bum51's dashboard

bum51 replied on April 24, 2009 01:29 to the question "Playing video in transparent mode" in Schillmania!:

for what I said the "To:" should actually be
(_s.wmode?'<param name="wmode" value="'+_s.wmode+'" /> ':'')

bum51 started following the question "Is it possible to play video in full screen video?" in Schillmania!.

bum51 replied on April 23, 2009 04:24 to the question "Playing video in transparent mode" in Schillmania!:

I'd just like to note that that's not the only bug pertaining to that issue:
At line 604 you do this:
(_s.useHighPerformance && !_s.useMovieStar?'<param name="wmode" value="'+_s.wmode+'" /> ':'')
when it should be this (or similar):
(_s.wmode?'<param name="wmode" value="'+_s.wmode+'" /> ':'')
This way checks to see if the parameter is needed, since the value has already previously been set

bum51 replied on April 23, 2009 04:05 to the problem "Videos interfere with waveform and eq datas" in Schillmania!:

Ahh yes! that would make sense! The video host we are using is separate to conserve bandwidth.
I completely never thought of that though! Fortunately there's an off switch for the visualizations anyway so when I run into any sort of error I just turn them off.

bum51 reported a problem in Schillmania! on April 22, 2009 04:58:

Videos interfere with waveform and eq datas

Not that this is a big deal, as I don't expect many people to run into this issue, but for my site I have both mp3 files and video files. I typically display a visualization when playing the mp3s, but when you switch to a video file it displays the video instead (I'm impressed with this feature of soundmanager BTW).
Now this itself works seamlessly. The problem is when the user switches back to regular mp3s, there are no longer any visualizations, as there seems to no longer be any "data".
For now it's only an experimental feature anyway, so I can warn users about it, but I just thought that you should know about this use case. NOTE: I only tested it in Firefox 3 so far.

Kudos as always for the amazing API

bum51 started following the idea "HTML5 Audio" in Schillmania!.

bum51 replied on February 24, 2009 06:45 to the problem "Soundmanager crashes if a parent of the sm2-container element changes class names." in Schillmania!:

After a bit of testing, it seems that if I change the height through JS instead of the CSS, I can even go to height:0px and it still works just fine =)
I'll need to test a little bit more but this is really good news for me!

bum51 replied on February 24, 2009 06:40 to the problem "Soundmanager crashes if a parent of the sm2-container element changes class names." in Schillmania!:

In my case I don't do a display:none, but instead a height:0. And I certainly make sure not to change the DOM to affect it at all ;-).
In essence it's the same result, but for some reason it only started happening when I updated SoundManager.
I was using V2.78a.20080920 before I updated to the latest, which has some significant changes. My guess is that it is somehow related to the MovieStar feature that came in afterward (awesome addition BTW).
For now as long as I can reposition it it should be fine, but I'll post here if I find a good solution though for sure.
If I can at least change the size of the container element on-the-fly to a non-zero number then I should be able to at least handle it in the JS if not the CSS, which is ok by me.

Thanks for the advice =)

bum51 reported a problem in Schillmania! on February 23, 2009 04:39:

Soundmanager crashes if a parent of the sm2-container element changes class names.

As the title says.

I make the sm2-container in my special location on the page and when css classes change above it, it breaks soundmanager.

One note is that I am hiding the parent, by changing the height of the element as part of the css class. And it seems to be specific to Firefox, though I didn't do much testing.

As far as things go it's not a problem for me at the moment, but it certainly should at least be a known issue if it happens for anyone else.

bum51 started following the problem "Flash exception/errors with Flash 10 plugin and SM2 using either Flash 8 or Flash 9 .SWFs" in Schillmania!.

bum51 shared an idea in Schillmania! on October 06, 2008 18:58:

unload vs. destroySound

So for a while I've been having a small problem in my pre-loader script for mp3s. I've realised that I am just commiting the "unload" command and not actually destroying the sounds, and so if I go and create the sounds again later, with the same ID and such, with autoLoad set to true, then the file does not actually begin to load until it starts playing.

Now there IS an obvious fix for this; switch to destroySound (should be better on memory anyway, and is now my plan when I have access to it), or even just check to see if it exists already and start loading it. But it definitely took a while for me to find that tiny bug, so I was thinking, in case anyone else runs into the same error, could there be a way that SM2 checks when "re-creating" an existing one if any of the settings have changed? I've found that if it already exists then it just ignores any changes (such as volume as well, which could be very important), which has been a bit of a pain. ♦

Anyway, that's my 2 cents.

Also, great job on the MovieStar stuff, that's bloody brilliant! You are my hero!

bum51 replied on October 02, 2008 14:43 to the problem "Waveform and Frequency data issues" in Schillmania!:

I honestly didn't think there would be any sort of a workaround. I just figured it would be a good thing for you and other users to know about if they didn't already. And maybe there'll be a fix for it in an upcoming version of Flash anyway.

