Coul FDev roll back the planetary update and shaders?

They both look like "works ok for me" or "works ok for some", which is what I was saying. My gx970 is old and slow, and the game now post update 9 is playable. It should be faster, no doubt, but it's playable so for me it "works ok". I would hope it will work even better with more optimisations, as I suspect I'm looking at roughly where las gen console versions would be.
I don't get your logic. You're saying that you're ok with the FPS on your machine so the game work ok even if it should works better ?
But if the game should work better, that imply that the game doesn't work ok ? What is the logical subtlety that i miss ?

it's playable so for me it "works ok"
What do the inverted commas mean ? Does that mean that the game works ok but only from your point of you ? If it's the case, what the difference with the "Work on my machine" certification ?
 
I don't get your logic. You're saying that you're ok with the FPS on your machine so the game work ok even if it should works better ?
But if the game should work better, that imply that the game doesn't work ok ? What is the logical subtlety that i miss ?
Gosh!

It just appeared to me that saying "I can play with 60 FPS with 98% GPU load" cannot be tempered with "But it would be better if I could have that 60 FPS with 30% GPU load"... Is that not logical?
 
With the 2080S I use the difference is between half and 1 third the FPS of Horizons in Odyssey. I'm only using the one GPU, granted.
Which illustrates beautifully just how unpredicatable performance in Odyssey is, I'd say.
And to be fair, there very well could be bugs with how SLI is implemented in Odyssey beyond it just being inefficient. The FPS does seems to kind of strangely lock-in to the mid to low 30s, and for a couple of the Odyssey updates (maybe it was 4 and 5 or 6, or something like that) I was getting into the 40s and 50s, and that was before I repasted my GPUs and got them running cooler and more like new again – it gave me a decent boost in FPS performance in Horizons from the high 100s to the mid 200s, but very strangely didn't seem to change Odyssey's performance at all. I'm not sure how rigorously Frontier have tested SLI in Odyssey, if at all. 🤷‍♂️

I'll do some more comparisons and post them in the thread I made that you suggested sooner or later.
 
Last edited:
Gosh!

It just appeared to me that saying "I can play with 60 FPS with 98% GPU load" cannot be tempered with "But it would be better if I could have that 60 FPS with 30% GPU load"... Is that not logical?
A bit off topic with the conversation, no ?
It seems that we mostly agree that the Odyssey should work better, not that it would be better if it could.
No one has mentioned the GPU load. And without more information, we can't even say if 60fps with 98% GPU load is ok or not.
We are talking about the logic of correlating personal feelings and configuration to judge the overall performance of Odyssey (the fps is ok for me on my config so the game work ok. Or sometime "work ok").
 
Does it matter what we think about other's opinions on it though, other than those at Frontier, I suppose?

If it works well for me (which it doesn't), that doesn't invalidate the legitimacy of it working poorly for someone else. Likewise the other way around.

...

I think most of us can likely agree that Odyssey does seem to have some significant performance issues for a significant number of players. But if not, well, you know, this is the internet after all. :giggle:

That said, it does seem to be improving for many players as well, though not for me nor apparently some others so far.
 
Last edited:
Hmm, well, if we assume a certain given level of detail in ships and SRVs above world surfaces, I'm not sure why that same total level of detail couldn't be used. What I mean by that is that in a ship or an SRV, you're seeing more overall so the PPI (as it were) is in decent quality. That same level of PPI of clarity closer up shouldn't necessarily be more demanding, though they would need to rescale the textures accordingly somehow. Sort of like when you're higher up above a world's surface it isn't generating the full clarity of everything as though you were on the surface. That would be a huge waste of resources. So being on foot would just warrant a similar approach, at least in that regard.

Either way, I have some unused VRAM, so some higher resolution textures shouldn't be too much of an issue. I've taken some nice 16K screenshots in Horizons after all.

But yeah, maybe lighting. My only issue with it in Horizons was multiple light sources.

Granted, I'm not using VR, and I generally only play at 1080p.
I guess what I really think is fdev made a decision when originally designing (umm.gif) the features that were going to be implemented for EDO, and that was to upgrade lightng/textures/planetary surfaces/planetary generation & topography instead of just adding the new features in and (as you say) just upping the texture levels to work better at that scale.

And 2 years ago, or even 1 year ago, I can't imagine any of us here would have said anything other than "YES! That's what to do!" if we'd been asked if we wanted all of it.

And it makes sense, I've no doubt the devs and artists were desperate to bring ED up to scratch in terms of graphics, I remember one of the devs pre-release saying something along the lines of "finally we have pbr".

The problem is it's broken, in some way. Maybe more than one. Best intentions etc.

