Eurofighter Typhoon
-
Manos cheer up mate
Im working also on my own plane and pit so i endorse you to keep it up
You are the few independent modelers around so we are a rare especies if you know what i mean.
I meant to talk to you long ago but time slips way
We will keep in touch.
BTW - Since the 2GB problem as been solved for 4.34 do you plan to make a more detailed Poly model? Or do we still keep with the 30K limit? Seems a little outdated this thought?
I have my airplane in the 100k polys by now and he cockpit around 130k polys soi don’t expect any sympathy around here
-
Since the 2GB problem as been solved for 4.34 do you plan to make a more detailed Poly model? Or do we still keep with the 30K limit? Seems a little outdated this thought?
AFAIK following the general rules for other products we will implement to some theaters, the basic idea is keep a Lod 1-2-3-4 format as per bms rules for lod1 size/polys, then provide an extra lod0 that can replace the lod1 at any time with much more details and polys.
I have my airplane in the 100k polys by now and he cockpit around 130k polys soi don’t expect any sympathy around here
I’ve been testing JH models in the same 100k area counts with no notifiable issues in my setup. Also managed to gather some 3 million polys in a scene with no problem…
-
… if Manos manage to make models fitting with BMS requirement, we would certainly be happy to integrate his model in stock install.
Nothing prevent him to also create side a with higher poly-count to replaced LOD0 as an external mod …
-
… if Manos manage to make models fitting with BMS requirement, we would certainly be happy to integrate his model in stock install.
Nothing prevent him to also create side a with higher poly-count to replaced LOD0 as an external mod …
The one im finishing up now is about 56-57k tris. If the team wants an 35k i have it all ready in my hdd [emoji6]
Στάλθηκε από το SM-J500F μου χρησιμοποιώντας Tapatalk
-
AFAIK following the general rules for other products we will implement to some theaters, the basic idea is keep a Lod 1-2-3-4 format as per bms rules for lod1 size/polys, then provide an extra lod0 that can replace the lod1 at any time with much more details and polys.
I’ve been testing JH models in the same 100k area counts with no notifiable issues in my setup. Also managed to gather some 3 million polys in a scene with no problem…
That’s one question that keeps buzzing me… why do we still have to have separate lods for the cockpit 1 for the main frame and another for the gauges + 1 for the night lights etc etc? Can’t we have only one complete Lod for all. I have tested in my rig and work just fine with no noticeable bugs or FPS drops.
About the polys so did i and my PC is old as Shxxx!!! We have to start evolving somehow to the Low/medium poly cont to a new era i guess?
And in the next 4/5 years BMS will be a completely new game. We know the path to take
Anyway thanks Raptor
-
The one im finishing up now is about 56-57k tris. If the team wants an 35k i have it all ready in my hdd [emoji6]
Στάλθηκε από το SM-J500F μου χρησιμοποιώντας Tapatalk
Hello,
When it deals with triangles, here is the new policy :
- We shall to keep 30,000 triangles for most cases.
- However, we may increase this limit to 50,000, as an exemption, for models that require more.
In your case, We can’t accept 56k-57k. Sukhoi Su-33 was the aircraft that was cleared with the most triangles, with 47k.
I would suggest you to reduce as much as you can your tri-count.
Nevertheless, great model !
Could you check the canopy ? I may be wrong, but it looks too long to my eyes… What do you think ?
Your canopy looks longer…
Regards,
Radium
-
Hello,
When it deals with triangles, here is the new policy :
In your case, We can’t accept 56k-57k. Sukhoi Su-33 was the aircraft that was cleared with the most triangles, with 47k.
I would suggest you to reduce as much as you can your tri-count.
Nevertheless, great model !
Could you check the canopy ? I may be wrong, but it looks too long to my eyes… What do you think ?
Your canopy looks longer…
Regards,
Radium
He already said that he as as 35k for the team so…
-
He already said that he as as 35k for the team so…
I don’t understand your post as I can read fine.
Manos said :
If the team wants an 35k i have it all ready in my hdd
I just explain why, we can’t accept :
The one im finishing up now is about 56-57k tris.
Cheers,
Radium
-
I don’t understand your post as I can read fine.
Manos said :
I just explain why, we can’t accept :
Cheers,
Radium
I just meant that he understands that his baseline model is not accepted so he as a 35K poly for the team. He understood perfectly before you posted
Jezz keep cool man
-
The one im finishing up now is about 56-57k tris. If the team wants an 35k i have it all ready in my hdd [emoji6]
Στάλθηκε από το SM-J500F μου χρησιμοποιώντας Tapatalk
Cool!
-
Cool!
Main differences are less detailed afterburner pedals and the various scoops and aux intakes are on the skin and not 3d modelled [emoji6]
About the canopy Radium said i will see when back the differences im not home till tuesday!
