Anthem D2/D2v/AVM50/AVM50v/ARC1 tweaking guide - Page 1141

MACCA350,

I can partially confirm your recent results.

I can confirm that for 5.1 (but *NOT* for 7.1) DTS-HD MA Bitstream input, the D2v decodes to a +3dB hotter result than with LPCM input (with the exception of the LF/RF channels, due to their previously reported decoding problem). However my tests show no such discrepancy between LPCM and Bitstream input for traditional, 5.1 DTS tracks, nor for 7.1 DTS-HD MA tracks (again, with the exception there of the previously reported LF/RF problem).

In addition, I discovered what appear to be a couple decoding level bugs in the PS3.

I also re-confirmed my suspicion that using DTS and DTS-HD MA tracks to set speaker levels, even DTS tracks from calibration discs, is likely to produce problems. These tracks are fraught with inherent level problems IN THE CONTENT. As it turns out, the 7.1 DTS-HD MA track on AIX is correctly matched for channel levels, but as you showed, the 5.1 DTS track on that disc is not.

As always, I recommend using LPCM tracks (as from the AIX disc) for actually setting speaker levels.

---------------------------------------------

Test tracks used:

For DTS, I used the 5.1 DTS Channel ID track from Avia Pro Audio Test Tones calibration SD-DVD, and also the 5.1 DTS track from Disney World of Wonder calibration SD-DVD.

I did not use the 5.1 DTS Channel ID track from AIX Audio Calibration Blu-ray since, as you've recently confirmed, that specific track has a level encoding error for the surrounds and subwoofer. It shouldn't matter since the error is in the content and thus the LPCM and Bitstream versions would both see that track the same way, but I decided not to use it anyway.

For DTS-HD MA, I used the 7.1 DTS-HD MA Channel ID track from AIX Audio Calibration Blu-ray, and also the 5.1 and the 7.1 DTS-HD MA tracks from Disney World of Wonder calibration Blu-ray.

---------------------------------------------

For Bitstream input, I played the tracks from the Oppo 93. For LPCM input I played the tracks from the Oppo 93 and also from the PS3. In the Oppo, Secondary Audio was OFF and the HDMI 1 output was used.

For the 5.1 tracks, I played using my existing 5.1 speaker setup in the D2v. Thus Audio Surround Mode was NONE. THX was also OFF. I have no "temporary" level adjustments set. My standard ARC configuration was used.

For the 7.1 tracks, I temporarily told the D2v that I had two Rear speakers which are not actually wired up. Thus, again, Audio Surround Mode was NONE, THX was OFF, and no "temporary" level adjustments were set. Again, my standard ARC configuration was used. The lack of ARC data for the Rears is not a problem since they are not wired (I can only hear the portion of Rear channel bass steered to the subwoofer by the default 80Hz crossover).

So no PLIIx or other forms of expansion from fewer channels to more speakers, no down-mixing, and no THX processing -- i.e., nothing to complicate the result. Also for the 7.1 DTS-HD MA track, the decoders in the PS3, the Oppo and the D2v are all seeing that a 7.1 decode is required, so the special decoding rules imposed by DTS-HD MA when a 7.1 track is decoded for 5.1 speakers are NOT being applied.

(Of course I can't actually check the Rear speaker levels for the 7.1 tracks.)

---------------------------------------

My Radio Shack digital SPL meter was used following my usual practice: That is, set Main Volume in the D2v so that the intended base test level yields 75dB SPL with the meter set to the 70dB range ("C" Weighting; Slow Response). I did that using the Center channel for each track. Read comparisons against that base up to +/- 3dB SPL directly from the SPL meter. For variations more than +/- 3dB, adjust Main Volume to yield 75dB SPL again and take that adjustment as the reading. I.e., if you have to raise Main Volume +8dB then the comparison tone is -8dB too soft. This procedure eliminates the error in the Radio Shack meter at the upper or lower end of each range, and also the variation between range settings on that meter. It assumes the Main Volume control on the D2v is accurate, which has been separately confirmed.

The absolute setting of Main Volume to get the base level correct for each track can be used to compare the output level between tracks. The comparison readings for the channels within a given track tell the uniformity of output for that track.

