GEOSATpro Can the microHD do 5.1 audio?

I have noticed that the mix and content of the audio channels on 103W sometimes changes from program to program (for example, the Today show has a different mix than the Olympic coverage).

That is my observation as well. Seems that the discrete audio mixes on the PIDs are not consistent as Pepper suggests.
Yeah they probably are not consistent. But all I am suggesting, and would be perfectly happy with, is something that doesn't require you to know or even care what they are: if there are multiple streams containing different channels of audio (in this example, probably four streams each with two channels), an option "somewhere" for a custom audio processing that encodes 5.1 from whatever pid:channel the user decides to assign to each of 6 positions. I hope that makes sense. If I was good at graphics I'd draw a picture of what the popup would look like.

Upon thought, the "somewhere" would probably be as an additional option in the green-button popup that lets you select the audio pids. If there's more than one, add a "Combine" option at the bottom which leads to a box containing six fields [LeftFront, Center, RightFront, LeftRear, RightRear, LowFreq] and each of these could contain the designation of one channel from one of the existing audio pids. With the default if you haven't changed it yet being LeftFront and RightFront as L and R from the first pid, and the others empty.

This has the advantage of tasking the user, instead of you, with figuring out which audio channel goes where. And it's obscure enough most casual users probably wouldn't find it. Just one more feature that would set this box apart as more advanced than any of the competition.
 
This has the advantage of tasking the user, instead of you, with figuring out which audio channel goes where. And it's obscure enough most casual users probably wouldn't find it. Just one more feature that would set this box apart as more advanced than any of the competition.

Pepper, I understand what you want, but it probably isn't going to happen unless someone takes an interest in researching and taking on this as a side project. 5.1 encoding is about balancing and timing and not just about selecting what audio plays on which speaker. It also seems that the applications that hobbyists are using on their PCs are 3-4x and not realtime.....

Please find out how affiliates process the information and what equipment they use. Once the basics of how NBC uses these audio feeds is known, then we can determine if it is feasible to devote development time and money.....

Is anyone using a PCI DVBS card to tune and a PC manage the NBC audio sources?
 
Just for giggles I took a look at one of the TS files from HD02 when they were playing "War: What Is It Good For", or something like that.

The gory details are at this link: http://www.certsoft.com/FTA/hd02.txt

Just the essentials:
 

Attachments

  • hd02.png
    hd02.png
    14.9 KB · Views: 158
they would probably scramble those signals once somebody figured out how to put all that audio together , lol. I'm surprised they haven't done it already.
 
I just happened to tune into NBC west on 103 this afternoon and they were doing an audio test where they were sending tones down and it was identifying which audio channel was what. They also had it in text on the video. I hit record but probably got the tail end of it. I'll check it later to see if I got anything worth using.
 
Did a bit of research on this and here's what I found. The DaySequerra UpMix renders any two channel audio source (stereo, matrix encoded stereo or LtRt) as surround sound. Because UpMix renders sound independently of the transport environment, it will compliment analog, digital, linear or compressed mediums. Using the DTS Neural Audio process, UpMix can simultaneously position individual elements within a rich surround field creating unparalleled image stability and realistic granularity. This approach avoids taking “artistic license” with content by placing audio exactly where it would be heard in a professional LEDE (Live End Dead End) listening environment. For example, mono or pan-pot stereo will image in front of the listener on the center channel, where stereo containing depth information will surround the listener using the full capabilities of their surround decoder.
UpMix can be used as a stand-alone unit to monitor stereo production or in tandem with the DaySequerra DownMix as a complete 5.1 transport solution. Surround content that has been encoded by UpMix will be reproduced just as it was heard in the original 5.1 mix. UpMix is DSP based and will never need calibration or alignment.



So basically, the audio is set up like this: L - Left (AES Input 1)
R - Right (AES Input 2)
C - Center (AES Input 3) Ls - Left Surround (AES Input 5)
LFE- Low Frequency (AES Input 4)
Ls - Left Surround (AES Input 5)
Rs – Right Surround (AES Input 6)

So, to answer the question as to why you can't hear certain stems of audio, it's because they were not produced in 5.1 but in traditional LR stereo.
UpMix-Front-1260px-030912.jpg
I'm waiting to hear back from an engineer friend at NBC who can give me a bit more specific information. More to follow.
 
Last edited:
Thanks for the research. Interesting information! Looking forward to the additional insight that the engineer might share.

A Google search yielded that the process is proprietary for DaySequerra.

I have seen their booth at NAB, but had never stopped. Will have to put it on my map next year.
 
Further to my post above, NBC transmits their audio sent as stems...a discreet left channel, right channel, center channel, etc, etc, etc. The actual encoding into 5.1 is then handled at the station.. This is supposed to help resolve some audio issues as local stations switch from network to local programming and back. This is why when listening to the raw feed, you can only hear certain portions of the programming.
 
Thanks for the research. Interesting information! Looking forward to the additional insight that the engineer might share.

A Google search yielded that the process is proprietary for DaySequerra.