So basically I can see your point and I can see that yes it would have been possible* to add the headline EDO features in to Horizons, but I think it would have been so underwhelming.

*caveat: it might not have been possible, we don't know. Atmospheres just might not have been possible without a large part of the new code that is causing the problems.
 
I don't get your logic. You're saying that you're ok with the FPS on your machine so the game work ok even if it should works better ?
But if the game should work better, that imply that the game doesn't work ok ? What is the logical subtlety that i miss ?
Yes, I am saying that I'm ok with the FPS on my machine so the game works ok, even if it should work better. I expect them to improve it further.

I think you are dancing on the head of a pin, and I'm not sure why. But that's also ok.

What do the inverted commas mean ? Does that mean that the game works ok but only from your point of you ? If it's the case, what the difference with the "Work on my machine" certification ?
I'm not sure what any of this means.

I'll restate my simple view: EDO runs okay (not great) on my gtx970, and is playable. I hope and expect it to improve further. Your mileage may vary, and I imply no prediction or guarantee of performance for any other user, or hardware.
 
I'm no graphics expert, but surely that simple shader for the atmospheric effects can't be that disruptive, can it?
I wouldn't have thought so, no, as I said I imagine fdev could have added EDO features in to Horizons but my point was just that we users don't know if there was some technical problem with that, or something else.

As I said above, I imagine upgrading everything was a creative decision which none of us would have disagreed with if asked at the time.
 
I guess what I really think is fdev made a decision when originally designing (umm.gif) the features that were going to be implemented for EDO, and that was to upgrade lightng/textures/planetary surfaces/planetary generation & topography instead of just adding the new features in and (as you say) just upping the texture levels to work better at that scale.

And 2 years ago, or even 1 year ago, I can't imagine any of us here would have said anything other than "YES! That's what to do!" if we'd been asked if we wanted all of it.

And it makes sense, I've no doubt the devs and artists were desperate to bring ED up to scratch in terms of graphics, I remember one of the devs pre-release saying something along the lines of "finally we have pbr".

The problem is it's broken, in some way. Maybe more than one. Best intentions etc.

So basically I can see your point and I can see that yes it would have been possible* to add the headline EDO features in to Horizons, but I think it would have been so underwhelming.

*caveat: it might not have been possible, we don't know. Atmospheres just might not have been possible without a large part of the new code that is causing the problems.
That's fair, but likewise there's no way I would have expected to take a seven fold hit to performance. I'm saying this with the assumption that they're going to continue sorting it out where they are able. It's only if they aren't able to significantly improve performance for me sooner or later that I'd rather they stuck with the graphics rendering of Horizons.

I've been looking at some gear for other computer related interests, and preemptively avoiding the scalpers or outwitting them has been like playing 3D chess crossed with wac-a-mole. I don't see getting new graphics cards anytime soon, but also, I don't think they should really be needed for the game for me anyway.
 
Last edited:
I think the takeaway from these recent debates is that "expected performance" is fundamentally different than "acceptable performance." The former is objective, the latter subjective.

When people say performance is still bad even after Update 8, they are probably talking about expected performance. When someone says "its good enough for me now" is talking subjectively. What's more, you can hold both of those opinions at the same time; expected performance can still be "bad" even if you consider the current performance as "acceptable" for yourself. You dont HAVE to explicitly take one side or the other.

And again, expected performance is objective but it's also comparative. I've mentioned it before, but expected performance is determined by how one game stacks up to every other game, and what engines they use and what the visual outcomes are relative to performance. A game rendering a simple skybox and nothing else, but getting only maximum 45fps would probably be considered objectively bad if it's using standard techniques. But a game simulating Rayleigh scattering in real time with volumetric clouds etc etc running at 45fps would be considered pretty good.

Ergo, Odyssey running at a max of 60fps in settlements despite the game not really employing any unique or advanced rendering techniques, and overall visually being pretty unremarkable compared to other relatively new games, means the game performance is objectively not good, even if 60fps max is "fine" for you.
 
Well, not just. I think there are more issues with Odyssey's performance in comparison to Horizons, though yes, SLI is a very significant one for me. SLI alone shouldn't reasonably account for as huge a difference of 30~35 FPS from 230~260 FPS; a factor of about seven times. I'm not using 7 cards in SLI after all, and SLI isn't 100% efficient.
To be fair, you are kind of an edge case. I would need to check but I think I get ~200 FPS on pretty recent hardware in the Odyssey hangar (which is your benchmark IIRC). Obviously I don't get 1400 FPS in Horizons, so I guess the comparison is slightly flawed.

No excuse for the bad performance on your end, just pointing out that (generally) Odyssey doesn't perform 7 times worse my experience.
 