Readings made this way are probably useful to detect variations greater than +/- 1dB. They are CERTAINLY useful to detect variations of +/- 3dB or more.

NOTE: It is normal for calibration test tones like this to show my subwoofer as almost exactly +3dB hotter than the other speaker channels due to the effects of Room Gain in my ARC solution. I eliminate that from the comparisons, since it's only any sub variation above or below that expected, "normal" result that are of interest.

---------------------------------------

Here are the results:

(Again, the D2v Main Volume choices below are based on Center channel.)

Avia Pro SD-DVD DTS 5.1 Channel ID
Oppo 93 Bitstream: D2v Main Volume -16; LF/RF/LS/RS -3dB low, Sub -6dB low
Oppo 93 LPCM: D2v Main Volume -16; LF/RF/LS/RS -3dB low, Sub -6dB low
PS3 LPCM: D2v Main Volume -16; LF/RF/LS/RS/Sub -3dB low

Disney World of Wonder SD-DVD DTS 5.1
Oppo 93 Bitstream: D2v Main Volume -17; LF/RF -3dB low, LS/RS -4dB low, Sub -1dB low
Oppo 93 LPCM: D2v Main Volume -17; LF/RF -3dB low, LS/RS -4dB low, Sub -1dB low
PS3 LPCM: D2v Main Volume -17; LF/RF -3dB low, LS/RS -4dB low, Sub +2dB high

-----------------------------

AIX Audio Calibration Blu-ray DTS-HD MA 7.1 Channel ID
Oppo 93 Bitstream: D2v Main Volume -12.5; LF/RF -3dB low
Oppo 93 LPCM: D2v Main Volume -12.5; all channels matched
PS3 LPCM: D2v Main Volume -10; all channels match

Disney World of Wonder Blu-ray DTS-HD MA 7.1
Oppo 93 Bitstream: D2v Main Volume -17; LF/RF -6dB low, LS/RS -4dB low
Oppo 93 LPCM: D2v Main Volume -17; LF/RF -3dB low, LS/RS -4dB low
PS3 LPCM: D2v Main Volume -14.5; LF/RF/LS/RS -3.5dB low

Disney World of Wonder Blu-ray DTS-HD MA 5.1
Oppo 93 Bitstream: D2v Main Volume -20; LF/RF -6dB low, LS/RS -4dB low
Oppo 93 LPCM: D2v Main Volume -17; LF/RF -3dB low, LS/RS -4dB low
PS3 LPCM: D2v Main Volume -17; LF/RF -3dB low, LS/RS -4dB low

-----------------------------------------------------

Conclusions:

1) There are level issues in the content in *BOTH* DTS 5.1 tracks tried. Combining that with the level issues we already know about in the AIX Audio Calibration Blu-ray DTS 5.1 track, the only conclusion is that DTS tracks are really bad to use for setting speaker levels. We can only hope the folks actually authoring movie tracks do a better job.

2) For the DTS 5.1 tracks there is NO LEVEL DIFFERENCE between LPCM and Bitstream input into the D2v.

3) For the DTS 5.1 tracks, the PS3 decodes the Subwoofer level +3dB hotter than either the Oppo or the D2v. Majority rules: The PS3 is wrong.

4) The DTS-HD MA 5.1 track also has content level issues.

5) There *IS* a basic level difference between LPCM and Bitstream input into the D2v for DTS-HD MA 5.1 tracks. The D2v renders the Bitstream +3dB hotter than the decoders in the PS3 or in the Oppo.

6) The previously known problem of the D2v decoding LF/RF -3dB too low for DTS-HD MA Bitstream input is present in this 5.1 track.

7) If you combine (5) and (6) it looks like what's really going on here is that when the D2v decodes this DTS-HD MA 5.1 Bitstream the LF/RF channels are CORRECT (compared to LPCM input). It's all the OTHER channels that are +3dB hotter than they should be! That would suggest the proper correction in the D2v is to use temporary trims to lower "Center" -3dB, "Surrounds-Rears" -3dB, and also "Subwoofer" -3dB when playing DTS-HD MA 5.1 Bitstreams. That has the net effect of raising LF/RF +3dB compared to the other channels, but also keeps the absolute levels matched against LPCM input of this same track.

