstance, is MongoDB any better?
For clarification: I'm using PHP and Apache on a typical LAMP stack (development not production) and working on a platform to enable creative collaboration between artists. So, I would have several artists collaborating on the same files, and I would like to be able to search inside those files, index them, keep all metadata together and employ sharding. I really seems like MongoDB is the way to go.
Thanks!
Michael
If you want to store binary files in the database, you can serve them directly without writing them to a filesystem first. All you have to do is write a PHP script (or whatever) to send the data in response to an incoming HTTP request, with the proper media type in the response HTTP header.
So if we're talking about images, in your HTML you'd do something like this:
<img src="/your_img_script.php?image=1234"></img>
Then just write your_img_script.php to lookup image ID 1234 and dump the data out with a image/jpeg (or whatever) content type header.

Embed code in video file

I'm sorry if the question is ambiguous, I'll try to explain.
I'm working on an existing PHP download script for videos and some parts of it are broken. There's code in there that's supposed to place a specific member code inside the video file before download, but it doesn't work. Here's the code:
//embed user's code in video file
$fpTarget = fopen($filename, "a");
fwrite($fpTarget, $member_code);
fclose($fpTarget);
$member_code is a random 6-character code.
Now, this would make sense to me if it were a text file, but since it's a video file, how could this possibly work and what is it supposed to do? If the member code is somehow added to the video, how can I see it after download it? I have no experience with video files, so any help is appreciated (a modification of the available code or new code would be equally welcome).
I'm sorry I can't give a more precise description of what the code is supposed to do, I'm trying to figure that out myself.
It may work, depending on the format/type of the video. MPG files are fairly tolerant of "noise" in a file and players would skip over your code because it doesn't look like valid video frame data.
Other formats/players may puke, because the format requires certain data be at specific offsets relative to the end of the file, which you've now shifted by 6 characters.
Your best bet is to figure see if whatever format you're serving up has provisions for metadata in its specifications. e.g. there might be support for a comment field somewhere that you can simply slap the code into.
However, if you're doing all this for 'security' or tracking unauthorized sharing of the video, then simply writing the number into a header is fairly easy to bypass. A better bet would be to watermark the video somehow so that the code is embedded in the actual video data, so that "This video belongs to member XYZ only" is displayed while playing.
You don't write to the content of the file directly, not like you would with a text file. As you've noticed, this effectively corrupts the video and you have no way of reasonably reading the information.
For audio/video files, you write to meta-data that's packaged with the file. How this is packaged and what you can do with it generally depends heavily on the container format used for the file. (Remember that container and codec are two different things. The codec is the format used to encode the audio/video, the container is the file format in which that data stream is stored.)
A library like getID3 might be a good place to start. I've never used it, but it seems to be what you're looking for. What you would essentially do is write a value to the meta-data in the container (either a pre-defined value for that container or maybe a custom key/value pair, etc.) which would be part of the file. Then, when reading the file, you can get that data. (Now, that last part depends heavily on what's reading the file. The data is there, but not every player cares about it. You'll want to match up what you're writing to with what you usually see/read from the file's internal meta-data.)

Grabbing Information from a Remote MP3 File

I'm trying to extract MP3 ID3 Tag information from any remotely hosted mp3 file via PHP.
I've tried several times with fread and fseek to grab the last 128bytes of the file and extract the tag information...but had no success.
Here's and example of what I want to do:
http://www.digitalcoding.com/tools/read-remote-mp3-avi-file-id3-information.html
Any tips or code snippets would be appreciated :)
With HTTP you'd download a range of data from the resource. You can pick an arbitrary range and it can even consist of multiple parts like "X-Y" or "X-Y,A-B" etc.
Servers however don't always support that and then you'll be served the full file instead.
When using something like curl, you can use its CURLOPT_RANGE option (or the command line tool's -r).

