Simply random stuff

[mandrix]

I remember it being discussed in that long thread regarding Dolby TrueHD...although whether it's actually perceptible is a separately story. It's not limited to Atmos, DTS:X in object mode does the same thing, its just the nature of object-based audio at home. You can't deliver the full 118 objects or however many it is, so you have to compromise and use some psychoacoustics to get the same effect when rendered on larger speaker arrays.

OK. Well my interest is limited to 7.1.4. If you come across any documentation I'd be eager to read it.

 

[inspclouseau]

Many years ago I actually made my living by working in various fields of photography. At one point I had the pleasure of working with two photographers from The Missoulian. They were publishing a "coffee table" book of the beauty of Montana. As you might expect it was strongly outdoor & wild life shots.They came to KC because, thanks to Hallmark cards, we had a thriving industry of photography, pro photo labs, color sep houses, and the country's leading fine art publisher, Lowell Press. And coming to KC meant they could be involved directly with every step of the process. Like going into a restaurant's kitchen to tell them exactly how much salt you want.

My main point I'm slow to getting around to, is that so many of your pics remind me of that project. Just casual shots on a phone? I'd say it looks great on my monitor. I really enjoy your photography & I hope you blow up some of them for wall display.

Edit: Then again I guess you can just step outside your back door & see it even bigger, eh?

Sonik Wiz, First off THANK you! I usually try not try to get too personal. on the internet. But I wanted to say this~ I used to hunt, fish and be out in the woods as often as I could. From just a little sht kid, and throughout any spare time from work and raising a family. BLESSED! This until pretty recently. Seen so many incredible things in my life, but never had a camera during those times. I have had 3 strokes in the last 5 years. The last two really messed me up pretty good. Lost much of the feeling on my left side etc etc You know stroke related 'stuff'. This was originally due to an injured neck that caused a dissected carotid. Got hit HARD. I never knew until years later the damage that was caused. My melted, dumb ass face did finally go back into whatever it was before. My left hand/ side though, a bit more usable but very little feeling at all. I was not able to talk or write for about a month. I was a lefty, but them days are over now. Now I am a compete CLOD. Go throw a baseball wrong handed. You know the deal. I break shit a lot. So frustrating. Now in the last couple of years, I have been prescribed a blood thinner and man I HATE that crap. I often come back into the house with a small cut on hand or arm/ leg etc and looks like a murder scene. Ridiculous. Scares Kathy and I do not liking doing that. I have done enough of that. She is why/how I am here. She has always been everything to me. (Different story and I get upset) Anyway, I can drive and now if I want to take off on my own, (runnin' around in the mountains next to us.) she gets scared still and I do understand. We've been friends since 11 years old, married as teenagers. (the thumb image) She says I will go with ya too. Geezzz.. I do not want her making herself go and I KNOW why. Hey..this is pretty good typing practice..... Anyway.................
I get off track ..
Gear~ I use a few cameras now and I just wing it. Learned stuff as I went. I sure enjoy it and believe me ANYONE can! Just save up and get one, then run around with it figuring it out. I do not use cell phones for anything except if it rings.. and even then I tend to ignore it.. I us Canon dslrs and their lenses. The Canon menus were always easiest for me to understand. I actually had a dude and then a few months later a lady get hold of me wanting to purchase a couple of images from me. I gave them to them to keep. Not selling. Anyway..this episode of 'this my life' is now over. I just enjoy others liking what I can come up with on a camera sometimes. All the very best to you and yours. Oh by the way. MIZZOOLA! We stay clear of cities... This image was nice because I just sat on the front deck an got this one while drinking coffee..View attachment 1741026214344.jpeg

 

[mandrix]

Dolby Atmos on Blu-ray is not entirely lossless (due to the way it stores heights as objects and how it extracts them) and Auro3D's static height speaker channels are lossy. Well, DTS:X has largely been unused and forgotten, but has anybody done any experiments with it? Because from what DTS claims, unlike Auro3D and Atmos, DTS:X can actually store up to a 7.1.4 mix on Blu-ray losslessly, before going into object-based shenanigans.

OK. I'm calling BS. I can't find any reference that says there is anything lossy in TrueHD/Atmos. I will require proof to the contrary.
Not from forum posts, from Dolby.
The gauntlet has been thrown.

 

[Mr. Afternoon]

OK. I'm calling BS. I can't find any reference that says there is anything lossy in TrueHD/Atmos. I will require proof to the contrary.
Not from forum posts, from Dolby.
The gauntlet has been thrown.

Well, lossy is not exactly the right term in a traditional sense. Atmos stores it's object data losslessly in TrueHD, it's just the method of nulling those objects from the base mix and placing them is not a perfect null. I'm still trying to find the source, if I don't find it, pretend I never mentioned it.

