Developer Diary 300 - Discussion

[Jaws2002]

2 hours ago, Avimimus said:

Interesting!

7 July 1943
The convoy made up of the cargo-ships Ardeal and Varna and escorted by NMS Marasesti, NMS Marasti, NMS Stihi, NMS Dumitrescu and an R-boot. Two other R-boots and 6 MFPs later joined the convoy. During the night it was attacked by a submarine (Sc-201), which NMS Marasesti manage to damage severely. In the morning, at about 0600 hours, the convoy was attacked by four DB-3Fs. NMS Marasesti opened fire and shot down one of them. The sailors claimed that they saw the fighters shoot down two Soviet aircraft.

Here is what the two fighter pilots claimed:
Adj. av. Mihai Mihordea 1xBristol* at 0650 hours
Adj. av. Vasile Burcu 1xBristol at 0653 hours
Adj. av. Mihai Mihordea 1xBristol at 0656 hours

* The Romanian fighter pilots used to wrongly identify the DB-3F with the Bristol Beaufort.

1 August 1943
The convoy made up of the cargo-ships PLM 16, Prodromos, Kassa and Serose and escorted by NMS Marasti, NMS Murgescu, NMS Ghiculescu, NMS Dumitrescu and 2 R-boots was attacked at 0937 hours in the vicinity of Yevpatoria by two Soviet torpedo-bombers, coming from the south. The fire from the NMS Marasti and NMS Murgescu (probably the Romanian ship with the most aircraft shot-down) made them launch from far away (some 3,000m) and the torpedoes missed the destroyer. The airplanes machine-gunned the Xanten submarine-hunter, which had some casualties on board. Two IAR-80Cs of the 49th Squadron attacked the DB-3Fs and severely damaged them. The pilots did not see them crash, but they disappeared from German radar and the sailors claimed that the aircraft were shot down, so they were later confirmed.

Lt. av. Gheorghe Butnaru 1xBristol flying the IAR-80C no. 271
Serg. TR. av. Radu Costache 1xBristol flying the IAR-80C no. 241

6 September 1943
The convoy made up of the cargo-ships Burgas and KT 25 and escorted by NMS Regina Maria, NMS Regele Ferdinand, NMS Ghiculescu and 3 R-boots was attacked near Sevastopol by five Soviet bombers at 1220 hours and at 1251 by another four. The fighters intervened and claimed two kills, of which one was confirmed.

Adj. av. Stefan Dumitrescu 1xBoston III* flying the IAR-80C no. 271
Serg. TR. av. Radu Costache 1xBoston III probable flying the IAR-80C no. 274

* Probably also DB-3Fs

27 December 1943
Adj. av. Mircea Mazilu 1xBoston III at 0730 hours

30 January 1944
Adj. av. Mircea Simion 1xBoston III 3.5 km NE Yevpatoria at 1510 hours

3 February 1944
Lt. av. Gheorghe Butnaru 1xBoston III N of Ak. Mafelka
I have also found a mention of a German tug convoy being attacked near Odessa on 5 February and four Soviet aircraft being lost. Maybe this claim is actually for 5 February.

16 March 1944
Adj. av. Mihai Mihordea 1xBoston III at 1105 hours
Adj. av. Mihai Mihordea 1xBoston III at 1108 hours
Adj. av. Eugen Toflan 1xIl–2 at 1112 hours

[Avimimus]

Well... now I just want a DB-3F and IAR-80C XD

[Boomerang]

Happy 20th Anniversary IL-2 Sturmovik! 

Congratulations on the 300th DD too!.  It's great to see IL-2 continuing to shine as bright as the true star it is! 

 

[Gambit21]

17 hours ago, Soilworker said:

 

I reckon it'll be bump mapping, look at the leather on the seats, no way they did that with polys. ?

 

 

Normal map, but same concept.

