Originally Posted by Josh Z
Did you try switching your Blu-ray player to PCM output so you could compare DSU with Neural:X in that scene? I'd be curious what happens with regard to that helicopter pan.
Speaking as someone reading this thread with interest but who does not have a DTS:X capable receiver yet, your description here seems to be consistent with the theory that Neural:X behaves a lot like the Auromatic upmixer, and largely copies information from the ground level channels to duplicate it up top. If a 5.1 or 7.1 mix has a lot of activity going on in the left and right and surrounds, that will translate to a lot of activity up above those speakers too.
But, again, I haven't been able to actually hear this for myself yet.
I did do an A/B with PCM out of the player. The difference is much less pronounced than my initial impression. I used the same beach storming scene in Edge of Tomorrow.
The helicopter pan with DSU is much closer to Neural X than I first thought. The Neural X pan was still better. I listened several times to each to put my finger on why I liked the Neural X better: It just sounded more precise. The DSU pan still did a good right to left pan but it sounded much more like ambiance, less pinpoint but the difference is much less dramatic than I first thought. I'm guessing there's plenty of people who would prefer that effect more, less localization, etc. But for me, I do like a little more precision.
The rest of the battle scene had lots of ear level weapon fire. I listened to both upmixers with and without the base 7 channels. I liked the Neural X better. Again it was the precision. It achieved imaging of a shell whizzing just above the left side of my head and landing at some point to the right and in front of me. That scene is full of such weapon fire at or just above ear level. With DSU, the shells whizzing by were more ambient sounding, not as precise.
My initial impression that Neural X puts more in the overheads still stands but on second thought it not as easy as that. It does seem that what it does choose to steer to the overhead is louder. In fact, I had to turn my ceiling speakers levels down with Neural X. I had them +3 db hot from where Audyssey set them for use with Atmos/DSU. Turned it down to +1.5 db because I felt the tops in Neural X were too overwhelming. However, there are points in that beach scene that's not always the case.
1) At one point while they were getting ready to drop out of the copters onto the beach, the copter takes a hit and the soldiers have to prematurely drop. Neural X put tons of very loud crashing noises in the tops. The crashing noises were extremely faint in DSU tops. It choose to leave the majority of that effect in the base layer. Even at that, the base layer crashing noises were not as loud as Neural X effects in the top speakers. It seemed to me that it was exaggerated and a bit excessive.
So it seems easy to conclude Neural X just puts more up there, however:
2) During the battle on the beach, both upmixers placed some shouting and screams in the tops. Typical phrasing and exhortations you'd expect to hear in a frantic battle. DSU in this case put much more into the tops. I could hear many distinct voices and phrases. Specifically, I heard "Let's lock and load" being said from one of the tops. The exact same moment with Neural X, nothing. I dunno, it probably showed up in the base somewhere. I was specifically looking for it in the tops though. Neural choose to make those vocalizations steered into the tops into ambiance, more muffled shouting noises. So pretty much exactly the opposite of how I thought of the helicopter pan.
So this is all based on a 10 minute segment of one movie. I'll have to listen some more but I'm liking Neural X more than DSU right now. I wish I could somehow upmix the Super Bowl in Neural. Have watched tons of football in DSU, would like to compare. Simple summary, both are good. Closer than I originally thought. The things Neural X steers to the tops seem louder. I have no experience with Auro but from descriptions of it, it does not seem like Neural X works that way. The top speakers are not repeating what the base layer is doing.