[ronbaby]

OK, well, I see that there are, give or take, about 5 gazillion posts here with various/numerous people having trouble of one form or another with getting audio out via their HDMI-out ports, so obviously I'm not alone. And I do apologize if the specific issue/problem I'm having/reporting is already well known to most or all of the serious XBMC hacks around here, but I do feel that I should speak up and say something about this specific issue.

Based on the evidence and my limited understanding, my guess is that my hardware is just too brain dead to do what I am expecting of it, but hopefully somebody here will tell me for sure. Separate from that however, I've spent (wasted?) hours and hours of time trying to pin this problem down, and I will say that I do think that XBMC could have been, and probably should have been more helpful and informative during this process. At the very least, it really should *not* have been offering me options that were just never going to work, in practice, given my setup and hardware.

Briefly, I have an MSI X340-218US laptop. It has an HDMI-out port. I have the HDMI-out of this laptop wired into one of the HDMI input ports on my Sony STR-DH520 receiver. (The receiver in question knows how to decode _all_manner of both Dolby and DTS encodings.)

I've already spent quite some time getting the video working perfectly with this setup. Now however, audio output coming out via the laptop's HDMI port is driving me slightly nuts. As long as all I want is downmixed 2channel stereo, all is well. However when I try to get 5.1 Dolby Digital to be passed down through the HDMI port from the laptop to the receiver... well... it just plain doesn't work.

I suspect that the reason for this may perhaps be because the company that constructed my laptop (MSI) was really being cheap about it, and they just didn't build the thing with the kind of hardware that would be capable of passing multichannel digital audio out the HDMI port. (But I sure would like it if someone could verify that for me.) What I do know is that under Windows, when I go and look at the audio devices, and then look specifically at the HDMI "device" (you know, as opposed to the built-in speakers device) the properties seems to say that "maximum output channels" is 2. I've tried to see if I could "fix" this by installing the latest and greatest Realtek audio drivers package (R2.70) and I've verified that the driver version did indeed change (to the latest) but "Properties" still says that the max channels for the HDMI device is 2. (I should perhaps mention that apparently, the thing that is actually responsible for implementing the "Intel[R] High Definition Audio" on this specific laptop is actually a Realtek ALC888S. CPU is an Intel Core2 solo.)

So anyway, can anyone here tell me if my hardware is just to stupid and/or brain dead to simply do pass-through on the Dolby or DTS encoded audio? (I don't see why this should be in the least bit hard for the hardware to implement, but what do I know? I don't build chips for a living.)

Anyway, for the moment, it is indeed my default working assumption that my hardware (and/or the most current available Windows driver for it) simply cannot do pass-through of encoded digital audio.

Moving right along, and with that assumption in mind...

At this point I would like to bitch and moan a bit about XBMC itself. Assuming, as I do for now, that my hardware simply can't do the pass through of digital audio via the HDMI-out port, then why, oh why, is XBMC offering me a veritable plethora of audio output options, the majority of which, apparently, do not have a prayer in hell of actually working on the hardware that I have?

To be more specific, under System->Settings->System->Audio Output there is an "Audio output" option that can have one of three values:

1) Analog
2) Optical/Coax
3) HDMI

Well, in the first place, this laptop has no optical or coax digital audio output ports. Why doesn't XBMC know that? Why is it trying to confuse me with options that are utterly irrelevant to the hardware that I actually have... options that, if selected, are absolutely guaranteed to insure that audio output from XBMC *will not work* ? This does not seem in the least bit "user friendly" to me. (I consider this to be a bug in the XBMC UI. The option shouldn't be there if it is only going to cause pain, suffering, gnashing of teeth and tearing of hair.)

Second, why am I even being offered an "HDMI" option here. I tried that option, repeatedly, with numerous other settings on numerous other audio-related options and it NEVER worked. Not even a little. I mean like not at all. Period. (I consider this to be yet another bug in the XBMC UI. The option shouldn't be there if it is only going to cause pain, suffering, gnashing of teeth and tearing of hair.)

Briefly, the *only* option from the above three that allowed me to get *any* audio out and through my receiver was the "Analog" option... a fact which is damn counter-intuitive, since the "analog" output in question most definitely *is* being passed down to my receiver via the (definitely digital) HDMI connection. Given that fact, one would naturally be inclined to wonder why "Analog" *must* be chosen here, and why the "HDMI" option, paradoxically, doesn't give you any audio at all.