 

[mandrix]

Well, lossy is not exactly the right term in a traditional sense. Atmos stores it's object data losslessly in TrueHD, it's just the method of nulling those objects from the base mix and placing them is not a perfect null. I'm still trying to find the source, if I don't find it, pretend I never mentioned it.

What exactly defines the "base mix"? Is this different from bed channels?

 

[Mr. Afternoon]

What exactly defines the "base mix"? Is this different from bed channels?

I meant the bed channels...sorry my terminology is out of wack.

TrueHD stores Atmos as a bed stream and objects, and those objects are nulled out of the bed stream (similar to CD-4) and placed in the speaker array. From what I read (I think the thread actually got deleted due to the amount of discourse, I can't seem to find it), the nulling process that Dolby uses in TrueHD Atmos is not perfectly lossless; it doesn't result in a perfect null.

I'm actually going to try do a null test between a 7.1.4 mix and a decoded Atmos file that was created from a 7.1.4 file to see if they null out perfectly...if they don't, I guess that's proof?

DTS:X claims to be able to carry 7.1.4 (and 5.1.4) without encoding to objects...while Atmos, if I understand correctly from the white paper, does not allow anything past 7.1.2 before object encoding. If Dolby's "nulling" process isn't perfect and causes artifacts (even if it's inaudible), and DTS:X truly can carry unaltered 5.1.4 or 7.1.4, I think that's moral grounds to recommend DTS:X as a carrier for music mixed to 7.1.4 (as there's a substantial amount of Atmos mixes that are really just 7.1.4, and don't utilize the additional speakers offered by larger speaker arrays) rather than Atmos. Plus, DTS doesn't restrict who buys their encoder (although they do charge 3x the price! Ouch!).

 

[mandrix]

I meant the bed channels...sorry my terminology is out of wack.

TrueHD stores Atmos as a bed stream and objects, and those objects are nulled out of the bed stream (similar to CD-4) and placed in the speaker array. From what I read (I think the thread actually got deleted due to the amount of discourse, I can't seem to find it), the nulling process that Dolby uses in TrueHD Atmos is not perfectly lossless; it doesn't result in a perfect null.

I'm actually going to try do a null test between a 7.1.4 mix and a decoded Atmos file that was created from a 7.1.4 file to see if they null out perfectly...if they don't, I guess that's proof?

DTS:X claims to be able to carry 7.1.4 (and 5.1.4) without encoding to objects...while Atmos, if I understand correctly from the white paper, does not allow anything past 7.1.2 before object encoding. If Dolby's "nulling" process isn't perfect and causes artifacts (even if it's inaudible), and DTS:X truly can carry unaltered 5.1.4 or 7.1.4, I think that's moral grounds to recommend DTS:X as a carrier for music mixed to 7.1.4 (as there's a substantial amount of Atmos mixes that are really just 7.1.4, and don't utilize the additional speakers offered by larger speaker arrays) rather than Atmos. Plus, DTS doesn't restrict who buys their encoder (although they do charge 3x the price! Ouch!).

Well...I would rather see something definitive from Dolby. You well be right, but I never heard it in that way before, that is, having to null the objects between the bed and height? layers.
I notice that with the lossy encoding from the streamers, there is often only one bed channel, Lfe, and the rest is objects, at least according to Mediainfo.
Since I seldom break down a Blu Ray beyond .iso, I don't think I've examined on a track level with Mediainfo in quite some time how it recognizes bed/objects.

Still, it's fascinating. My only goal here is to get more knowledge.

 

[Mr. Afternoon]

Well...I would rather see something definitive from Dolby. You well be right, but I never heard it in that way before, that is, having to null the objects between the bed and height? layers.
I notice that with the lossy encoding from the streamers, there is often only one bed channel, Lfe, and the rest is objects, at least according to Mediainfo.
Since I seldom break down a Blu Ray beyond .iso, I don't think I've examined on a track level with Mediainfo in quite some time how it recognizes bed/objects.

Still, it's fascinating. My only goal here is to get more knowledge.

Dolby does definitely define a nulling process is happening, as Dolby uses nulling for actually storing channels. This is described in the Dolby TrueHD White Paper, and can be witnessed looking at the source code for open-source decoders such as FFMPEG. I'll recap what I said in the missing thread.

See, when TrueHD was still Meridian Lossless Packing and still being developed, one of the requirements was to minimize the work lower-end players have to do. So, MLP was designed to store a downmix (which could be decoded by low-end stereo players easily) and recovery channels. What MLP would do on higher-end multichannel players is take the downmix, null out the rears, center, and sub from it, and place them in the respective channels. The end result would be a reconstructed 5.1 mix that perfectly matches the original source file.

When MLP got changed to TrueHD and 7.1 sound hit the scene on Blu-ray, Dolby thought, "Hey, we'll just do the same thing, it makes sense and it's backwards compatible" so they made the two extra channels for 7.1 was made into extra data that 5.1 players could ignore. Same lossless nulling principle.