[LukeFF]

4 hours ago, Jaws2002 said:

7 July 1943
The convoy made up of the cargo-ships Ardeal and Varna and escorted by NMS Marasesti, NMS Marasti, NMS Stihi, NMS Dumitrescu and an R-boot. Two other R-boots and 6 MFPs later joined the convoy. During the night it was attacked by a submarine (Sc-201), which NMS Marasesti manage to damage severely. In the morning, at about 0600 hours, the convoy was attacked by four DB-3Fs. NMS Marasesti opened fire and shot down one of them. The sailors claimed that they saw the fighters shoot down two Soviet aircraft.

Here is what the two fighter pilots claimed:
Adj. av. Mihai Mihordea 1xBristol* at 0650 hours
Adj. av. Vasile Burcu 1xBristol at 0653 hours
Adj. av. Mihai Mihordea 1xBristol at 0656 hours

* The Romanian fighter pilots used to wrongly identify the DB-3F with the Bristol Beaufort.

1 August 1943
The convoy made up of the cargo-ships PLM 16, Prodromos, Kassa and Serose and escorted by NMS Marasti, NMS Murgescu, NMS Ghiculescu, NMS Dumitrescu and 2 R-boots was attacked at 0937 hours in the vicinity of Yevpatoria by two Soviet torpedo-bombers, coming from the south. The fire from the NMS Marasti and NMS Murgescu (probably the Romanian ship with the most aircraft shot-down) made them launch from far away (some 3,000m) and the torpedoes missed the destroyer. The airplanes machine-gunned the Xanten submarine-hunter, which had some casualties on board. Two IAR-80Cs of the 49th Squadron attacked the DB-3Fs and severely damaged them. The pilots did not see them crash, but they disappeared from German radar and the sailors claimed that the aircraft were shot down, so they were later confirmed.

Lt. av. Gheorghe Butnaru 1xBristol flying the IAR-80C no. 271
Serg. TR. av. Radu Costache 1xBristol flying the IAR-80C no. 241

6 September 1943
The convoy made up of the cargo-ships Burgas and KT 25 and escorted by NMS Regina Maria, NMS Regele Ferdinand, NMS Ghiculescu and 3 R-boots was attacked near Sevastopol by five Soviet bombers at 1220 hours and at 1251 by another four. The fighters intervened and claimed two kills, of which one was confirmed.

Adj. av. Stefan Dumitrescu 1xBoston III* flying the IAR-80C no. 271
Serg. TR. av. Radu Costache 1xBoston III probable flying the IAR-80C no. 274

* Probably also DB-3Fs

27 December 1943
Adj. av. Mircea Mazilu 1xBoston III at 0730 hours

30 January 1944
Adj. av. Mircea Simion 1xBoston III 3.5 km NE Yevpatoria at 1510 hours

3 February 1944
Lt. av. Gheorghe Butnaru 1xBoston III N of Ak. Mafelka
I have also found a mention of a German tug convoy being attacked near Odessa on 5 February and four Soviet aircraft being lost. Maybe this claim is actually for 5 February.

16 March 1944
Adj. av. Mihai Mihordea 1xBoston III at 1105 hours
Adj. av. Mihai Mihordea 1xBoston III at 1108 hours
Adj. av. Eugen Toflan 1xIl–2 at 1112 hours

 

To go back in time just a bit, Volume 4 of Black Cross/Red Star talks about both IAR 80s and 81s at Stalingrad in the fall of 1942 (Grupul 8 and 6, respectively). 

[Hanu]

On 11/22/2021 at 6:37 PM, LukeFF said:

Looking awesome! Getting to drop troops over Normandy at night with flak exploding all around is going to be a lot of fun:

 

 

Too true!

Fun fact: after the war Finns bought 19 DC-3/C-47/C-53 for FAF and Aero (national airline company). Several planed had had history in Normandy, but one partitcular place, SN: 42-100646, Register number DO-7 during Air Force and OH-LCB during Aero had a history of transporting paratroopers of 101th Airborne, 506th PIR, 2nd platoon, Easy company. This plane's jumpmaster was none else than Richard "Dick" Winters.  ???

[Livai]

7 hours ago, Gambit21 said:

Normal map, but same concept.

 

Normal Maps do not contain any height information whatsoever. and leather on the seats is not flat........