Moving right along...

Separately, I have to also take issue with the options available under System->Settings->System->Audio Output->Audio Output Device. I was presented with three options here:

1) DirectSound: Primary Sound Driver
2) WASAPI: Speakers (Realtek High Definition Audio)
3) WASAPI: HDMI Device (Intel® High Definition Audio HDMI)

As I learned (via tedious trial and error), there was _no_ functional difference between #1 and #3 in my case, i.e. on my specific hardware. So why offer the user multiple options that in practice do exactly the same thing? (I consider this to be yet another bug in the XBMC UI. The option shouldn't be there if it is only going to cause pain, suffering, gnashing of teeth and tearing of hair.)

(Oh! And by the way... Do I have a "secondary" sound driver? A tertiary? A quaternary? If so, why wasn't I given options for those?)

More importantly however, I don't suppose that it ever occurred to any of the XBMC developers, but all three of the options in question are effectively greek (i.e. untranslatably meaningless) to the average Joe Schmo XBMC user... AND THIS IS NOT DUE TO ANY FAULT ON THE PART OF JOE SCHMO. (I know, because I am one.) Why should an end user have to take a Community College course all about the current industry nomenclature for various flavors of audio driver technology if he just wants to plop down of his couch and watch his bloody movie?

OK, fine. So because my audio wasn't working, a had to go and Google for "DirectSound" and "WASAPI" and then I had to go and read the respective Wikipedia pages for both of those, just to make sure that there wasn't something ``obvious'' that *I* had screwed up before I could come here and ask for some help. But the average Joe Schmo shouldn't have to be geeked-up on all this stuff just to get audio output working right with XBMC. The options presented to the user could be and, in my rarely humble opinion, should be so simple and obvious that even a cave man can do it... you know, like this:

1) Internal Speakers
2) HDMI Output Port

*That* would be user friendly. The current set of options for configuring audio output are anything but. (They were clearly written by geek for geeks. Mind you I have nothing against geeks. I are one. I just don't think that a UI which is designed for the masses should have e prerequisite of geek-level courses before an Average Joe can operate it or make it do something useful.)

OK, that's all my ranting for today.

Now, does anybody happen to know whether or not my specific hardware is even capable of digital audio pass-through? Or have I just been chasing my tail for nothing?

[helta]

you took all that time to post a ranting wall of text, and you didn't even post a debug log. http://wiki.xbmc.org/index.php?title=Log..._Debugging

[ronbaby]

Here you go...

ftp://ftp.tristatelogic.com/pub/xbmc.log-20121015034700

.. not that it is going to matter much to anything I have said.

The bottom line is that either (a) the native chipset (Intel 4 Series Express Chipset / GS45) plus the latest Windoze driver for that supports Dolby/DTS pass-through or else (b) it doesn't. (It should not be necessary to look at the XBMC log file to determine that, one way or the other. I feel sure that somebody somewhere on this planet probably does know what the hardware is and isn't capable of... totally independently of whatever XBMC may elect to scribble into its log file.)

If the correct answer is (a), pass-through _is_ supported, then XBMC is quite clearly blowing it by failing to make that (reasonably simple thing) work properly.

If on the other hand, the correct answer is (b), pass through of 5.7/7.1 Dolby or DTS is in fact _not_ supported by the chipset... and all indications I am looking at now suggests that it isn't... then why is XBMC offering me up all of these various advanced audio output options, when essentially none of them are going to produce anything other than agony, frustration, and lots and lots of lost time?

Either way, XBMC has what I would consider a "bug". Others may simply call it a non-feature, but after having wasted hours and hours investigating this, I can assure you most solemnly that from and end-luser perspective, it is a "bug", one way or the other.


Regards,
rfg


P.S. I talked to a support tech @ MSI late this afternoon. I made sure that he clearly understood my question, and I challenged him a bit, just to make sure that the answer he was giving me was not just the first one that came into his head. He confirmed for me repeatedly that the machine in question (MSI X340-218US with Intel GS45 chipset including Intel X4500MHD) simply will not do digital pass-through on any encoded audio to the HDMI port.

