TESPort-Discussion

[bob196045]

Thank you vm


Enjoy
Bob

[The Grey Wizard]

After what I have read, I have come to the conclusion that I should make the map from scratch. I only have 512 mb of ram and I never have the height map editor crash on me, unless of course my ESP just hit above 20megs, which is normal. Then I convert to Master and it runs fine. If you have 2.5 gigs, it is simply impossible, for this simple reason. The landmass makes an ESP more then 20megs all at once. With this, the esp cannot compute anymore info and must be a master to withold the extensive amount of land data. It's not a matter of how much ram you have, but simply that it cannot be done with an OB esp. If I am wrong Lightwave, do correct me, after all, you program the ESP format, and I just use it.Anyway, I won't have a problem making the height map in a considerable time. I can just make it, but it will need to be smoothed just like Lightwave's maps simply because height mapping gets messy, and nothing beats using the landscape tool in the render window. Each claim can do this individually says I. It's your call all in all KuKu, but I seriously reccomend that you let me make it. No offence to your genius Lightwave, but I feel personally that this is the most rational course of action for moving production of Vvardenfell forward. You being a very rational man youself, I am sure you can sway my mind otherwise should you disagree.

[Lightwave]

Originally posted by The Grey Wizard
With this, the esp cannot compute anymore info and must be a master to withold the extensive amount of land data. It's not a matter of how much ram you have, but simply that it cannot be done with an OB esp. If I am wrong Lightwave, do correct me, after all, you program the ESP format, and I just use it.Anyway, I won't have a problem making the height map in a considerable time. I can just make it, but it will need to be smoothed just like Lightwave's maps simply because height mapping gets messy, and nothing beats using the landscape tool in the render window. Each claim can do this individually says I.

Atm, I'm seeing no technical problems with the 2X Vvardenfell (58Mb) ESP landmass I posted: To generate the LOD without crashing I've realized one just has to load the ESP as inactive - it requires only 303Mb of memory to load this way (Oblivion.esm is not required, but combined they need 476Mb). Loaded as active it requires a little under 2Gb and this is very close to the maximum memory Windows XP (32-bit) will give a single application. Thus is the price of such a large landmass - really any landmass which is 2x2 the surface area of the original Vvardenfell would totter perilously close to the edge of XP's limit when loaded as active. Luck rather than judgement kept my scaled up export just below this! I don't know how Windows2000 and NT4 fair by comparison. Luckily there is no begging need for anyone to load this landmass as active anyway, but you can if set enough virtual memory and are prepared to wait ... a while.

Btw, here's a screenshot of Red Mountain area, a portion of the 2x2 Vvardenfell landmass displayed in the heightmap editor preview window (loaded inactive.)

My PC is 2 years old, 1Gb RAM, 300Gb HD, Athlon64 3000+, but only takes 3 seconds for my PC to load the full landmass in to TES4 CS (as inactive).
There's no 20Mb ESP land limit that I'm aware of - unless that's for those worldspaces ticked as 'Small Worldspaces' maybe; this 2x Vvardenfell landmass is an ESP and certainly much larger. Still, I'm curious where you've heard of such a limit and would like to learn more, but it's not something I've experienced. I've loaded 900Mb ESP landscapes in to MW too (I tried a 4x4 Vvardenfell a few days ago - i.e. 16 times the surface area!)

To load the above landmass as active requires the best part of 2Gb of memory. It takes longer to load on my PC too since 1Gb of that is virtual memory, but once loaded (5 - 10 minutes) it's quite OK to use and is easy to start smoothing problems. Just don't try to generate LOD whilst it's loaded active, or the combined memory requirements of TES and the heightmap editor exceeds XP's hard limits and the whole thing crashes - LOD can only be done when loaded inactive.

AFAIK the only reason to want to load this landmass as active anyway is to smooth out the original landmass so individuals start with a nicer model. But it's by no means necessary: Any individual that smooths out the landscape in their own ESPs will override the master's copy when both are loaded in the CS or in game. Eventually these ESPs will have to be combined with the master and logically they have to over-ride the master landscape records. On that basis assuming each cell will have been textured and modified (if only slightly) by a individuals, we could totally do away with the master landmass ESP/ESM when the whole project is combined since the entire land will already be stored in the combined ESPs.

I've honestly nothing against you producing a new landmass, it would still be interesting to see (and could be fun making it), but I don't think the necessity for it is in any way technical at this stage. At the end of the day either approach from what you say will probably demand a similar amount of work by individuals to quickly smoothen out any oddities in the CS render window.

As an aside I don't think permission would need to be asked to use the landmass I've generated any more. Although the general profile is pretty close to the original, it's no longer even a direct copy to scale due to the quirks I introduced (and the quirks affect most cells, some more than others). Brushing to correct each problem caused by the ribs means almost the entire landscape will end being hand modified at least slightly before the project is completed.

Btw KuKulzA: Is there some kind of testing plan on how to decide which landmass might be chosen as the better suited for the project though (e.g. visiting areas in game, LOD or some kind of map view?)

Lightwave

p.s. I've been dabbling with another side project, a program to import a RAW, BMP, or CSV (commar delimited file) containing heightmap data in to Morrowind (and similarly Oblivion), which would be far more versatile than the TES4 importer 1024x1024 16-bit quad importer, that is once I get some of the little join quirks finished. Technically it should then be able import large RAW or BMP images scaled up in an image editor, without the ribbed quirks and cell join problems my previous method produced.