And thanks! I was actually going to try keeping an XML record of the FFT of the sounds and get those when I got the sounds and then you (and Flash I suppose, but it's your library that I'm using so Kudos to you) gave me the beautiful spectrumData and I rejoiced immensely as I was able to get a working visualization for every song by the end of day 1 rather than having to run full FFTW on every song and keep a full record of it, which would have probably taken months, or I would have given up ;-) This way is WAAAYYY more efficient in like every way possible.

The only (err... other) problem with it now is Internet Explorer, as it uses the canvas tags, but hopefully we'll get some HTML 5 love in IE8.
Still waiting for canvas 3d contexts though >_>

bum51 reported a problem in Schillmania! on September 30, 2008 01:57:

Waveform and Frequency data issues

I don't think that it's a problem with SM2 itself, but rather something to do with flash. The issue happens when the user opens another browser tab/window (same browser) that uses flash (with or without sounds I believe)

The issue I've been having is that flash seems to stop giving me any new "data". It seems to keep passing the same data that I already have. It still updates the times and such, but it's like it just forgets about the datas. If I close the other tab then once a new song starts playing, it'll start the data again, but not until after the first one is actually done playing and the next one starts.

As I said, I don't think it's an SM2 issue, but it certainly is quite weird.

http://www.bixtools.com:8088/

If you want to try it out. (Also, thanks so much to Scott for SM2 because otherwise I would never have been able to do any of that ;-)

bum51 replied on September 25, 2008 14:47 to the problem "sound.onload is not called for missing mp3 files using Flash 9" in Schillmania!:

Haha, I guess I should pay more attention to the Docs ;-)
I use an alternate method for checking for whether files exist before trying to load the songs in the first place. I never actually tried it that way.

bum51 replied on September 25, 2008 14:37 to the problem "sound.onload is not called for missing mp3 files using Flash 9" in Schillmania!:

Well techincally the onload shouldn't be called because it never loads. However I think there's a way to check the status of it, try using the readyState property.
Also, instead of ".play()" I would suggest setting "autoPlay:true".

@Scott: could there possibly be an "onFail()" event or something similar for when the file does not exist or SM2 can't access it for whatever reason?

bum51 replied on September 23, 2008 14:25 to the question "Loading files from user hdd! is it possible" in Schillmania!:

I'm pretty sure that the browsers block that kind of thing because if you could do that then it would be a HUGE security hole that there's no way to really deal with it.
Flash itself however might have the capability, but not likely within a browser, which more or less defeats the purpose. Best-case scenario (I believe) is that you upload it and if they choose not to keep it then you just delete it after they cancel.

bum51 replied on September 21, 2008 05:45 to the discussion "Soundmanager not working in IE but all other browsers" in Schillmania!:

I've been using numbers as ids and it seems to work pretty well for me in all browsers..... The dash might be doing it though. One thing I've made sure to do though is turn my numbers into strings before I hand them over (eg. id=n+""), that might help.

bum51 replied on September 11, 2008 20:04 to the question "buffering timeout" in Schillmania!:

Cool, looks like a good method. One other suggestion I might have is instead of checking in relation to a dialup modem, try getting the ID3s for bit-rate (if that exists) and estimate the required speed to load each one without having to buffer. Not sure if/how that would be done though..... I haven't really played around with the ID3s much.

bum51 replied on September 10, 2008 03:32 to the question "buffering timeout" in Schillmania!:

I think you could hypothetically check the number of miliseconds between "whileplaying" calls to see if it's not working.

eg.
var tm=null;
var buffers=0;
var maxBuffers=3
isBuffering=function (){
buffers++;
if (buffers>=maxBuffers){
//Go to next song here
}else {
tm=setTimeout("isBuffering()",1000)
}
}

whilePlaying=function(){
try{
clearTimeout(tm)
}catch(err){}
tm=setTimeout("isBuffering()",1000)
}
onFinish=function(){
try{
clearTimeout(tm)
}catch(err){}
buffers=0;
}

And of course you can change the number of seconds fairly easily. I did it that way in case it happens more than once for a song, of course you may want to do "simultaneous buffers" instead.

bum51 replied on September 02, 2008 16:14 to the question "What about flashblock ?" in Schillmania!:

As far as opera goes you could easily get around that by simply giving the object a ".visible" class. Or even better, you could detect it by checking for the css content.

As far as the rest goes, you could try having the flash video show up first (a loading box centered on the page) that disappears when finished and triggers the SM2 loading.
That way the users would always have the chance to click on it before it outright fails.

Of course, that would create a bit of a delay to loading the page, but you could have that as a setting like just about everything else is.