I have seen their booth at NAB, but had never stopped. Will have to put it on my map next year.

They're also big into HD Radio as well.

It makes sense what NBC is doing. Instead of dictating what audio they need to play out, they give the option to the stations instead.
 
A bit more info on the NBC Audio.....

The PIDs depend on which service you are looking for:

CH 9 / NBC East:

  • Video: 0517
    Audio 1: 0751
    Audio 2: 0752
    Audio 3: 0753
    Audio 4: 0754
    VANC: 8144

CH 10 / NBC Central:

  • Video: 0518
    Audio 1: 0761
    Audio 2: 0762
    Audio 3: 0763
    Audio 4: 0764
    VANC: 8145

CH 11 / NBC Mountain:

  • Video: 0519
    Audio 1: 0771
    Audio 2: 0772
    Audio 3: 0773
    Audio 4: 0774
    VANC: 8146

CH 12 / NBC West:

  • Video: 0520
    Audio 1: 0781
    Audio 2: 0782
    Audio 3: 0783
    Audio 4: 0784
    VANC: 8147

The 1290 needs to be set for phase aligned audio.

The embedded audio should look like this:

  • 1: Main Program 5.1 Front Left
    2: Main Program 5.1 Front Right
    3: Main Program 5.1 Center
    4: Main Program 5.1 Low Frequency Effects
    5: Main Program 5.1 Rear Left
    6: Main Program 5.1 Rear Right
    7: SAP Channel
    8: Audio Description Service

If you need a stereo, it's necessary to generate your own downmix... unlike ABC, which sends stereo on 9 and 10.

Phase aligned audio keeps all the audio stems in phase with one another.

VANC stands for Vertical Ancillary Data Space. Here resides network test signals, closed captioning, network cuing and insertion signals, etc. Very similar to analogs vertical interval, but allowing a lot more data to be stuffed in the space!


Here's a bit more...

The NBC Skypath equipment rack uses Miranda equipment. But there are a number of ways to accomplish it: Cobalt Digital, for instance, makes cards that can do the downmix and hand off either AES or analog audio... or it can re-embed the downmix back into the original video. You can even take the analog outputs straight from the receiver and hit a mixer to create your stereo downmix, if you don't want to invest in a permanent solution.

To create a Left Only / Right Only (Lo/Ro) downmix:
Feed receiver analog outputs 1, 2, 3, 5, and 6 to an analog mixer, and set the mixer like this:

Channel 1 (Left Front): PAN LEFT, UNITY GAIN

Channel 2 (Right Front): PAN RIGHT, UNITY GAIN

Channel 3 (Center): PAN CENTER, –3dB GAIN

Channel 4, the low frequency effects doesn't get used in the downmix.

Channel 5 (Left Rear): PAN LEFT, –6dB GAIN

Channel 6 (Right Rear): PAN RIGHT, –6dB GAIN

This method creates a good stereo and mono-compatible downmix, but it does destroy front/rear separation for people trying to use a Dolby Pro Logic upmixer later. If that is important, you need to create a Lt/Rt downmix, which involves phase shifting channels 5 and 6 by 90 degrees.
 
Last edited:
Here's a photo of the NBC Skypath rack......a friend of mine at an unnamed NBC affiliate took it for me. The top two items are Quintech RF splitters. The next four are Tandberg RX-1290 IRDs....a few other unnamed items.....Miranda Densite rack frame that contain the downmix boards for converting the audio to 5.1. NBC has full control over this rack. Some of the IRDs are tuned to the NBC mux on AMC - 1 (Ku) and AMC - 18 (C Band). Station also has override control as well. The six devices in the rack to the left are AJA FS-1 Up/Down/Cross Converters.

NBC_Rack_1.jpg
 
Last edited:
Thanks for the research. Interesting information! Looking forward to the additional insight that the engineer might share.

A Google search yielded that the process is proprietary for DaySequerra.

I have seen their booth at NAB, but had never stopped. Will have to put it on my map next year.
There are other devices.....Miranda and Cobalt both have cards/devices that provide downmix capabilities.
 
Here's the breakdown for CBS audio:

Embedded HD-SDI
Group 1
Pair1
Ch1 (L)=SAP-or-DVS
Ch2 (R)=silence


Embedded HD-SDI
Group 1
Pair2
Stereo PCM
LT/RT

Embedded HD-SDI
Group 2
Pair1
Dolby E encoded (inaudible
data signal)
SMPTE 302
CH1/2 L/R
CH3/4 C/LFE
CH5/6 LS/RS
CH7/8 SAP-or-
DVS/silence

Stand alone
AES 1 BNC Ch1 (L)=SAP-or-DVS
Ch2 (R)=silence

AES 2 BNC Stereo PCM
LT/RT

AES 3 BNC
Dolby E encoded (inaudible
data signal)
CH1/2 L/R
CH3/4 C/LFE
CH5/6 LS/RS
CH7/8 SAP-or-
DVS/silence
 

Users Who Are Viewing This Thread (Total: 0, Members: 0, Guests: 0)

Who Read This Thread (Total Members: 1)

Latest posts