I'm still not 100% sure that that is really true, and I will continue to investigate, but I sure as hell have not found any clear indication anywhere that it is false.

P.P.S. So far, the one and only little tidbit of useful information I have found that relates the audio "pass through" feature/non-feature to _any_ Intel chipsets is this:

http://www.asrock.com/nettop/overview.us...T%20Series

wherein ASRock is selling a cute little box that contains (among other things) an Intel HM65/HM67 chipset which they, ASRock, are proud to point on _does_ in fact support pass-through.

It may perhaps be telling that, according to this:

http://en.wikipedia.org/wiki/List_of_Int...s_chipsets

the GS45 chipset was released in September of '08 whereas the HM65 didn't hit the streets until January 9, 2011.

From these facts it might be reasonable to infer that Intel didn't even produce _any_ chipsets that could do something as simple as passing 5.1 audio to/through HDMI until the year 2011.

Personally, I wouldn't put it past them, even though by the year 2011, neither HDMI nor Dolby Digital nor DTS was exactly "new" anymore.

(And here all this time I had thought that Intel was pushing high-end multi-media crap on everybody, you know, so that they could sell more silicon to us all. What was the point if it took them until 2011 just to make a chipset that could handle something as simple and as already well-established as 5.1 pass through? It's kinda nuts.)

P.P.P.S. If I had only been offered *just* these options for "Audio output":

1) Analog
2) Optical/Coax
3) HDMI

and then *just* these options for "Audio Output Device":

1) DirectSound: Primary Sound Driver
2) WASAPI: Speakers (Realtek High Definition Audio)
3) WASAPI: HDMI Device (Intel® High Definition Audio HDMI)

then right now I would only be grousing about that fact that (as my setup proves) "Analog" and "HDMI" are quite clearly *not* mutually exclusive, and that therefore, the options offered for "Audio output" quite clearly make no logical sense at all. But the thing that really got me chasing my own tail around in circles was the fact that XBMC _also_ offered me options for "Dolby Digital (AC3) Capable Receiver" and "DTS Capable Receiver". OK, my receiver has all of those capabilities. But it seems now that none of them can actually be *utilized* from XBMC on my specific hardware. So why was I even offered these options? I mean what was that... a tease? It sure made me believe that if my receiver could handle these things (which it can) that then XBMC would be both willing and able to cooperate. But apparently, not. (Not from my specific laptop anyway. Hopefully the new one I just bough off eBay specifically to run XBMC will not have this same drain bamage.)

I guess that I'm just disappointed, because I've only been trying to use XBMC for about a week and a half, and I read somewhere that "user friendliness" was an integral part of the XBMC manifesto. Well, I'm sure that XBMC most probably has lots of features and lots of possible uses, from being able to suck down RSS feeds to (maybe, for all I know) allowing me to watch re-runs of Happy Days on Nickelodeon. But frankly, I don't really care about any of that stuff. I only came to XBMC because I wanted something that would display *my* video and play *my* audio content. And so far, just getting it configured properly to do just those two simple things has proved to be anything but "user friendly". Just simply configuring the audio output properly has turned into a big can of worms. And more to the point, I don't think that it really *had* to be like this.

[jjd-uk]

You seem to be getting confused between the Realtek Audio controller which should only be for the speakers and analog outs and Intel High Definition Audio. Updating the Realtek drivers does nothing for the HDMI as the HDMI is controlled by the Intel GMA4500M graphics so you need the correct Intel High Definition Audio driver for the HDMI, this driver should be part of the Intel HD Graphics driver package.

Download the latest Intel drivers from Intel Download Center

Once installed ensure the HDMI device is set as default in Sound



You then need to check what windows reports as the capabilities of the device in Supported Formats



If you've got the correct drivers you should hopefully see at least 6 in Max Number of Supported Channels and Encoded Formats should report DTS Audio & Dolby Digital, if not then there is still something wrong the OS level.

Once Encoded Formats reports DTS Audio & Dolby Digital the final thing to check is that the two Exclusive Mode settings ticked



Once this is done you should be good to go in XBMC with WASAPI: HDMI Device (Intel® High Definition Audio HDMI) being the preferred device you should be using.

[ronbaby]

Thank you. I should be able to get some time to update the Intel drivers later today.