So at this point, TrueHD stored its audio like this:

Section 1: Stereo Downmix
Section 2: Recovery Channels to Reconstruct 5.1 Mix
Section 3: Recovery Channels to Reconstruct 7.1 Mix

And if your player didn't support a section it could ignore it. Well Dolby calls them TrueHD sub streams but people get that confused with actual sub streams of a feature so I'm calling them sections.

Now with Atmos, Dolby did the same thing again, adding a 4th section. However, we've got objects now, and so we need to store some position data and stuff.

The thing that calls into question and requires testing to verify claims is that unlike 5.1 and 7.1, Atmos in TrueHD is not getting encoded from a static set of channels that it can perfectly match the data of. Atmos is encoding from a ADM BWF, which means the Atmos encoder is doing some processing to fold that humongous data down to a normal stream, and the decoder is doing some extra work reproducing the positions of said objects and nulling them out from the bed. And object nulling from the bed may not be entirely lossless in theory, as it may be relying on some psychoacoustic nulling rather than pure simple math (at least that was what was implied).

Rather long message from someone like me but I hope that explains where I'm coming from. I am trying to understand how Atmos audio at home is stored, but it's a bit harder when the secrets are locked away!

 

[mandrix]

Dolby does definitely define a nulling process is happening, as Dolby uses nulling for actually storing channels. This is described in the Dolby TrueHD White Paper, and can be witnessed looking at the source code for open-source decoders such as FFMPEG. I'll recap what I said in the missing thread.

See, when TrueHD was still Meridian Lossless Packing and still being developed, one of the requirements was to minimize the work lower-end players have to do. So, MLP was designed to store a downmix (which could be decoded by low-end stereo players easily) and recovery channels. What MLP would do on higher-end multichannel players is take the downmix, null out the rears, center, and sub from it, and place them in the respective channels. The end result would be a reconstructed 5.1 mix that perfectly matches the original source file.

When MLP got changed to TrueHD and 7.1 sound hit the scene on Blu-ray, Dolby thought, "Hey, we'll just do the same thing, it makes sense and it's backwards compatible" so they made the two extra channels for 7.1 was made into extra data that 5.1 players could ignore. Same lossless nulling principle.

So at this point, TrueHD stored its audio like this:

Section 1: Stereo Downmix
Section 2: Recovery Channels to Reconstruct 5.1 Mix
Section 3: Recovery Channels to Reconstruct 7.1 Mix

And if your player didn't support a section it could ignore it. Well Dolby calls them TrueHD sub streams but people get that confused with actual sub streams of a feature so I'm calling them sections.

Now with Atmos, Dolby did the same thing again, adding a 4th section. However, we've got objects now, and so we need to store some position data and stuff.

The thing that calls into question and requires testing to verify claims is that unlike 5.1 and 7.1, Atmos in TrueHD is not getting encoded from a static set of channels that it can perfectly match the data of. Atmos is encoding from a ADM BWF, which means the Atmos encoder is doing some processing to fold that humongous data down to a normal stream, and the decoder is doing some extra work reproducing the positions of said objects and nulling them out from the bed. And object nulling from the bed may not be entirely lossless in theory, as it may be relying on some psychoacoustic nulling rather than pure simple math (at least that was what was implied).

Rather long message from someone like me but I hope that explains where I'm coming from. I am trying to understand how Atmos audio at home is stored, but it's a bit harder when the secrets are locked away!

As always, I respect your input.

So in effect, you're saying that Dolby has not defined exactly what this process is?
I get what you're saying from a theoretical point of view, and in some respects it makes sense, ...and I'm fairly well versed in MLP as pertains to the original Meridian concept for DVD-Audio, but without evidence it still lacks traction.

If this is indeed secret sauce by Dolby, I guess I would ask for definitive proof. I would not confirm or deny without more information. That information may/may not be available, IDK.

 

[Mr. Afternoon]

As always, I respect your input.

So in effect, you're saying that Dolby has not defined exactly what this process is?
I get what you're saying from a theoretical point of view, and in some respects it makes sense, ...and I'm fairly well versed in MLP as pertains to the original Meridian concept for DVD-Audio, but without evidence it still lacks traction.

If this is indeed secret sauce by Dolby, I guess I would ask for definitive proof. I would not confirm or deny without more information. That information may/may not be available, IDK.

Dolby has defined it for the regular 5.1/7.1 stuff but not Atmos, yeah. That's where my skepticism comes into play. Again, planning to do that 7.1.4 discrete vs decoded Atmos test, cause that's at least SOME sort of evidence. When I get the time.

 

[mandrix]

Dolby has defined it for the regular 5.1/7.1 stuff but not Atmos, yeah. That's where my skepticism comes into play. Again, planning to do that 7.1.4 discrete vs decoded Atmos test, cause that's at least SOME sort of evidence. When I get the time.

Sounds like a plan!