8) The Disney World of Wonder DTS-HD MA 7.1 track has THE SAME content level issues as its 5.1 version. Of course I don't know what might be going on in the Rears.

9) NOTE: There is *NOT* a basic level difference between LPCM and Bitstream input into the D2v for DTS-HD MA 7.1 from the Oppo! However, the PS3 appears to decode this track -2.5dB too low. (The remaining +/- 0.5dB difference in the other speakers in the the PS3 LPCM output is in the noise of what I can measure with the SPL meter.) You could argue that the D2v is decoding too high compared to the PS3, but there is no such difference when compared to the Oppo. Again, majority rules. The PS3 is wrong.

10) The DTS-HD MA 7.1 Bitstreams also exhibit the previously reported decoding problem in the D2v where LF/RF are decoded -3dB too low. Unlike the 5.1 case, however, there is no countering rise in basic level to compensate. So for the Bitstream 7.1 DTS-HD MA tracks the workaround really is to just raise LF and RF +3dB with the temporary level setting controls. NOTE: I'm not sure the temporary adjustment memory distinguishes between 5.1 and 7.1 track input. So it may be best to just raise LF/RF by +3dB as previously discussed in this thread, and ignore the +3dB over all rise in basic level when playing a 5.1 DTS-HD MA Bitstream.

11) The AIX DTS-HD MA 7.1 track has no inherent content level problems.

12) There is *NOT* a basic level difference between LPCM and Bitstream of this track from the Oppo. However, as with the other 7.1 DTS-HD MA track, the PS3 is decoding this track -2.5dB too low. Again, majority rules. The PS3 is wrong.

13) As with (10) above, when the D2v decodes this AIX DTS-HD MA 7.1 track, LF/RF come out -3dB too low. This is the previously reported problem.

14) Comparing these tracks against each other shows they are pretty well matched in level (although there's no good reason why that must be the case) EXCEPT that the AIX 7.1 DTS-HD MA track is roughly -5dB softer than the other test tracks. That's just an authoring choice of course. Again, there's no cause to presume that two different test tracks will match in level.
--Bob

Thanks for running these tests Bob.


I believe this is due to an option in the PS3 to enable button noises to be mixed into the output. From memory I had a similar difference but when I disabled this feature those differences disappeared. From memory it's under the audio options section.

Btw I'm using an old 40GB 'fat' PS3 and ran the system update prior to running the tests, not sure if that makes any difference though.

Cheers

I also have a fat PS3 (i.e., one that can not do Bitstream output of TrueHD or DTS-HD MA tracks). My PS3 is not the original hardware however, as it ALSO won't play SACD discs.

All of the hardware I use has been updated to the latest firmware.

The only PS3 setting I can find that's like what you describe is the one that enables the PS3's own "Key Tones" when navigating the XMB menu. That really shouldn't have any effect at all on the way it decodes Blu-ray audio to LPCM.

I passed on these results to Anthem.
--Bob

Yep, that's the setting.

The main concern I have with the level increases the D2v applies, +3dB on MA 5.1 and +3dB with PLIIx on MA 5.1 and LPCM 5.1 is that it may be doing it in the digital domain. If that's the case then it will be clipping peaks in the digital domain. I'm hoping the level increases are being performed in the analog volume control, in which case there's no issue.......well apart from the fact there are level issues but it shouldn't cause additional signal clipping.

Out of interest the way these movies are mastered is quite interesting. Some are encoded with regular clipping on the disc already, some never clip yet the clipped signal is still there, some never clip and signal shows no signs of prior clipping and some never get close to using the full dynamic range. Interesting but not really relevant

Cheer

Yeah, there's a guy who posts here under the handle FilmMixer who could probably tell you more than you'd ever care to know about how corners might get cut, or what else could go wrong, to produce these variations of mastering and transfer issues.