Inline <video> large file with buffer

I'm trying to build a site using HTML5's video tag so that I can share some movies I have made. Their sizes are pretty big (>500 MB), and when I watch them from outside my network, it seems like it's trying to download the whole thing before showing it. I'm wondering how I can make it so that they can be downloaded and watched at the same time.
I'm using php and javascript to build the site, although if there are libraries or techniques available in other languages, I'm more than happy to hear about them.
Video files on the web sometimes need to be encoded in a special way in order for them to be played while downloading. In order for flash based videos to work, data called "moov" must be moved from the end of the stream to the start. A program called mp4 FastStart can do this for you.
Programs like HandBrake have a "web" option that also does this when encoding. The data basically contains the length of the video, etc. Typically this was at the end of the file. However when the web came along that meant downloading the entire thing before being able to play.
Can you tell us what format the video is?

is there anyway to hide swf path?

Is there anyway to hide .swf path showing up from webpage?
you can use a php script something like getswf.php?name=flash.swf in your flash tags. Then create getswf.php script to respond with output of flash.swf file, and keep flash.swf file in a directory outside of public directory.
There may be ways to make it difficult to view. However, nothing you do can stop an intelligent adversary from using a tool like Fiddler to monitor web traffic and undo all your obfuscation.
I think this is a fairly pointless exercise. Any resource that gets sent to a browser, be it an image, sound, flash movie, even flv files loaded by the YouTube player, can all be saved to disk fairly easily.
As Justin mentions, Fiddler can achieve this easily.
As others answered it there are ways to do it using your script. If you are looking for a paid option check this out Media Vault.
Hiding the URL to the swf file might be quite a challenge but there are other things you can do if you're wanting to more closely protect the video/data being displayed by the swf.
I'll run through a couple of ideas in the order I think them most obfuscated with the least first. Bare in mind that most of these techniques merely make it harder to get to the information/video rather than making it impossible to obtain.
The main idea most sites tend to follow is that of having the swf as a player and the content in another file somewhere else, usually an flv or mp4 etc.
Add flv location through Javascript
This technique is as basic as it sounds. You have your swf player on the page and pass a new variable too it (such as 'file') with the location to the flv file using Javascript. If you're already loading your content with some kind of JS flash module then all the easier to begin implimenting.
Obfuscating flv location through XML
Another techniqe I've seen used quite recently is that of having an XML document as a paremeter to the swf player and then the flash player itself resolves the URL of the flv from a node in the XML. It's easy to get to the flv URL if you want to but it does make it that little bit harder.
Token access
This technique can be used in conjunction with any of the above two. You basically ensure that your flv files can only be accessed with the use of a special token otherwise the page returns a HTTP error. The token would be understood by the flash player and the server and upon the player making a request for the flv, a token must be included (usuallu the token itself is obfuscated in some way that it cannot be easily mimmicked through a simple GET request).
Domain access
Very similar to the above however in this case, the flv file will only be loaded when the requesting URL is a specific site. All other requests will be denied (such as directly hitting the flv location in your browser.
As stated above, none of these methods make it impossible to get hold of your flash material. If it's on the web (or any network for that matter), it's possibly a target. You'll usually find for most things that making it harder to obtain will deter a lot of those who would otherwise have been privy to downloading your content.
Completely hiding the URL to the swf
If your only criteria is to hide the URL then hiding it behind a URL rewrite is the best option I can think of.
Your swf might be at /location/flash/player.swf?file=summer.flv and then you could do a URL rewrite to something like /vacations/summer2011/.
This way the URL to the swf is completely hidden away and this should satisfy your desire to hide the swf path.
The answer is NO.
You may not want to believe it. but it is a fact. you can do all you like to obfuscate it. but the browser needs to find it. if the browser can find it so can anyone else.
A server side script that acts as a loader will hide the real path to the file, but to what purpose. the end result is the file is still available.
If you want a simple answer, so simple people can't find it then the people here have given you some suggestions. obfuscation is the best you can hope to achieve.
Alternately only allow approved users access to the file. that way they need to log in to get access to it. but if you want it available publicly then its well, public!
DC

Categories