For whatever it's worth, I did actually visit (for the first time ever) the "Intel Download Center" just yesterday, not long before I posted, but something there convinced me that I either (a) already had the latest chipset drivers or else (b) the Intel drivers might not provide the same capabilities as what the system vendor supplied. (I forget which.) But I'll go back now and give that another whack.

One thing that I do know is that on the second of the three screens you posted, currently it most definitely says "Max Number of Channels: 2" and also, the list of Encoded Formats supported is empty.

But anyway, yes, this page I found:

http://www.mythtv.org/wiki/User_Manual:H...Passthroug

indicates that "Cantiga" chipsets (which is apparently what I have) does support pass-through. So we'll see.

[jjd-uk]

I did a Google earlier which suggested Intel GMAX4500 should support up to 8 channels of PCM, so should do DD & DTS 5.1 which takes the equivalent bandwidth of 2 channels of PCM.

See http://www.anandtech.com/show/2627/5

[ronbaby]

OK, so before I say anything else, I want to extend my sincere thanks to jjd-uk. It takes a kind and generous soul to look past the acrimony and to try to help someone who has bitched and moaned as much as I have about this HDMI audio output issue.

That having been said, I'm afraid that I must also report that it now appears that installing the latest and greatest chipset drivers available from the Intel Downloads Center (for my specific chipset) may have been (and most probably was) an extraordinarily Bad Idea.

After I did that (and then also performed the obligatory shutdown and reboot), I now find that whereas I was hoping that this would cause me to now see an HDMI audio output device for which "Max Number of Channels:" was something greater than two (and whereas I was also hoping to see "Encoded Formats:" chock full of multi-channel compressed audio codecs, as shown in jjd-uk's posting above), now, sadly, It appears that the primary effect of installing the official Intel chipset drivers was to cause my HDMI audio output device to entirely disappear from the conscious mind of Windows 7.

No, I'm not kidding.

Now, when I look at the first of the three screens that jjd-uk posted (above), although I formerly had two devices listed there... HDMI and speakers... now the only thing that's listed anymore as an audio output device is speakers (which I assume means just the tiny crappy ones in the laptop itself).

In short, installing the official Intel chipset driver seems to have actually taken me back a step. Now, it would seem, I am not going to be able to even get the 2 channel audio to flow (down the HDMI cable to my receiver) that I was formerly getting... i.e. before I ``upgraded'' the chipset driver.

(Perhaps this should go into an FAQ somewhere: If you have a Cantiga chipset on a something that has an HDMI port, then DO NOT install the latest Intel chipset drivers. You'll regret it.)

Sigh. I don't like Windows, and I don't normally even use Windows. But it does seem like I am going to have to learn now, at long last, how to do a Windows system roll-back, you know, to get back to how the system was _before_ I ``upgraded'' the chipset driver(s).

I will do that and then keep on looking for solutions.

One test I am already preparing to do is the one suggested on this page:

http://www.mythtv.org/wiki/User_Manual:H...assthrough

for the Cantiga chipset, which involves running mythfrontend and then checking some things under that. (This should at least give me one more opinion about whether or not my specific hardware can or can't pass-through Dolby Digital and/or DTS.)

I am in-progress right now building the X.org stuff and the mythtv stuff under FreeBSD on the MSI laptop (in preparation for performing this test) and I will report here on whatever I find, once I run the test.

Certainly however, the anandtech.com page that jjd-uk cited, and also the material that I cited on the mythtv.org web site do seem to support the view that Cantiga chipsets ought to be able to pass-though either Dolby or DTS 5.1 or even 7.1 audio to the HDMI port. I should note also however that Post #125 in this thread:

http://forum.xbmc.org/showthread.php?tid...pid=665017

says quite plainly that the Cantiga chipset cannot handle High Bit Rate (HBR) audio. Assuming that's true, then I'll never get either Dolby TrueHD nor DTS-HD flowing out of the HDMI port on my specific MSI laptop. But that's OK. If all I can get is lossy 5.1, then I will still be a pretty happy camper.


Regards,
rfg


P.S. It probably goes without saying, but I'll say it anyway... Nothing I've said in this message in any way negates the points I've made previously about the rather entirely sub-optimal aspects of the stock Confluence skin when it comes to configuring audio output. In my opinion, this part of the UI is in need of a serious re-think.