[AEthelraedUnraed]

1 hour ago, Livai said:

Normal Maps do not contain any height information whatsoever. and leather on the seats is not flat........

Actually, Normal Maps are a more advanced and more modern technique than bump mapping to achieve the same effect. See https://www.cgdirector.com/normal-vs-displacement-vs-bump-maps/.

Edited November 24, 2021 by AEthelraedUnraed

[Soilworker]

9 hours ago, Gambit21 said:

 

Normal map, but same concept.

 

Ah yes, of course. ?

[Livai]

1 hour ago, AEthelraedUnraed said:

Actually, Normal Maps are a more advanced and more modern technique than bump mapping to achieve the same effect. See https://www.cgdirector.com/normal-vs-displacement-vs-bump-maps/.

 

Bump mapping and Normal Maps are the same but with one big difference Bump Mapping use "RGB" coordinate where Normal Maps use "XYZ" coordinate. This makes Normal maps better than Bump Mapping. What Normal Maps not have are height information. For Height information you can use displacement map or use the Alpha Channel from Normal Maps for Height informations ( Normal Maps in this Game are without Alpha Channel is the Alpha Channel white means no information inside ).  I was more talking about Normal Map + Displacement Map

 

 

Edited November 24, 2021 by Livai

[Soilworker]

2 hours ago, Livai said:

 

Normal Maps do not contain any height information whatsoever. and leather on the seats is not flat........

 

Actually if you go back and look at the leather on the seats the edges are pretty straight, I'm pretty sure it is normal mapping.

 

I can't really see why they would go with displacement for something as shallow as the texture of leather considering its higher performance cost, especially on seats which is something pretty much out of your normal line of sight. 

[ScotsmanFlyingscotsman]

Congratulations on the milestone, been here since the early version and will support this too in the future, keep up the good work,

Retirement has arrived for me, so I got the Bravo Throttle system recently, this should be brilliant with it. Thanks Devs 

Edited November 24, 2021 by scotsmanFlyingscotsman

[AEthelraedUnraed]

1 hour ago, Livai said:

Bump mapping and Normal Maps are the same but with one big difference Bump Mapping use "RGB" coordinate where Normal Maps use "XYZ" coordinate. This makes Normal maps better than Bump Mapping. What Normal Maps not have are height information. For Height information you can use displacement map or use the Alpha Channel from Normal Maps for Height informations ( Normal Maps in this Game are without Alpha Channel is the Alpha Channel white means no information inside ).  I was more talking about Normal Map + Displacement Map

Sorry, but your post doesn't make sense. First, RGB and XYZ are exactly the same thing, it's just the name you give to the channels. I could name them ABC if I wanted to, and it would still be the same thing.

 

Secondly, it's also not true that Bump Mapping uses three channels. It uses one, namely the height. That's precisely why Normal Mapping is more accurate than Bump Mapping: you've got three times the information density, so you're able to accurately depict the slope of an object, which is impossible with Bump Mapping.

 

Lastly, it's not true that the leather on the seats is not flat (which you claimed in your previous post). You can clearly see in the posted cockpit images that the leather is in fact flat. Especially in this image:

You can clearly see from the straight edge in the top left corner that the leather doesn't have any "real" additional depth, but rather the depth effect is implied by Normal Mapping. There are techniques that do simulate depth, such as Parallax Occlusion Mapping, but these are much more computationally expensive than Normal Mapping (which is rather straightforward from a shading perspective) so it wouldn't make sense to use any of these techniques in cases were normal mapping suffices (such as here; the leather actually looks pretty good IMO even without "real" depth).

[Missionbug]

However the C-47 interior effect is done it certainly is impressive I have to say.

 

My own thoughts are that it might be sculpted, there are 3d programs around that essentially allow the artist to do just that, they replicate working with clay and as such it could be that the quilt effect is done in the same way that artists use clay or we as children might use plasticene modelling material flattened out and then use some form of pointed object to scribe the detail in.

 

The 3d program I use is very old, however, many of the new generation allow the import of materials created in one to be integrated into others to allow the artist to create a scene such as what we see in sim work.

