Video Quality -- Asking For Feedback

Hey, everyone. I've been failing to crack the code on YouTube video quality, but I may have something that I'm willing to tolerate for now. I'm looking for some feedback, and below are two videos that are identical except for one thing. I'm not going to tell you what that thing is, and while some of you may notice it, others may not (it's okay if you do not). Now, this isn't a challenge to figure out what's different because it should be fairly obvious to most, but I'm asking that if you watch a specific part of one video, that you at least watch the same part of the other video to compare. The video is 6:20 in length and features my low-level, mixed-class team in scenes from two different dungeons. Plenty of mistakes were made during these dungeon runs, but we're not here to focus on those. >_>

As for viewing the videos... I'll ask that you watch them in a VP9-capable browser—so, either Google Chrome or Opera. If you must watch the video in Firefox because you absolutely abhor the other two choices, then you will be limited to 1080p60 using the AVC codec (slightly less quality), and I ask that you state that you could only watch the video in FireFox if you're voting or giving feedback. You also don't have to watch the videos at any special resolution, but the higher the resolutionb the better quality you should get (1440p60 requires ~15Mbps downstream in order to avoid constant buffering).

Thanks.


Video 01

Video 02

Quote:

---

Originally Posted by MiRai

I'm not going to tell you what that thing is, and while some of you may notice it, others may not (it's okay if you do not). Now, this isn't a challenge to figure out what's different because it should be fairly obvious to most

---

TEASE!

Honestly, I'm of mixed feelings. I like the scene-wide consistent clarity of the first one from the perspective of being able to keep track of what's going on a bit better. That said, from the perspective of the player, I can't help but wonder how that difference would fare in an area that caused spiky framerates; yours are pretty consistent through this vid, though it is a zone that's somewhat easy on the renderer. Honestly, I've never had a beefy enough system to develop a personal baseline for running that level of detail at those resolutions/framerates, so there's that. :P

All that aside, it still looks better than most of the gaming vids I watch on youtube. ;)

Very good once the HD kicks in. Didn't put any effort into finding out whats different but both look exactly the same to me.

Obviously read only the first half part of your question and watched them side by side in firefox at 1080p. No difference to me.
Then read the second half of your post, watched them full screen side by side in chrome in 1440p, (proof !) and still can't tell the difference. It goes like: hm the color tone of the left is lighter, and 2 seconds later: nope it's actually lighter on the right one. Etc ...

New shady way to get views?

Quote:

---

Originally Posted by Ughmahedhurtz

TEASE!

---

Haha -- I had to give a reason for why I wanted people to watch two of the same video, and I assure everyone that they are the exact same video, content-wise, rendered two different ways.

However, I'm glad that the initial response is that people don't notice a difference, because I personally prefer one version to the other. :)

Quote:

---

Originally Posted by Ughmahedhurtz

Honestly, I'm of mixed feelings. I like the scene-wide consistent clarity of the first one from the perspective of being able to keep track of what's going on a bit better. That said, from the perspective of the player, I can't help but wonder how that difference would fare in an area that caused spiky framerates; yours are pretty consistent through this vid, though it is a zone that's somewhat easy on the renderer. Honestly, I've never had a beefy enough system to develop a personal baseline for running that level of detail at those resolutions/framerates, so there's that. :P

---

Would you mind elaborating on this? I'm not sure I fully understand. Answered via PM.

Quote:

---

Originally Posted by zenga

Obviously read only the first half part of your question and watched them side by side in firefox at 1080p. No difference to me.
Then read the second half of your post, watched them full screen side by side in chrome in 1440p, (proof !) and still can't tell the difference. It goes like: hm the color tone of the left is lighter, and 2 seconds later: nope it's actually lighter on the right one. Etc ...

New shady way to get views?

---

No color correction has been added to either video, and if I wanted views, I'd make the videos public, or go post this same type of thread on reddit. :)

Thanks to everyone for the feedback, so far.

Quote:

---

Originally Posted by MiRai

However, I'm glad that the initial response is that people don't notice a difference, because I personally prefer one version to the other. :)

---

If it turns out to be a toss-up so you'll simply pick your own preference, will you reveal whether you went for #1 or #2 and why, or will it just be something to further enhance the mystery surrounding the man, the myth and the legend that is... MiRai? :)

Quote:

---

Originally Posted by MultiBear

If it turns out to be a toss-up so you'll simply pick your own preference, will you reveal whether you went for #1 or #2 and why, or will it just be something to further enhance the mystery surrounding the man, the myth and the legend that is... MiRai? :)

---

Yes, I will reveal which one I choose in the end, as well as the pros and cons that come with it. :)

I would have to say I like both.

After some careful looking, I think I see some slightly darker areas in video 2, that in video 1 have some detail that comes throughand with video2 there is some blurring in various movements whereas in video1 I couldn't spot them, but you have to be looking real close, and doing a frame by frame comparison to see any difference to my eye.

Personally I also think I could be imagining it all.

I initially thought there were colour differences too, but after moving things around, I see that is my monitor. When I moved the video over to my "professional colour accurate" monitor, those differences went away.

Might have to change from Auto to a specific HD resolution to compare them.

Pretty sure I am viewing at 360p for the first 15 seconds before the HD flag is on, and once it turns on it doesn't say which of the three HD resolutions I am viewing at.

I was watching both at 1440, and waited until the obvious change from 480 to something much nicer before I started playing pause, play, pause. Then I picked a cached timecode so it didn't need to redownload, and began bouncing back and forth to it.

I considered downloading them and doing an offline compare, but then I wasn't sure if I was going to get format changes.

But a frame by frame does show little differences. e.g.
Top pic is from video 1, bottom from video 2. I've replayed and grabbed this frame a bunch of times from different sessions to make sure that the video is displaying correctly and coming down at 1440 on both sides.
Look at the blur trail from the mouse overlay (among other things that don't look quite as crisp).
http://i.imgur.com/JdJ73RV.png


There are static bits that look exactly the same to me, like the light fixture on the wall on the right, the floor, the back wall and fixtures on the left. The corpse eater being engulfed. But then there are the blurred things. The corpse eaters to the right, the Frantic guys on the left, the girly night elf (?) in the inferno, the star lighting effects.

Maybe it is just my end. I don't see this at all while it is actually playing. I have to get real picky and go frame by frame to see it.

One looks a little blurry during movement, the other does not.

I could however be imagining it! this is fun, please please let us know the differences.

I almost forgot about this thread... sorry. :)

The second video did, in fact, have motion blur added to it which, in my opinion, helps makes things a bit smoother (and more action-packed). Adding the effect does not increase the file size, but does increase the render time by a very large amount. However, I've since been working on some other recording and rendering techniques that I'm pretty pleased with, so I doubt I'll be going with the motion blur.

Thanks, again, for the feedback.