LDraw.org Discussion Forums

Full Version: Technic 2019
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Technic 2019 models thread.
[Image: 42095.png]
Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6b
Rendered with Studio 2.0
[Image: 42096.jpg]
[Image: 42096_-_porsche_911_rsr_-_back.jpg]
Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6b
Rendered with Studio 2.0
[Image: 42093.jpg]
Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6b
Rendered with Studio 2.0
[Image: 42094.jpg]
Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6b
Rendered with Studio 2.0
MPD updated: winch rope was not properly embedded.
[Image: 42092.jpg]
Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6b
Rendered with Studio 2.0
[Image: 42099.jpg]
Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6c
Rendered with Studio 2.0
The render looks really nice, but the color to me looks like traditional yellow was used instead of the yellow-ish orange of the actual set.
(2019-10-12, 1:45)Travis Cobbs Wrote: [ -> ]The render looks really nice, but the color to me looks like traditional yellow was used instead of the yellow-ish orange of the actual set.

Assuming that "Flame Yellowish Orange" is the correct color (and it certainly sounds correct), I would argue that the RGB value in LDConfig.ldr is wrong. It definitely doesn't match the color on the real model that I see with my eyes, either purely as an RGB color, applied to geometry in LDView (with any of LDView's lighting options), or in your high-quality render. Your render seems to match the color I see in LDView.
(2019-10-12, 2:00)Travis Cobbs Wrote: [ -> ]Assuming that "Flame Yellowish Orange" is the correct color (and it certainly sounds correct), I would argue that the RGB value in LDConfig.ldr is wrong. It definitely doesn't match the color on the real model that I see with my eyes, either purely as an RGB color, applied to geometry in LDView (with any of LDView's lighting options), or in your high-quality render. Your render seems to match the color I see in LDView.

maybe we should try one of the other RGBs listed for 191:
https://wiki.ldraw.org/wiki/User:Owen_Bu...ison_Chart
(2019-10-12, 2:50)Orion Pobursky Wrote: [ -> ]maybe we should try one of the other RGBs listed for 191:
https://wiki.ldraw.org/wiki/User:Owen_Bu...ison_Chart

LDD's color looks the most accurate to me, but since I don't have a calibrated monitor, I guess the problem could be on my end.
(2019-10-12, 4:14)Travis Cobbs Wrote: [ -> ]LDD's color looks the most accurate to me, but since I don't have a calibrated monitor, I guess the problem could be on my end.
I agree that the LDD color looks better, and actually I use the latest beta ldconfig.ldr which uses LDD color data. But I have no control over Studio render color (I guess I could patch the value in the renderer source file, but that's another story...)
[Image: 42110.jpg]

Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6c
Rendered with Studio 2.0

LDCad Universal joint placement helper script was heaven sent for this model!
Awesome !
[Image: 42100.jpg]
[Image: 42100_-_liebherr_r_9800-back.jpg]
Download MPD (OMR compliant)
Known errors: No stickers.
Done with LDCad 1.6c
Rendered with Studio 2.0
Just out of curiosity, I've noticed that by far the most common "error" of any OMR submission is "no stickers." I always create the stickers for any model I build (if they're not already in the library), but it's entirely possible that the stickers I'm creating wouldn't be OMR-compliant if I were to submit them. Is that the main reason people don't bother with the stickers—because there's no officially "acceptable" way to create them? Or is it just that most modelers don't undertake the extra effort?
(2019-11-23, 1:34)N. W. Perry Wrote: [ -> ]Just out of curiosity, I've noticed that by far the most common "error" of any OMR submission is "no stickers." I always create the stickers for any model I build (if they're not already in the library), but it's entirely possible that the stickers I'm creating wouldn't be OMR-compliant if I were to submit them. Is that the main reason people don't bother with the stickers—because there's no officially "acceptable" way to create them? Or is it just that most modelers don't undertake the extra effort?

For me it's 2 reasons:
- Most of my sticker sheets are in storage. I download instructions from Lego so I can alt-Tab between LDCad and my PDF reader but not such luck on stickers.
- There's currently no method for including textures with OMR models. This is because I wrote the OMR spec before we had any TEXMAP parts in the library.
(2019-11-23, 1:34)N. W. Perry Wrote: [ -> ]Or is it just that most modelers don't undertake the extra effort?

Yep! You have also to consider that not all model builders are able to author a part or sticker. In addition such projects are long runners. It took almost a year and half to get the 10246 - Detective’s Office done 'cos of a few missing parts.

w.
(2019-11-23, 11:07)Willy Tschager Wrote: [ -> ]Yep! You have also to consider that not all model builders are able to author a part or sticker. In addition such projects are long runners. It took almost a year and half to get the 10246 - Detective’s Office done 'cos of a few missing parts.
The other problem is that most stickers (and to a lesser extent, patterns) are not so reusable, so it is less "useful" to model them, compared to create a new shape that can be used in other sets, MOCs, etc...
(2019-11-23, 11:07)Willy Tschager Wrote: [ -> ]Yep! You have also to consider that not all model builders are able to author a part or sticker. In addition such projects are long runners. It took almost a year and half to get the 10246 - Detective’s Office done 'cos of a few missing parts.

w.

Hmm, well I'm definitely stuck on authoring any complex new parts for now, but I didn't run into any great difficulty with stickers. I've used a couple of different methods, both involving Studio Part Designer. For set 5550 I applied patterns directly to the parts that are meant to receive stickers. Then I realized that a bona-fide sticker part should be its own thing, made up of a box primitive with a pattern applied to it, so I did that for the "data sheet" sticker in set 7181. (I've attached both files—would either of these be acceptable in the OMR?)
(2019-11-24, 2:09)N. W. Perry Wrote: [ -> ]Hmm, well I'm definitely stuck on authoring any complex new parts for now, but I didn't run into any great difficulty with stickers. I've used a couple of different methods, both involving Studio Part Designer. For set 5550 I applied patterns directly to the parts that are meant to receive stickers. Then I realized that a bona-fide sticker part should be its own thing, made up of a box primitive with a pattern applied to it, so I did that for the "data sheet" sticker in set 7181. (I've attached both files—would either of these be acceptable in the OMR?)