 

Hopefully those in the know can put us out our misery and explain just how the C-47 interior was done.

 

Take care and be safe.

 

Wishing you all the very best, Pete.

[ShamrockOneFive]

Looks like Ugra Media are pulling out all of the stops to give us a nice looking C-47 interior to me!

[Gambit21]

5 hours ago, Livai said:

 

Normal Maps do not contain any height information whatsoever. and leather on the seats is not flat........

 

No.

What matters and what we’re talking about is the render, not creating actual geometry (displacement) which is too costly for game engines.

 

Thus normal maps which are baked from a texture map or procedural material. Bump maps are not typical for this application.

 

 

Edited November 24, 2021 by Gambit21

[Livai]

2 hours ago, Soilworker said:

 

Actually if you go back and look at the leather on the seats the edges are pretty straight, I'm pretty sure it is normal mapping.

 

I can't really see why they would go with displacement for something as shallow as the texture of leather considering its higher performance cost, especially on seats which is something pretty much out of your normal line of sight. 

 

What looks for you more like Leather? Not the best example because we talk about seats but still it is Leather. I prefer the last one just because I have much more details.

 

-> BTW The look of the leather is largely defined by bump map, normal map and displacement map. Thin leather has more wrinkles and sharper folds, while thick leather has larger and softer details.

 

 

1 hour ago, AEthelraedUnraed said:

Sorry, but your post doesn't make sense. First, RGB and XYZ are exactly the same thing, it's just the name you give to the channels. I could name them ABC if I wanted to, and it would still be the same thing.

 

Secondly, it's also not true that Bump Mapping uses three channels. It uses one, namely the height. That's precisely why Normal Mapping is more accurate than Bump Mapping: you've got three times the information density, so you're able to accurately depict the slope of an object, which is impossible with Bump Mapping.

 

Lastly, it's not true that the leather on the seats is not flat (which you claimed in your previous post). You can clearly see in the posted cockpit images that the leather is in fact flat. Especially in this image:

You can clearly see from the straight edge in the top left corner that the leather doesn't have any "real" additional depth, but rather the depth effect is implied by Normal Mapping. There are techniques that do simulate depth, such as Parallax Occlusion Mapping, but these are much more computationally expensive than Normal Mapping (which is rather straightforward from a shading perspective) so it wouldn't make sense to use any of these techniques in cases were normal mapping suffices (such as here; the leather actually looks pretty good IMO even without "real" depth).

 

You can add Bump map to Normal Map even different Bump maps another Normal Map or Bump map and yes it improves something but still some materials are largely defined by bump map, normal map and the displacement map.

 

RGB and XYZ are not the same thing to some degree yes. You need XYZ how & where light and shadows appear.

 

-> As long the pilot not sits on the leather seat - missing Height information - was some kind of joke

 

For example about techniques in (Total War Shogun 2) what I modded where adding Height information into the Alpha Channel from Normal Maps did the same thing and looked the same as using their Displacement Shader. This showed me to archive the same result and quality there is a way, just easier.....................

[Jason_Williams]

Guys please. This discussion about bumps and normals is a waste of energy and a bit off-topic.

 

Jason

[AEthelraedUnraed]

59 minutes ago, Jason_Williams said:

Guys please. This discussion about bumps and normals is a waste of energy and a bit off-topic.

You're probably right, and I think everything that needed to be said on the topic of Normal vs. Bump mapping has been said by now. I will add one last reply to Missionbug's post however, as it specifically asks about the C-47 as shown in the DD and also touches upon the poly budget (since you specifically talk about the poly budget in the DD, IMHO it is on topic to explain what this is and how Ugra Media tries to adhere to it).

 

4 hours ago, Missionbug said:

My own thoughts are that it might be sculpted, there are 3d programs around that essentially allow the artist to do just that, they replicate working with clay and as such it could be that the quilt effect is done in the same way that artists use clay or we as children might use plasticene modelling material flattened out and then use some form of pointed object to scribe the detail in.

 

The 3d program I use is very old, however, many of the new generation allow the import of materials created in one to be integrated into others to allow the artist to create a scene such as what we see in sim work.