[jjd-uk]

Damn, maybe I've mistakenly pointed you at the wrong drivers, if you haven't already rolled back try the Intel Driver Update Utility at http://www.intel.com/p/en_US/support/det...ht_dctop10

---

Also just in case you don't have them to hand the MSI provided drivers can be found at http://www.msi.com/product/nb/X340.html#...=Win7%2032

[jjd-uk]

says quite plainly that the Cantiga chipset cannot handle High Bit Rate (HBR) audio. Assuming that's true, then I'll never get either Dolby TrueHD nor DTS-HD flowing out of the HDMI port on my specific MSI laptop. But that's OK. If all I can get is lossy 5.1, then I will still be a pretty happy camper.

Oh, and what this means is that Dolby TrueHD & DTS-HD can't be bitstreamed for decoding on the AVR, however the 8 channel PCM support means HD audio formats can be decoded on the laptop and passed over the HDMI as 7.1 PCM to the AVR..

[ronbaby]

My sincere thanks again to jjd-uk for being so willing to share knowledge.

As regards to the Intel drivers... yes, the ``new'' ones that I just installed (that made my HDMI audio disappear entirely) came to me via the Intel Driver Update Utility. (I know so little about Windows drivers generally that I decided to let this auto-magical program decide for me what new driver(s) I should have. In retrospect, and with 20/20 hindsight, it now appears to have been a mistake to have placed such trust in this specific bit of Intel wizardry.)

As regards to the official drivers from MSI.... THANKS! I quite certainly had already looked myself for a set of official MSI drivers on MSI's web site, but for some reason, the page that *I* ultimately found was most certainly *not* the one that you (jjd-uk) just pointed me to.

OHHHHHHH! I just now went back and looked again, and *now* I know where I took a seriously Wrong Turn. Notice that if one simply goes to this URL:

http://www.msi.com/service/download/

and then chooses option #2, and then selects "X-Slim" series (which is what my laptop is part of), the x340 model isn't even listed there anymore! (Geeezzz! I only bought the thing about 2 and 1/2 years ago... April 2010! But I guess that MSI wants to treat as as antiquated and obsolete already. Greedy bastards probably just want me to buy a new one.) So anyway, x340, not being on the list, I choose x370, you know, as the next closest thing that *is* on the list, and then "Win 7" and that gets me here:

http://www.msi.com/product/nb/X370.html#...=Win7%2064

and on *that* page, as you'll notice, the only thing that looks like it might even be vaguely related to this whole HDMI audio output issue is the latest & greatest RealTek audio driver (6.0.1.6267) and as it happens, I ended up downloading & installing that exact same version from the Realtek site a bit later on anyway, so I had even that part covered (although as I have now learned, the Realtek stuff only drives the built-in speakers, so even that driver was and is irrelevant to what I'm trying to do).

Question: Do I have to go through the whole rigamarole of backing up and restoring to the last Windows checkpoint (or whatever they call it)? Or can I just effectively override the (apparently broken) Intel chipset driver by just installing the MSI-suppled version of same from the MSI web site over the top of the broken Intel one? (As I've said, I don't normally work on Windows, so I have no idea how this stuff works.)

Anyway, regardless of what I have to do to get there, I most certainly _do_ want to (and intend to) give a try to the latest MSI-supplied GS45/4500MHD chipset driver(s). I don't know that they will help any, but they probably won't hurt. (And at the very least, I should at least be able to get back stereo over HDMI, which is at least something.)

As regards to having the laptop decode TrueHD & DTS-HD and then just shipping it down the HDMI pipe as 7.1 PCM to the AVR... yea! That sounds great! (No pun intended.) If I can do that, then that will be very nice indeed. Point of curiosity: If I do that, then will the audio streams seen by the AVR be lossy, or not? (Not that I really care. I'll be happy either way... assuming of course thatAVR can grok 7.1 or even 5.1 channels of PCM coming down to it from the other end of an HDMI connection. I don't know at the moment whether it can or not, and I'm not even sure how to find out, other than to just try it and see. These are the kinds of in-depth details that seem to routinely get left out of ``consumer'' oriented manuals.)

