Win 7 XBMC Dharma RC1 No AC3 when using HDMI Audio [debug log posted at #9] - Printable Version
+- XBMC Community Forum (http://forum.xbmc.org)
+-- Forum: Help and Support (/forumdisplay.php?fid=33)
+--- Forum: XBMC General Help and Support (/forumdisplay.php?fid=111)
+---- Forum: Windows support (/forumdisplay.php?fid=59)
+---- Thread: Win 7 XBMC Dharma RC1 No AC3 when using HDMI Audio [debug log posted at #9] (/showthread.php?tid=86121)
- G|zO - 2010-12-11 12:25
I also tryed Boxee, and the sound worked perfect!
me too. - willrom - 2010-12-17 06:02
Everything worked great in Camelot. Broke in Dharma betas, still broken.
ATI 5000 series card. Win 7 64bit ent.
Everything still works in VLC.
DTS works in Dharma. Have to change audio output to analog for AC3 audio to transmit.
- cartman1337 - 2010-12-19 23:21
Wow! I was amazed when I downloaded something called "Release Candidate 1" of Dharma and found it was completely ruined as far as audio output goes (no Dolby Digital bitstreaming working is equivalent to ruined in my book, as far as media players go). But now you've actually gone and RELEASED this, without having fixed the issue in question, despite several bug reports included in this thread, obviously seen by developers who've commented in this thread, and even dare to call it stable? You've just lost a LOT of cred in my book!
I just downloaded the release version, tried to play a Dolby Digital encoded movie, discovered it still doesn't work, tried all available audio settings without getting it to work. So I continued to uninstall it and reinstall the previous version, which as far as I'm concerned is still the latest STABLE version of XBMC...
So, what I'd like to know is, are anyone working on this rather serious bug? Enough users have joined this thread to suggest the problem is quite real, and now that you've just released it as "stable" I'm pretty sure there will be quite a lot more users with AMD 5xxx cards who'll also run into it. Further more, if you need me to provide any sort of usage dump or anything along those lines to get more to work on, let me know...
- CrystalP - 2010-12-20 04:02
We just had to release Dharma and move on. It had been way too long already. It means that not every bug could be fixed.
I compared boxee's code with ours and the way the sound device is opened is identical, so a fix will not be trivial, especially since no dev can repro. So if it's important to you, help find which the revision or a date range when the breakage happened. To do that, download the code and build older versions of xbmc unless you can find pre-built versions that are old enough.
Other than that, I hope the issue will be fixed with the AE branch merge, which will replace all audio code and do things more correctly.
- cartman1337 - 2010-12-20 18:46
CrystalP Wrote:We just had to release Dharma and move on. It had been way too long already. It means that not every bug could be fixed.
Why can't you just use the sound device part from version 9, which worked perfectly?
- cartman1337 - 2010-12-20 22:19
By the way, I've downloaded all the Dharma betas and nightlies I found still posted on the site, so I'll check them out in order, and see if this sound issue is OK in any of them, when I get around to it.
As for source code, I find the current release source, but where do I find the sources for any earlier releases? And do you have any instructions as to how to go ahead and compile in Windows, what I need etc.? I have programmed earlier in life, but it's been so long I don't have any compilers lying around any longer. As such, I do hope this issue can be figured out without it coming to me having to learn to program a new language, buy new compilers and stuff. Because it's not worth THAT much to me.
Still no AC3 - john.eyl - 2010-12-21 01:51
Cartman, I hear your frustration. Even though XBMC is free and pretty much a labour of love for the devs, it's still very frustrating to have something like this happen.
The only other example I have seen of things like this occurs when there is an EDID issue. It's supposedly fixed in later drivers for the 5xxx cards but I think I may look into it over Christmas. Otherwise I am stumped at the moment. I've tried Beta 2, 3, RC1 and RC2 with the same problem. I'll try Bet 1 tonight and go from there. Fortunately, there is a guide here http://wiki.xbmc.org/index.php?title=HOW-TO_compile_XBMC_for_Windows_from_source_code to compile XBMC. I'm a software engineer and work with all of the prerequisites on a daily basis - so I should be able to get it compiled.
CrystalP - out of curiosity, you say you can't get it repro'd. What are you testing it on? Are you using an ATI 5xxx card? What receiver do you have?
- CrystalP - 2010-12-21 11:04
The HOW-TO-BUILD guide is up to date and if you run into problems ask for help in the development section of the forum.
To get older versions of the source code, check out the code with TortoiseSVN, and then use the "Update to revision" function to go to a specific revision in the past.
There have been quite a few changes since 9.11 so dropping some 9.11 code in a recent version is not likely to work. Real bugs were also fixed, which would be reintroduced. I don't think the issue is in the code dealing directly with the device anyway. Passthrough of AC3 or DTS doesn't make a difference to that code, and it also happens with both DS and WASAPI.
As far as repro, I don't have an ATI 5xxx or an hdmi receiver, and no dev has that config either, as far as I know.
- steelman1991 - 2010-12-21 13:23
CrystalP Wrote:The HOW-TO-BUILD guide is up to date and if you run into problems ask for help in the development section of the forum.Gnif has both (donated by members as far as I'm aware) to develop the audio branch. Maybe he could have a look.
- CrystalP - 2010-12-21 19:29
hdmi receiver yes, ATI maybe, Windows definitely not...