 

Hopefully those in the know can put us out our misery and explain just how the C-47 interior was done.

I won't call myself an artist, but I do know a thing or two about 3d modelling and I'm pretty well versed in Blender, so I think I'm qualified to give an educated answer about how the C-47 interior was done. You're right about modern 3d modelling programs being able to add small details to the geometry where required. However, this quickly increases the amount of polygons in the mesh. Each poly takes a bit of GPU power to render, so in games such as this, they usually have a pretty tight poly budget to keep render times low (as Jason mentions in this DD). Therefore, they leave out all small details (such as the leather in this case) and use normal maps (or very rarely bump maps) to simulate these additional details. This can save literally thousands of polygons while it doesn't reduce the visual quality too much. If you know what to look for, you can see that the C-47 cockpit too uses a relatively low-poly model with normal mapping (don't get me wrong - low-poly here still means many thousands). This is how Ugra Media creates this beautifully detailed cockpit while still adhering to IL2's poly budget. (The only exception to this general rule are large animation studios like Pixar that can afford to spend hours rendering a single frame just to get that extra bit of detail.)

 

These normal maps are either created for a specific texture if this texture is often repeated, as is the case here, or created from a high-poly version of the model (that does contain all small details in 3d) that is then projected onto the lower-poly model.

[Gambit21]

1 hour ago, AEthelraedUnraed said:

You're probably right, and I think everything that needed to be said on the topic of Normal vs. Bump mapping has been said by now. I will add one last reply to Missionbug's post however, as it specifically asks about the C-47 as shown in the DD and also touches upon the poly budget (since you specifically talk about the poly budget in the DD, IMHO it is on topic to explain what this is and how Ugra Media tries to adhere to it).

 

I won't call myself an artist, but I do know a thing or two about 3d modelling and I'm pretty well versed in Blender, so I think I'm qualified to give an educated answer about how the C-47 interior was done. You're right about modern 3d modelling programs being able to add small details to the geometry where required. However, this quickly increases the amount of polygons in the mesh. Each poly takes a bit of GPU power to render, so in games such as this, they usually have a pretty tight poly budget to keep render times low (as Jason mentions in this DD). Therefore, they leave out all small details (such as the leather in this case) and use normal maps (or very rarely bump maps) to simulate these additional details. This can save literally thousands of polygons while it doesn't reduce the visual quality too much. If you know what to look for, you can see that the C-47 cockpit too uses a relatively low-poly model with normal mapping (don't get me wrong - low-poly here still means many thousands). This is how Ugra Media creates this beautifully detailed cockpit while still adhering to IL2's poly budget. (The only exception to this general rule are large animation studios like Pixar that can afford to spend hours rendering a single frame just to get that extra bit of detail.)

 

These normal maps are either created for a specific texture if this texture is often repeated, as is the case here, or created from a high-poly version of the model (that does contain all small details in 3d) that is then projected onto the lower-poly model.

 

That’s correct.

Speaking of the beautiful C-47 in this development diary - the object can be sculpted (natively in most 3D apps now) and then the results exported as a Normal map. Underlying polygons remain the same. Whatever was done regarding the C-47 that we’re discussing here  - it’s coming along nicely and great work!

 

Thanks for sharing Jason.

Gears are turning.

 

Lots of potential here.

 

 

 

 

Edited November 24, 2021 by Gambit21

[BlitzPig_EL]

I wonder if a C47 can do a loop, or other ACM?

 

HMMMMM....

 

?

 

 

 

[Gambit21]

9 minutes ago, BlitzPig_EL said:

I wonder if a C47 can do a loop, or other ACM?

 

HMMMMM....

 

?

 

 

 

 

I don’t know - but damn sure can drop supplies near the front lines and ferry around a few VIP’s.

 

[BlitzPig_EL]

I plan to check out the edges of her performance envelope. Be sure!

 

 

[Avimimus]

2 hours ago, BlitzPig_EL said:

I wonder if a C47 can do a loop, or other ACM?

 

HMMMMM....

 

?