Anyway, I'm still working on testing some things under FreeBSD, and for whatever it's worth the signs there are encouraging... so far. The dmesg output (boot-time kernel message) have this to say about the audio hardware:

hdacc0: <Realtek ALC888 HDA CODEC> at cad 0 on hdac0
hdaa0: <Realtek ALC888 Audio Function Group> at nid 1 on hdacc0
pcm0: <Realtek ALC888 (Analog 2.0+HP/2.0)> at nid 27,20 and 24 on hdaa0
pcm1: <Realtek ALC888 (Onboard Analog Mic)> at nid 18 on hdaa0
hdacc1: <Intel Cantiga HDA CODEC> at cad 2 on hdac0
hdaa1: <Intel Cantiga Audio Function Group> at nid 1 on hdacc1
pcm2: <Intel Cantiga (HDMI 8ch)> at nid 3 on hdaa1

None of that is likely to mean much to Windows folks, but I though that I would share it anyway.

The part that says "(HDMI 8ch)" is certainly encouraging.

[ronbaby]

OK, some test results...

I was not able to perform the suggested MythTV test under FreeBSD, because apparently, one is not able to get past first base with MythTV unless and until one se's up a MthyTV database _server_. (You can't even get past the config screens without that, apparently.) I didn't feel like going through all of the rigamarole necessary to setup a MythTV server just for a sound test, but I found this other suggested way of testing 5.1 under FreeBSD:

http://aksyom.blogspot.com/2010/08/surro...x-and.html

(I should probably mention here that fortunately, ICH9, aka ICH9M-SFF, which is apparently part of the whole chipset I have, _is_ supported under FreeBSD.)

Anyway, thank God that this guy put this page up! Otherwise I would still be fumbling around trying to merely test 5.1 output over HDMI from FreeBSD.

The short version is that you download a couple of test files (which he gives URLs for) and then configure mplayer so that it knows to talk to the OSS drivers, _and_ so that it knows to output 6 channels. Thank god also for this page, which describes exactly how to do those things:

http://www.mplayerhq.hu/DOCS/HTML/en/adv...round.html

So I downloaded the test files, built & installed mplayer, configured it for OSS drivers and 6 channel output, and then just told mplayer to play each of the two files (.ac3 and .wma) in turn.

The results were at once both encouraging and also slightly disappointing.

When playing the .ac3 (Dolby Digital, 5.1) test file, the front LED display on my STR-DH520 lit up like a Christmas tree. The big-letters part of the LED display started a crawl which announced very unambiguously "Dolby Digital 5,.1". Meanwhile, all five of the speaker indicators (Front-left, Center, Front-Right, Rear-left, Rear-right) all lit up, along with the LFE (Low Frequency Effects, .1) indicator. Also, the tiny little Dolby logo LED lit up, followed by the associated (small) following capital `D' (for "digital"). In short, I made it!

When playing the .wma file the big-letters part of the LED display on the receiver said only "LPCM 48kHz" but again, all five of the little speaker indicator LEDs lit up, along with the LFE indicator. So again, success.

That's it for the good news. That bad news is that despite what all the wiz-bang LED indicators on the front of the receiver were saying, what I actually *heard* was only "Center... Front Right... Rear Left... Rear Right" (all coming out of the proper speakers, except for center, because I actually don't have a front-center speaker, and my receiver knows this, so it helpfully played the word "center" equally out of both the left-front and right-front speakers).

Problem is: What the bleep happened to Front-Left!?!?! In neither of these playbacks (.ac3 & .wma) did I ever hear the words "front left". (I also never heard any of the alleged grumbling at the end, because in addition to only having 4 normal speakers, I ain't got no LFE woffer... yet.)

Hummm... well, because everything else seems to be working so splendidly, I'm going to be uncharacteristically generous for the moment and just assume that every time a new kind of encoded audio bitstream begins to be sent down to the receiver, the receiver has to think for a second or two before its little brain fully kicks in and it starts applying the proper codec, and that thus, I never head the initial words "left front".

Well, it's a theory anyway. :-)

Meanwhile, later that same lifetime...

I couldn't stand not knowing, so I went and found another surround sound test file. The one here is good:

http://www.tfm.ro/win32-projects/test-avi-ac3/

and playing that through mplayer now I do indeed get all of the speakers enunciated, *and* as an added bonus, this test/demo file has a nice test of LFE, which sounds to me like it is perhaps just a sine wave tone. It starts out at 120Hz and slowly slides down to 20Hz. And even though I have no big fat woofer, per se, I could hear it all the way down from 120 to 20. (So obviously, my receiver is doing the Intelligent and Correct Thing.)