He does this stuff for a living (and has his very own Emmy showing he knows how to do it right!).
--Bob

So the level issues seem to be:

1) D2v PLIIx processing adds +3dB on Center, L/R, SL/R, SUB channels when the D2v is decoding DTS-HDMA 5.1 or LPCM 5.1
(It may may or may not be applying +3dB on the newly created Surround Back L/R channels as this has not been tested)
Reference, Reference

2) D2v (including OPPO and PS3) is applying DialNorm as per Dolby spec, ie DD and TrueHD bitstreams with -27dBFS are reduced by -4dB.
Reference, Reference

3) D2v (including OPPO and PS3) is applying applying DialNorm to Legacy DTS bitstreams with encoded DialNorm values.
(Note: WOTW DTS 5.1 track is encoded with a DialNorm value(-4dB). This seems to cause a -4dB DN Offset on DTSHD decoders. Need to test on old DTS decoder to confirm this was ignored by Legacy DTS decoders)
Reference, Reference

4) D2v is applying +3dB to all channels except L/R when decoding DTS-HDMA 5.1
Reference (7)

5) D2v is applying -3dB to L/R channels when decoding DTS-HDMA 7.1 (also some SL/SR content is being redirected to L/R speakers at low level, could also affect issue 4)
Reference, Reference (10), (13)

6) D2v is applying -6dB to Surround Back L/R channels when decoding TrueHD 7.1
(I have confirmed SBL/R are encoded at the same level as the Centre channel)
Reference

(I can add issues and references to specific posts later, so let me know if I've missed some or something needs correction/adjustment)

cheers

Looking at some of those older graphs I noted DialNorm +4 being displayed on the D2v. I just played those DD 5.1 and DD 2.0 tracks now and the display is indeed displaying "+4dB Offset".

As I noted earlier only the TrueHD track on the AIX disc has a DialNorm value(-27dBFS) encoded. Both the DD 5.1 and DD 2.0 tracks have no DialNorm Value........either that or it is a value of -31dBFS and my software is not telling me because it calls for no adjustment in the Dolby spec(I'll have to investigate that one further).

If we assume those AIX DD 5.1 and DD 2.0 tracks are encoded with -31dBFS DialNorm. My software ignores DialNorm and shows they are encoded at the same level as the TrueHD, DTS, DTS-HDMA and LPCM tracks. Then it seems the D2v is not making an adjustment on -31dBFS DD tracks BUT it's telling you to increase the volume by +4dB to bring that track up to the THX standard of -27dBFS DialNorm playback.

This means that the DialNorm value is wrong in either the TrueHD track or in the DD tracks. Based on all the information I'd say the DN value is correct in the TrueHD track and incorrect in the DD tracks.
Because
(1)the D2v playing back TrueHD 4dB lower than LPCM, DTS and DTS-HDMA tracks, and
(2)the D2v seems to be conforming to the Dolby spec for DialNorm output(ignoring what it displays)

The problem here seems to be that it lowers DD(and maybe TrueHD) tracks encoded with -27dBFS by -4dB to output at the Dolby spec of -31dBFS BUT it doesn't tell you that you need to increase volume by +4dB to output at the THX spec of -27dBFS DialNorm output.
The only time it seems to tell you to increase the volume +4dB to bring Dolby track levels up to THX spec is on DD(and maybe TrueHD) tracks ONLY IF they are encoded with -31dBFS DialNorm.

I'll need to run some further tests, I think nutting this one out will be my next project.

cheers

BTW those AIX test tones are mastered at -22.5dBFS RMS, so theoretically they should playback at 82.5dB(C) if you use them to set reference...........theoretically!


Just looking at Tron DTS-HDMA 7.1 and there is no 0dBFS or signal clipping in the LFE channel(max signal is -0.275dBFS) BUT the other channels have a LOT of signal clipping.
It looks like they mixed it and there was heaps of signal clipping so they just used that mix and reduced levels by exactly -2dB and encoded it instead of remixing. All the clipping is in the encode its just at -2.000dBFS not 0dBFS

Just for kicks here's some pics, the orange lines indicate a clipping like plateau detected.
TRON Centre channel:


TRON Centre channel close up(Timestamp 1:17:13.930 - 1:17:14.130)


Hmmm, nasty.


And here's one dealt with properly. No 0dBFS or signal clipping at all in any channel over the entire movie, max signal is -0.217dBFS.

