Thread: 1.0 vs ARR vs DT For Context

Originally Posted by GreenLeafy

This is so crazy jarring to look at. The lighting and model changes in the benchmark do such an insane disservice to literally every single one of these - they all look completely stripped down of defined features.They just look extremely dull and lack any sort of depth. The clothing texture improved though lol.

I really, really, really wish we had better control of the lighting in the benchmark. I found once I swapped over from the static standing pose and over to the "unique" pose and got real movement in the character, I saw a much better lighting response on the 7.0 models.

Originally Posted by Ath192

You're not, what you were asking was a really hard request, here it is though. I have also updated the Thread front post with it. I hope this provides you the details you need and helps everyone else make feedback or compare.



Remember to right click and open in new tab.

yea idk this makes ARR -> DT look good actually they are very close to me. DT just has obvious diff lighting.

Originally Posted by KhujaRela

I'm sorry to be a bit offtopic, but there is something that's been bugging me since I saw the first videos of 1.0 I feel they tend to be more realistic, and also the Elezen were better proportioned?? I saw Mother Myoune to look much better in 1.0, and with narrower shoulders (in my personal opinion)

I wonder why this changed so much with A realm reborn? I don't say I like ALL better in 1.0, I certainly not (since I prefer less realism in general, I like more colorful and stilized worlds). It's because the engine or change artistic direction? If someone knows...

They changed engines which will have an inherent effect on how the game looks and presents itself, but they also made the visuals worse on purpose so that hardware (at the time) could actually run it. 1.0 was very intensive for 2010 and people really struggled even with the best PCs.

That, and they wanted to streamline the development process. Obsessing over visuals will really slow down how quickly you can make things.

Thank you both! so it was an engine/specs thing, I believe this also has to do with current graphics change, since they said they would use from Dawntrail the same mesh number on all characters and races equally (also to make easier adjust clothing), so maybe this is affecting to some face features and causing the glitches on the benchmark...

With this clarified, I want to think now they didn't had enough time to fine-tune faces, or maybe they wanted to see the public reaction on benchmark first? I'm just guessing tho, until they talk about I don't think we know for real...

Originally Posted by KhujaRela

I want to think now they didn't had enough time to fine-tune faces

What's really strange is that there was a theory running around in the JP community that Dawntrail would release in late August. At Tokyo Fanfest, when they didn't announce the release date yet due to wanting to make sure they would not delay like EW, Yoshida even brought up a chat comment that said "it's gonna be October" to confirm that "no, it will be summer". Which proves that people were already expecting late summer.

And then at PAX he says "oh we wanted to release one week earlier, actually, but Elden Ring DLC" etc.

I was also mentally prepared for a late August release (even before reading about that theory), so when they announced early access late June I was "HUH?" (and suddenly I'm pressed for time with all the things I had planned to do before the 48h maintenance lol).

There was always bound to be some amount of negative reaction with the benchmark, but I think this is a lot more than what everyone expected. It does feel a little rushed (how did they miss the weird square shadow on the Hyur M shoulder, the dent in the Roegadyn M skull, the rigging issues on the Elezen F mouth, the scale artifacts on dark-skinned Au ra), and now they definitely don't have time to fix all that before release, they're probably in post-production already at this stage with the media tour coming up.

Content drought aside, I think no one would have been shocked if the release date had been in August (some would have complained, sure, as always, but then again... Elden Ring DLC), and with the same release date for the benchmark, it would have given them time to react to player feedback.

Also feedback wise: why do they send a survey to ask us what our favorite mounts and minions are to make funny statistics, and why did they not provide a survey with the benchmark when it was the appropriate time to gather players' opinion? The EN forums are an even bigger mess than usual since Sunday and if it's already hard for players to find the threads they were following in this tsunami of "here's my individual opinion on my own character, I love it/I hate it" threads, I can't imagine how much of a pain it's gonna be for the community team (or whoever looks at these threads) to try to synthesize what players' requests are.

To go back on topic:
Thanks for providing the comparative pictures! (I think you or maybe someone else mentioned it in the London Fanfest feedback thread already but couldn't find them at the time)

However, I also think that M'iqote F is probably the model the least affected by the benchmark changes (fangs aside ofc), so I'd be curious to see comparative pictures of all 1.0 races (if anyone has a private server around...)

... I'm gonna go insane I actually find the 1.0 lighting the most interesting.

I know the lighting environment is not the same (if you look at the Roegadyn's hair you can clearly see blue and orange lights, and the other screenshots are lit from behind), but it feels like the way light bounces on the skin feels the most varied. The Highlander especially, even if his eyes look dead you can really appreciate the "relief" of the skin, it's not all hard edges like ARR-EW but it's not all diluted like the benchmark either.

Same if you look at the M'iqote, there's just the right amount of gloss on her lips and nose without it looking too oily (looking at you, benchmark), and her shoulders manage to look so soft without looking like plastic.

I think the light overpowering the materials a lot less is also helping, like ARR-EW cruelly suffers from light overexposure, but it's still present in the benchmark if you look at the white parts and compare them to 1.0 (again, not the same environment and lighting, I know, but I feel the materials react differently to the light).

Now I'm looking at videos and damn... That's some really nice skin!!

You can still run the 1.0 benchmark. I ran it on my 2024 system, 7800x3d CPU, 7900xtx GPU, 64GB of ram, and even though I scored the highest bracket, it still stuttered and dropped frames. Because it's not anywhere near optimized. You can have the best PC in the world and it won't run right. 1.0 characters looked great, but 1.0 zones were a mess, and the performance was nothing worth talking about.

That said, I wish we could have character curves again.