The bottom line is that yes, 5.1 out via HDMI (both DD and LPCM at least) from the GS45 Cantiga chipset _does_ work, and either the douchbag first-level tech support drone who I talked to @ MSI doesn't know his ass from a hole in the ground or else he _may_ perhaps have just been trying to tell me that 5.1 out of this thing simply will not work OUT OF WINDOWS (perhaps because nobody ever got 'round to implementing it in the Windoze driver(s), or perhaps because of some obscure licensing hassles). I am still not far enough along to definitively either prove or disprove that 5.1 out from this chipset under Windows will work or not work, but I intend to keep working to find out, finally and definitively, what the story is on 5.1 out of this chipset under Windows, and then to report whatever I find here.

That's the least I can do... record my results here so that others who fall into these same holes in the future can perhaps at least google and find this thread and then become enlightened, as I am slowly becoming, thanks to jjd-uk and a whole lot of diligent messing around.

[jjd-uk]

Can you use this guide to check what the Intel Graphics Controller is reported as http://www.intel.com/support/graphics/sb/cs-009482.htm as I'd suggest still trying to get it working with Intel supplied drivers is best. I believe this is the graphics driver package you require.

Driver download http://downloadcenter.intel.com/Detail_D...9&lang=eng
Release Note http://downloadmirror.intel.com/20189/en...gfx_32.htm

Look at Readme.txt on that Intel download page for details on installing the Audio driver.

If however you want to cut your losses and don't want to persist with the Intel supplied drivers, then I'd just go ahead and install the Mobile Intel® GMA 4500MHD package from the MSI page, as this should include the Intel HD audio drivers for HDMI, and see what if any difference this makes. If no different then try the Intel® 5 Series Chipset driver package on the same page.

If still no different then you may want to consider getting advice over on AVSForum to see if anyone has direct experience of Intel GMA 4500MHD graphics, my only experience is with the later Core i3 graphics.

Rollback on Windows is done via System Restore, found under Accessories -> System Tools from All Programs on Start menu, that is if it was enabled on install.

Going back to HD Audio, there should be no perceptible difference between decoding the HD Audio format on the laptop for send as LPCM 7.1 and bitstreaming as TrueHD or DTS-MA, most want bitstreaming as it guarantees the audio hasn't been interfered with by the OS and they like to see the decoding led's lit on their AVR as proof their getting HD Audio.

[ronbaby]

Following the steps given in:

http://www.intel.com/support/graphics/sb/cs-009482.htm

my adapter is exactly as shown there, i.e. "Mobile Intel® 4 Series Express Chipset Family".

Regarding (broken) driver replacement, must I really do a System Restore? Will things become horribly hosed if I just simply install new/different drivers for the devices of interest over and on top of the old broken ones?

Regarding HD audio, yes, I myself am probably one of the people who would prefer to see all of the decoding indicators lit up on my receiver. It gives me a nice warm fuzzy feeling, and makes me feel that I didn't totally waste my money on this fancy-schmancy decodes-everything receiver.

Regards,
rfg


P.S. Not that anybody _here_ (in the Windows support forum) is likely to care, but I just finished building & installing xbmc on FreeBSD Version 9.1-RC2. Testing it with the test file available here:

http://www.tfm.ro/win32-projects/test-avi-ac3/

Video display is correct/acceptable, and I am able (with default settings) to get 2 channel audio down through the HDMI cable to my receiver, but so far, that's all. I will be asking FreeBSD multimedia gurus for guidance in setting up xbmc on FreeBSD for HDMI multi-channel encoded and LPCM pass-through. (It seems that xbmc for *NIX platforms is heavily ALAS-oriented. From what I understand however, FreeBSD isn't. Like not at all. So this may perhaps not go well.)

[jjd-uk]

Regarding (broken) driver replacement, must I really do a System Restore? Will things become horribly hosed if I just simply install new/different drivers for the devices of interest over and on top of the old broken ones?

You should be able to just go ahead and try the driver install, if any conflict was detected I would normally expect the new driver to uninstall the old driver, I just gave the System Restore info in case you wanted wanted to rollback to how it was.