By the way Happy Easter to all orthodox christians to this forum and the team. Celebrations still go on [emoji16]Στάλθηκε από το SM-J500F μου χρησιμοποιώντας Tapatalk
-
-
I found an other good pic about EF.
-
That’s one question that keeps buzzing me… why do we still have to have separate lods for the cockpit 1 for the main frame and another for the gauges + 1 for the night lights etc etc? Can’t we have only one complete Lod for all. I have tested in my rig and work just fine with no noticeable bugs or FPS drops.
3dckpit.dat for example
// the LOD models to use.
cockpitmodel 941;
cockpitmodel2 947;
cockpithudmodel 4;
cockpitcanopymodel 1037;
cockpitrttcanopymodel 6;
cockpitwingsmodel 1772;
cockpitmodel_empty 1802;
cockpitcanopymodel_empty 1803;
cockpitlegsmodel 2015;My logic states that having the cockpit splitted to smaller parts, as these are mentioned above, will make it much more easier and efficient to work on the individual parts and help populate the different cockpit variants using less model changes.
-
3dckpit.dat for example
// the LOD models to use.
cockpitmodel 941;
cockpitmodel2 947;
cockpithudmodel 4;
cockpitcanopymodel 1037;
cockpitrttcanopymodel 6;
cockpitwingsmodel 1772;
cockpitmodel_empty 1802;
cockpitcanopymodel_empty 1803;
cockpitlegsmodel 2015;My logic states that having the cockpit splitted to smaller parts, as these are mentioned above, will make it much more easier and efficient to work on the individual parts and help populate the different cockpit variants using less model changes.
Sure you are 100% correct and it makes perfect sense but…what if i’m only making one of a kind plane and one of a kind Pit?
-
Back Home. Checked the canopy. It doesnt seem to me longer but i believe needs a bit more Bubble shape i think. And maybe just a bit shorter. I will see what i can do But in general i think only the shape need fixing i hope with some Editpoly will fix it and i wont need to redo it
-
Cool!
Dee-Jay i was thinking regarding the F-4’s in the db.
If i remember well i had a converstation with Radium for them ,and also with Switch before he left the team, for the same reason (and that was the main reason some of the models came only with Lod0 to lod2)
E/F, ESK, EJ and of couse G models (maybe RF-4C as well) can have between them either lod3 and lod4 the same as in big distance the differences are too small to see i believe. But its my humble opinion and anyway beside the lights (witch my original lower lods missing) theres no dofs. DSK and M have their dedicated lower lods.
In general if a plane is like 500 miles away from you its difficult to say if it is an E,F or ESK or EJ you only see the main outlines witch to my eye at least looks the same you cannot even tell if it is a “hardwing” or “softwing” Phantom and also using common lods saves some space in the DB (also one of the reasons i made E and F to use the same model).
Sorry for the off-topic just wanted to clarify this
Cheers -
LOD can have the same shapes … but as long as there are textures on it … it must fit with LOD0 sharing the same Parent.dat file og the Parent folder , otherwise 3DDB Test will generate errors reports.
As long as 3DDB Test do not report for critical errors (labeled as *** in the .txt file) and along as distance transition makes the differences non visible (mainly colors, global shapes) , we are fine. of course, if transition occurs so far that we could even use a generic LOD, the related LOD is simply pointless. So it must remains relevant. No point in making a dedicated lower LOD if the transition occurs beyond visual range.
In general if a plane is like 500 miles away
Aat 500 miles it is even no longer rendered. LOD starts to be rendered at Deagg distance or the lower LOD distance (the lowest value of the two).
-
Back to topic.
Edited a bit the bubbleness (lol) of the canopy it seems a bit better now, might be a bit longer as Radium said but have to see how i can fix this without destroying completely the Uvmapping of the model at least to my eyes now looks a bit better
Now that i see a bit better the whole model i think is a bit longer…. Have to check it maybe a global scaling in x axes fix the issue, got to get some info on that
-
LOD can have the same shapes … but as long as there are textures on it … it must fit with LOD0 sharing the same Parent.dat file og the Parent folder , otherwise 3DDB Test will generate errors reports.
As long as 3DDB Test do not report for critical errors (labeled as *** in the .txt file) and along as distance transition makes the differences non visible (mainly colors, global shapes) , we are fine. of course, if transition occurs so far that we could even use a generic LOD, the related LOD is simply pointless. So it must remains relevant. No point in making a dedicated lower LOD if the transition occurs beyond visual range.
Aat 500 miles it is even no longer rendered. LOD starts to be rendered at Deagg distance or the lower LOD distance (the lowest value of the two).
So in general speaking if i just use the lod3-lod4 of lets say E/F model with the skins of the other parents (lets say EJ) then problem solved right?