[Lightwave]

Btw, the landmass all appears to be there, I told TEStroi to turn the OB worldspace in to a 16-bit bitmap, and here's the result (2Mb) - this is a low quality jpeg cut down from the original 60Mb BMP output file (5376x5362).

The main point is that the whole landmass appears to be there, I was a little perturbed when loading then saving the 150Mb TESPort converted ESP in the TES4 CS produced a smaller 58Mb file, but since it's all there I'm guessing Bethesda must have a more powerful ZLIB compression algorithm than the standard one (written by the authors of ZLIB) which TESPort uses, interesting to know (well, for me anyway. )

Lightwave

[KuKulzA]

seeing Lightwave's work, it doesn't look half bad :yes: but I also know Grey Wizard is very skilled...
I feel like I'm abusing you guys' skills in asking both of you to make Vvardenfell and then choose one...
but, that's what I am asking

to judge I'm going to need screenshots, not edited, just as they are...
- height of Red Mt. in the new compared to the old (side to side, 2 pictures in the one image)
- overhead 'topography' type of screenshot showing the height levels
- compare to old TES3 heightmap overhead to show the changes
- basin for Balmora (with the Odai river/groove running through it) to see how it compares
- screenshot from the top of Red Mt. to show LOD



but Grey Wizard, I want to see what you make...
in the end, the smoothing out and fine tuning of the land will have to be done in both cases... and for the two of you I have a feeling both of you guys produce excellent work, so it'll be tough choosing, but I would like to choose, so yes, Grey Wizard, green light, go for it. I hope to see your Vvardenfell soon so i can compare

[The Grey Wizard]

There is still one thing then. We need to run it by Bethesda and see if it's ok. The only reason I feel this way is because if we don't, and Beth asks us not to use it, I would prefer if the whole argument that Vality got to do it not arise. I have alot of interest in Dawn of Oblivion, and I would hate to see Beth come down on both of the mods. So there you have it. Converting somting from OB and downgrading to MW is one thing, but taking assets from MW and putting them in OB is a little more touchy, or so my heart tells me.

[Vality7]

The more I think about the whole converting thing, I have to agree with The Grey Wizard. Although I may have encouraged it before, after much thought, I think that you should go back to doing it as originally planned.

My reason being, is that you would have to ask for permission, just incase they try and do something about it later, and in doing so, you would most likely have to say, "but you let Vality do it", and then they would most likely have to reconsider my mod, which could mean getting my mod shut down.

I don't know exactly why Bethesda agreed to let me go ahead, but I know that it was a one off case. It may sound like a double standard, but up until now, my case was unique.

I wouldn't want to take the chance that my mod could get shut down, especially when The Grey Wizard is prepared to make the heightmap from scratch. The difference is that I am not, as there is no way my project will be able to continue without a premade heightmap, as I don't have the time nor patience to start from scratch.

The only reason my mod ever came to be is because I already had a heightmap premade. To explain how it began simply, I had a heightmap of Cyrodiil, which had been converted to a DEM by Wolfie's Den about 7-8 months ago, that I imported using MGen, and was just gonna release it as-is for people to see what it would look like, but then Shenna came to me with a story, and we decided that we could go from there, and make a mod out of it, as the major work had already been done. To avoid any future complications, I asked Bethesda if it would be okay to go ahead, and they agreed to let me do it.

TES4: Vvardenfell however, was not founded upon the idea of converting anything, as it has always been intended to do everything by hand.

There is of course the possibility that Bethesda would let you go ahead with doing it, but I think that chance is slim, and I have a feeling that it would lead to them questioning why they let me go ahead, and for my mods sake I really don't want that to happen.

I'm not saying my mod is more important than yours, or that Lightwave's hard work has been for nothing, but I feel that I don't really have a choice, as I don't want to risk it.

Of course its not the end of the world if Cyrodiil is cancelled, I can live with that, it would probably be good for me in a way, and would give me a reason to get outside more.

It is of course totally up to you though, as its your mod, but I hope you can understand my reasoning for feeling the way I do.

[Stratguy]

Kinda new here, third post after all, feeling a bit out of place, but I would like to give my opinion:

I think we should run this by Bethesda, without mentioning Vality. If we get a no, The Grey Wizard will make the heightmap as originally planned, without jeopardizing Dawn Of Oblivion... Why should we risk?

Look, big mods like this, have about 1 to 30 chance to be done for various reasons. One of them is time: First of, if this mod takes more then a year to make, we can start losing modders (I can see you already did in the past) due to lost interest in the mod because it is taking too long.

Secondly, if this mod won't be done until the release of TES5, we're gonna lose members AND interest of the community because everyone will flock to the new game. Take a look at Tamriel Rebuilt for Morrowind for an example...

Time is worth it's weight in gold and diamonds for us basicaly, and if we can cut it in any way, we should go ahead, because a readymade height map will save us considerable time, and since we're making it from TES3's original one, quality is assured.

[Vality7]

Originally posted by Stratguy
I think we should run this by Bethesda, without mentioning Vality.

Well that would be ideal..but I get the feeling someone will bring it up if they say no.

I really wish I had kept my mouth shut. :poorme: But when it comes to me and heightma...

[KuKulzA]

how about this...

I will ask Bethesda by PMing a DEV or someone..
I will not mention you Vality7, just say that we have this idea of importing Morrowind heightmap to TES4
and I'll ask if we are allowed to do that...

if they say no, I say, thank you for your time
and we'll HAVE TO go with Grey Wizard's heightmap made from scratch
if they say yes, its ok, then we can compare the two and decide

EDIT:
any suggestions for moderators/DEVs on ESForums to contact?