 

In the original Il-2... back when people used a modified mission to fly AI aircraft in external views (or maybe shortly after someone made a flyable mod)... I recall getting it into a negative attitude stall... probably a result of the simplistic flight model though! It'll be interesting to see what it is like in this sim. I recall stories of a C-47 flying aggressively down canyons in Burma, and doing low speed turns to escape a Japanese fighter...

[Dakpilot]

http://www.dc-3.co.za/dc-3-stories/277.html

 

They can certainly do ground loops

 

Cheers, Dakpilot 

[Dakpilot]

Not a DC-3 but a DC-4, but it is surprising to many how agile a transport aircraft can be when empty of cargo. Of note is no. 3 and 4 engines feathered on the flyby, not an "actual" loop more of an extreme wingover but you get the idea

 

Watch to the end

 

Cheers, Dakpilot 

[ITAF_Rani]

C 47 will be also in russian version with guns turret ?

[ROCKET_KNUT]

15 minutes ago, ITAF_Rani said:

C 47 will be also in russian version with guns turret ?

It will be a seperate purchase though if I remember correctly. The engines on the Li-2 are different and so might be some of the systems, cockpit gauges and so on. ?

[jeanba]

2 hours ago, ROCKET_KNUT said:

It will be a seperate purchase though if I remember correctly. The engines on the Li-2 are different and so might be some of the systems, cockpit gauges and so on. ?

Will we have the  "bomber version" of the Li2 ?

I think it was some kind of a "field mod"

Edited November 25, 2021 by jeanba

[kendo]

5 hours ago, Dakpilot said:

Not a DC-3 but a DC-4, but it is surprising to many how agile a transport aircraft can be when empty of cargo. Of note is no. 3 and 4 engines feathered on the flyby, not an "actual" loop more of an extreme wingover but you get the idea

 

...

 

Cheers, Dakpilot 

 

?  Scary.....!

[BlitzPig_EL]

I talked to a C17 pilot at an airshow and he said that when empty it could pull some surprising ACM "tricks".  Makes sense, what with a high lift wing, and an excess of thrust when empty.

[Avimimus]

9 hours ago, jeanba said:

Will we have the  "bomber version" of the Li2 ?

I think it was some kind of a "field mod"

 

Yes... it will be a separate airplane... made in the Soviet Union with different engines and systems. Hopefully we'll get some field mods for it (e.g. 8xFAB100 loadout, bombs dropped from cassettes out of the side door, turret options, fixed forward firing gun as an option) but it will be a separate airplane:

 

[LeftyHC]

Hi!

 

Not sure if someone already asked, but i'd like to know about the status of the new clouds. What's the situation on that?

 

Thank you.

[ScotsmanFlyingscotsman]

C-47 Loop, depends how much height you have, how empty your are, and what's shooting at you, If the wings are still on at the end of the loop,....that's a yes. (Let us know how it goes...and if we don't hear back from you......sorry) ?

 

Edited November 26, 2021 by scotsmanFlyingscotsman

[Lensman]

I've been with you since the beginning and still using the same MS Force Feedback Sidewinder 2 (thank you Bill!).

Congratulations for your extraordinary achievements.

Best wishes from England.

Phil

[Avimimus]

You know what we won't have? The version of the Lisunov Li-2 with 2xNS-45 45mm automatic cannons firing forward. At least I'm pretty sure there aren't enough references.

 

I do wonder if one could've looped it though.

[Dakpilot]

C-130 doing an actual loop 

 

Cheers, Dakpilot 

 

[Beebop]

On 11/26/2021 at 12:59 AM, LeftyHC said:

Hi!

 

Not sure if someone already asked, but i'd like to know about the status of the new clouds. What's the situation on that?

 

Thank you.

Last I looked they were still in the sky. 

[Vortice]

On 11/26/2021 at 8:59 AM, LeftyHC said:

Hi!

 

Not sure if someone already asked, but i'd like to know about the status of the new clouds. What's the situation on that?

 

Thank you.

 

Fluffy?

[Noisemaker]

1 hour ago, Vortice said:

 

Fluffy?

Nebulous...