[ronbaby]

Success! But I have lots more info to share...

I've now (re-)loaded up the latest & greatest chipset and graphics (4500MHD) drivers available from MSI onto the x340. They were not quite as easy to get hold of as I had hoped. Apparently, if you search around for them and then end of up at this page:

http://www.msi.com/product/desktop/x340.html

then you will come to nothing but grief, because from there, *all* of the download links are BROKEN. Instead you *must* go here:

http://www.msi.com/product/nb/X340.html

and although this page looks identical to the other one, from here the links to all of the downloads are *not* broken.

So anyway, as I was saying, I re-installed chipset and graphics drivers... the latest & greatest from MSI themselves. (The installer did tell me that I was replacing newer with older in the case of the graphic stuff, but I proceeded anyway.) At first, those didn't seem to improve the situation, HOWEVER, I then later re-installed (again) the latest & greatest 4500MDH graphics (and chipset?) driver available from Intel's automagical update service, and then, fiddling around with the stuff some more, I managed to make it all work, at last. Yep, 5.1 Dolby Digital, out of the laptop and across HDMI to the receiver. (I verified conclusively that I was in fact getting all 5.1 channels by running the nice .avi test file that I mentioned earlier, both under Windoze Media Player and also under XBMC.)

I learned a couple of two or three critical lessons during this final phase of my fiddling. I share these here in the hopes that other people struggling with this same crap in the future can maybe find this thread and save themselves a lot of time and bother.

1) In the end, *both* the latest graphics drivers for the 4500MHD available from the system manufacturer (MSI) *and* also the slightly newer ones available from Intel do work, and both allow up to 8 channels of Dolby Digital or DTS to flow out of the HDMI port on the laptop... UNDER THE RIGHT CIRCUMSTANCES. (See below.)

2) Apparently, for laptops like mine (which have both an internal display and also an HDMI output port), Microsoft, in its infinite wisdom, decided that if you don't have anything live plugged into the HDMI port AT THE MOMENT, then you don't really need to see anything HDMI-related displayed in your list of available audio devices. (This, it turns out, explains the vanishing HDMI audio output device that I mentioned earlier. While experimenting, I was moving my laptop back and forth between the living room and my office. When the laptop was in the living room and wired to my AV receiver... or at any rate to *something*... via HDMI, then an HDMI thingy of some sort was displayed in my list of audio devices. When the laptop was in the office however, and had nothing plugged into the HDMI port, the HDMI audio output port would mysteriously disappear entirely from the list of available audio output devices. I'm sure that Microsoft thought that they were doing end-lusers a favor but not showing them anything about any hardware which they could not make use of right at that moment in time, but personally, I found that this does far more harm than good.)

3) If you have a laptop which has an HDMI output port _and_ which uses an Intel graphics chipset, then even if you have your HDMI port wired into a Dolbly Digital / DTS capable AV receiver and even if that receiver is properly connected via HDMI to a flatscreen TV, and even if the receiver and the TV are power powered up, ready and waiting, before you even boot your laptop, Windows (7 and possibly others) will show you that you do indeed have an HDMI audio output device on your system, but it will flatly refuse to tell you almost anything else about it (properties) unless and until you actually use either Windows Control Panel or else the spiffier Intel-supplied display control program to switch from displaying on the internal display to the external (TV) display. Only after you have switched to the external display will Windows show you all of the actually available properties for the "Intel® High Definition Audio HDMI" thingy... you know... stuff like max # of channels, available bitrates, codecs, and so forth.


++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

OK, so now that I'm past all of that, I just feel the need to say again that the way audio setup is structured under Confluence is rather entirely sub-optimal. (It _still_ looks entirely sub-optimal to me, even now that I have my multi-channel encoded audio output working correctly over HDMI.)

There are clearly at least one or two audio output settings in the Confluence UI whose settings, at least on startup of XBMC, could be drawn, and in my opinion should be drawn from available information about current _Windows_ settings. And beyond that, the available audio setup options are just confusing, contradictory, and nonsensical.

But I won't bore people here with this rant anymore, since it is not relevant here. I need to take this onto the developers forum, which I certainly shall, in the fullness of time.

My thanks again to jjd-uk for all the patient helpfulness.