I think the takeaway from these recent debates is that "expected performance" is fundamentally different than "acceptable performance." The former is objective, the latter subjective.

When people say performance is still bad even after Update 8, they are probably talking about expected performance. When someone says "its good enough for me now" is talking subjectively. What's more, you can hold both of those opinions at the same time; expected performance can still be "bad" even if you consider the current performance as "acceptable" for yourself. You dont HAVE to explicitly take one side or the other.

And again, expected performance is objective but it's also comparative. I've mentioned it before, but expected performance is determined by how one game stacks up to every other game, and what engines they use and what the visual outcomes are relative to performance. A game rendering a simple skybox and nothing else, but getting only maximum 45fps would probably be considered objectively bad if it's using standard techniques. But a game simulating Rayleigh scattering in real time with volumetric clouds etc etc running at 45fps would be considered pretty good.

Ergo, Odyssey running at a max of 60fps in settlements despite the game not really employing any unique or advanced rendering techniques, and overall visually being pretty unremarkable compared to other relatively new games, means the game performance is objectively not good, even if 60fps max is "fine" for you.
Seems reasonable.
 
To be fair, you are kind of an edge case. I would need to check but I think I get ~200 FPS on pretty recent hardware in the Odyssey hangar (which is your benchmark IIRC). Obviously I don't get 1400 FPS in Horizons, so I guess the comparison is slightly flawed.

No excuse for the bad performance on your end, just pointing out that (generally) Odyssey doesn't perform 7 times worse my experience.
Oh, yeah, for sure. I was just saying how they perform in comparison for me. It very likely is more of an edge case, or I'd assume we'd be hearing a lot more about it from people much louder than me.
 
Last edited:
I think the takeaway from these recent debates is that "expected performance" is fundamentally different than "acceptable performance." The former is objective, the latter subjective.

When people say performance is still bad even after Update 8, they are probably talking about expected performance. When someone says "its good enough for me now" is talking subjectively. What's more, you can hold both of those opinions at the same time; expected performance can still be "bad" even if you consider the current performance as "acceptable" for yourself. You dont HAVE to explicitly take one side or the other.

And again, expected performance is objective but it's also comparative. I've mentioned it before, but expected performance is determined by how one game stacks up to every other game, and what engines they use and what the visual outcomes are relative to performance. A game rendering a simple skybox and nothing else, but getting only maximum 45fps would probably be considered objectively bad if it's using standard techniques. But a game simulating Rayleigh scattering in real time with volumetric clouds etc etc running at 45fps would be considered pretty good.

Ergo, Odyssey running at a max of 60fps in settlements despite the game not really employing any unique or advanced rendering techniques, and overall visually being pretty unremarkable compared to other relatively new games, means the game performance is objectively not good, even if 60fps max is "fine" for you.
I totally agree with that and I really like the words you chose.

So when some people say that the game works ok when it reaches acceptable performance for them on their machine, i disagree. The game will work ok when it reaches the expected performance on any machine FDev targets.
 
A bit off topic with the conversation, no ?
It seems that we mostly agree that the Odyssey should work better, not that it would be better if it could.
No one has mentioned the GPU load. And without more information, we can't even say if 60fps with 98% GPU load is ok or not.
We are talking about the logic of correlating personal feelings and configuration to judge the overall performance of Odyssey (the fps is ok for me on my config so the game work ok. Or sometime "work ok").
Look, I don't think there is a single person out there that thinks the performance is where it should be. I certainly believe that most people think it should be much better then what it is.

But that's not to say that they can have a good time with the game and enjoy it, because the frames per second is good enough for them to play. But this is all subjective on what is acceptable to make the game enjoyable.

I'm having a great time when I play regardless of the FPS issues. So yes, for me it works okay. It's not perfect and could and should be better for me, but it's good enough for me to play and enjoy.

As to rolling back, that's a big no no. From my understanding, the planetary tech was designed so it could have plant life everywhere, so we now have heat maps instead of pinpoint locations with limited amount of POIs. Rolling back would negatively hit many aspects of the odyssey as it would have a major knock-on effect of other gameplay processes. It's not going to happen.

What w need to do is wait for them to continue to make it better, and hopefully fix the repeating tiling issue.
 
I totally agree with that and I really like the words you chose.

So when some people say that the game works ok when it reaches acceptable performance for them on their machine, i disagree. The game will work ok when it reaches the expected performance on any machine FDev targets.
There's that pin again.
 
Compare the Halloween skeleton suits in-game to the store and it feels like something is a bit off.

I know exactly what you mean.

jason-mamoa-meme-1.jpg
 
Last edited:
Back
Top Bottom