Deep Impact Centre channel close up


Hmmm, nicely done You could say Deep Impact TrueHD has better dynamics than TRON DTS-HDMA looking at that centre channel......you'll just have to increase the volume +4dB when playing the TrueHD(-27dBFS DN) track to realise it

Last off topic pic, I promise
cheers

That doesent look good. But it just confirm my findings to. I Many film it seams to mé that the dialog is way to high vs example gunshot. Open range is a good example how well dynamic coud be.......low voices and high gunshots

Hi Harmony One Users,

I would be grateful is someone could advise me on how to get those commands requiring a "3 sec press and hold" to work on my D2v, eg. 7 (video proc), Mode (Scale Output) and Dynamics (Brightness, etc) to work.

Am sorry if this is the wrong thread to post, but I thought since many of you use this particular remote ..........

Thanks
Ben

Let me know when you find out.......I still have to keep the D2v remote hanging around just them.

Cheers

Have you tried to "teach" the Harmony remote those codes? I taught my harmony to do a single button press to turn on and off Zone 2... It's two commands, one to power on Zone 2 and another to power off Zone 2. It was the same to control Zone 2 volume. Which on the Anthem remote requires pressing the Zone 2 button followed by the power button.

Yeah, that's easy. But the 3-second hold down is a tough one I think. My Universal MX-950 does not currently have a command for it (to get into the video menu), but it would be nice to have. I have to reprogram the whole thing anyway soon for a bunch of new gear (Oppo, Elite PRO-70X5FD, new DirecTV box, etc.). Maybe I'll take a shot at it. But I certainly don't have a clue with a Harmony.

-Brian

In the case of the Harmony remotes, you need to ask Harmony support to change the coding of the specific buttons in your definition for the device.

They have the ability to switch any button between "send only once" (as for Power Toggle), "send repeated presses" (as for Volume Up/Down or the Arrow buttons) and "send a long, single press" which is what you want for these press-and-hold operations.

I don't believe there is a way to make this change yourself or to get the remote to "learn" the change.

They may switch you to an alternate definition for the device in their library which already has these adjustments.
--Bob

Thanks for your advice, Bob. I was hoping you guys have a way of accomplishing it without resorting to Logitech support.

Macca350, ManWithAPlan, I'll be back with the solution.

Ben

Logitec Support was great for me, you just give them your login information and they can do just about anything needed.

I can confirm that. It plays 4 dB lower than any other tracks on the disc.


All bitstreams carry a DN value. But yes, if it is -31, no gain change occurs.


Results (1) and (2) show the decoder is working correctly. The display of DN offset, perhaps not.


That would seem to be the nut of the issue.


It should not advise to change volume at all when DN = -31.

All the tracks on AIX are -31 except for TrueHD, which is -27.

Yep, what I meant was that whoever set the DialNorm value in those tracks made the error, the decoders operating correctly. ie either the TrueHD track should have been -31dbFS DN or the DD tracks should have been -27dBFS.


THX want DialNorm output at -27dBFS, why should it not advise to increase -31dBFS tracks +4dB to output at the THX spec?
Only reason I can think of is that will increase max output levels +4dB above the standard, but if it's not increased dialogue will be 4dB lower than expected.

Or is it that the DD tracks that are mastered at -31dFBS were done intentionally to circumvent the usual Dolby -4dB DN Offset? In which case increasing them +4dB will actually mean they are output at an equivalent -24dBFS.

Cheers

I have just seen Leon on bluray. Sometimes voices are louder than gunskot
why does there have to bee so much comprecion in the mix. films like that Will sound avfull played at reference level......screaming voices. I Cant remember Any dvd movies Sound that compreced

Does your bluray player have a setting for 'DRC' or 'Dynamic Range Compression' in its setup menu.
By turning it to 'on', and maybe even to 'auto', it will turn up the dialog and tone down the explosions so that you can watch movies without disturbing other people in your house.
It almost sounds like you might have 'DRC' turned on. Maybe its just the mix on LEON bluray, but might be worth checking your bluray player's settings.

Tom

I am planning to add a TV to a workout area in my house. I would like to use the HDMI2 output on my D2v. I currently have HDMI1 set to 1080p60 or 1080p24 to my projector depening upon the input. I do understand that the HDMI outputs are parallel and that the same output is shown on both. Here are my questions:

1. Does the HDMI2 output send out an audio as well as a video signal? If the answer is "yes", does it send out the audio signal regardless of the type of audio input (digital via HDMI, digital via coaxial, or analog)?

2. I assume I need to find a TV that can accept either 1080p60 or 1080p24 input in order to make use of the HDMI2 without having to make changes to the D2v setting when watching the TV?

3. With HDCP content do both the projector and TV have to be on to watch the source, or will the HDCP since if only one of the display devices are on?

Thank you,

Mike

The error is in the TrueHD track.


THX wants sounds recorded at -20 dBFS to give 85 dB SPL when the volume is set to Ref (0 dB). Tracks with -31 are the same loudness as most DTS tracks and all PCM tracks, and since none of them need gain correction, neither do -31 dB tracks. This assumes an AVR that does not apply the blanket +4 dB gain shift. In that case, the -31 tracks do need to be attenuated 4 dB to bring them into match. If you were discussing that case with the blanket offset in effect, you are correct, and I missed that essential factor in my reply.


I have heard that some studios have adopted that practice so as to eliminate any perceived level difference to other lossless tracks. But in at least one case, Air Force One, the overall level of th master was reduced 4 dB in order to allow for the downmix of SDDS 7.1 with less peak limiting. So that one is correctly set for -31.

Three questions that I'm sure have been covered long in the past:

1. Is 1.33 still the most current release for the D2? I know that there were betas floating around, but at this point I would only stick to final releases. The website doesn't seem to show downloads anymore for the D2.

2. Does the latest version of ARC (which I believe is 3.02) work with the D2?

3. Has there been any talk about a D3 (or equivalent) yet? As nice as the D2v is I couldn't imagine buying a pre/pro that doesn't have a network connection in 2011.

Thanks in advance,

David

1. Yes
2. Yes
3. Only rumors.

1) Yes, 1.33 is still the most current and the official release for the D2. v1.47f is the beta version that's also available for the D2. However, if you do decide to upgrade to v1.47f, I would suggest that you contact tech support first to confirm with them that you should do it because I don't believe it works for all D2s. So, it's better to be safe than sorry.

2) Yes, ARC v3.0.2 works with the D2.

3) I have not heard anything official about a D3.

1. I am not positive, but I believe so. i have changed HDMI 1 and 2 to my TV and still can get sound out of the TV as well as the AVM50v So it should not be an issue.

2. Yes. As long as you are not buying the absolutely cheapest TV ever they should accept 1080p60 and most do 108024p.. but yes. That said HDMI 2 is a pass through ie. the video processor does NOT process the signal in the AVM or D2 so it willbe whatever the source resolustion and refresh rate are. ie. DVD will pass at 480i or whatever the output of the player is. BD will be 1080/24p

3. No. I have had display 1 on then 1 and 2, just 2 then 2 and 1. You can mix and match with no issue. I do have audio and video drop out as the new IDED for the display is acquired , but then everything goes back to normal after about 2 seconds or so.

Hope that helps.

Has anyone heard anything official or unofficial about a D3 or AVM 60.

I agree Network and media contections are very important in todays world and I would love to have all that stuff. At least my Oppo does it.

Thank you for the replies.

I am confused about the above statement. I thought that the HDMI outputs were just parallel outputs and that both were procesed the same way and had the same output settings?

Thanks,

Mike

Mike,

I do not believe they are just parallel. As far as I know HDMI 1 is processed through the video chip on board the AVM/D2 and you can apply changes to the stream. i.e. video profile 1,2,3,4

HDMI 2 as far as I know is a pass through that feeds the same video/audio information as the input being used

i.e. a DVD being output by the source at 480i will be:
HDMI 1: upscaled to the video settings you have (video 1 at say 1080i/60) while
HDMI 2: will output the source video/audio at the 480i

If I am wrong someone please current me.

HDMI 2 also does not display the on screen information volume changes etc.

That's incorrect. HDMI 1 and 2 output the same, processed video output except that HDMI 2 doesn't include the on screen info displayed by the Anthem.
--Bob