Attached to this post? Nothing is listed. But I did the 7181 sticker already:
https://www.ldraw.org/cgi-bin/ptdetail.c...41698a.dat
(2019-11-24, 3:37)Orion Pobursky Wrote: [ -> ]Attached to this post? Nothing is listed. But I did the 7181 sticker already:
https://www.ldraw.org/cgi-bin/ptdetail.c...41698a.dat

Hmm, there should have been an edit—naturally, unless I embedded my custom part, you wouldn't know whether my model was compliant, so I removed the attachments. Looking at them now, I realize I never made them cross-compatible between Studio/Part Designer and LDraw—and indeed, the sticker (or patterned) parts I have created don't appear in LDCad. So I've got more of that process to look into; perhaps then I'll realize why most modelers avoid it. :-)
(2019-11-25, 17:27)N. W. Perry Wrote: [ -> ]Hmm, there should have been an edit—naturally, unless I embedded my custom part, you wouldn't know whether my model was compliant, so I removed the attachments. Looking at them now, I realize I never made them cross-compatible between Studio/Part Designer and LDraw—and indeed, the sticker (or patterned) parts I have created don't appear in LDCad. So I've got more of that process to look into; perhaps then I'll realize why most modelers avoid it. :-)

Far as I know studio uses it's own texture mapping extension on top of the LDraw part geometry.

So any textures will be lost when 'exporting' to LDraw.
(2019-11-25, 22:44)Roland Melkert Wrote: [ -> ]Far as I know studio uses it's own texture mapping extension on top of the LDraw part geometry.
So any textures will be lost when 'exporting' to LDraw.
Yes, texture image is Base64 encoded in Studio exported "LDraw" part file (quotes because the  is not Ldraw compliant!). See https://forums.ldraw.org/thread-23400-po...l#pid32052
Cherry Picker
[attachment=4317]

Extra Building Instructions: Tow Truck
[attachment=4448]

All files are OMR compliant
Done with LDCad 1.6c
Rendered with Studio 2.0
Looks like I forgot this model...
FYI 32905 is available on PT: https://www.ldraw.org/cgi-bin/ptdetail.cgi?s=32905
(2019-12-04, 8:27)Philippe Hurbain Wrote: [ -> ]FYI 32905 is available on PT: https://www.ldraw.org/cgi-bin/ptdetail.cgi?s=32905

OK, I'll need help with that. I got the part file and the required (unofficial) subfiles, which needed another part file and additionnal required (unofficial) subfiles. Crammed everything in my MPD file, ran through MPDCenter and I got the following errors :

Error in: '42088 - s\4716s02.dat' Filename should have 3 or 4 portions.
Error in: '42088 - s\4716s02.dat' Set number from MPD file name and DAT file name does not match.
Error in: '42088 - s\4716s02.dat' Filename not OMR conform because of the above error(s).
Error in: '42088 - 1-4cyloh.dat' Unofficial file 1-4cylih.dat needs to be added.
Error in: '42088 - s\4716s01.dat' Filename should have 3 or 4 portions.
Error in: '42088 - s\4716s01.dat' Set number from MPD file name and DAT file name does not match.
Error in: '42088 - s\4716s01.dat' Filename not OMR conform because of the above error(s).
Error in: '42088 - s\4716s03.dat' Filename should have 3 or 4 portions.
Error in: '42088 - s\4716s03.dat' Set number from MPD file name and DAT file name does not match.
Error in: '42088 - s\4716s03.dat' Filename not OMR conform because of the above error(s).
(2019-12-05, 0:02)Marc Belanger Wrote: [ -> ]Error in: '42088 - s\4716s03.dat' Filename not OMR conform because of the above error(s).
42088 - s\4716s03.dat should be s\42088 -4716s03.dat (and referred to as such).
Attached the model with correct embedding. I made embedding with LDCad, first removed "42088 - " prefix using file->cleanup detect. Then File -> cleanup -> Check "embed unofficial part AND add prefix -> use OMR.. button and key in 42088 set number.

Roland, would there be a way to properly embed part without removing prefix first?

And otherwise, I also got the buggy '42088 - s\4716s03.dat' when I tried to embed the wormscrew directly by right-click on it -> reorganize -> embed unofficial content -> fill in '42088 - ' as prefix. Roland, I guess this is a bug Wink
I'm planing to do a rework of these subfiles tonight.
(2019-12-05, 8:30)Philippe Hurbain Wrote: [ -> ]42088 - s\4716s03.dat should be s\42088 -4716s03.dat (and referred to as such).
Attached the model with correct embedding. I made embedding with LDCad, first removed "42088 - " prefix using file->cleanup detect. Then File -> cleanup -> Check "embed unofficial part AND add prefix -> use OMR.. button and key in 42088 set number.

Roland, would there be a way to properly embed part without removing prefix first?

And otherwise, I also got the buggy '42088 - s\4716s03.dat' when I tried to embed the wormscrew directly by right-click on it -> reorganize -> embed unofficial content -> fill in '42088 - ' as prefix. Roland, I guess this is a bug Wink

Sounds like a bug, it should use 'smart' prefixing (preserving the \s or whatever subfolder stated)

But if it already has the wrong prefix this approach will see it as part of the folder name. Which imho is the correct way to handle it. (crap in crap out Smile )

I'll check and make changes in 1.6d
(2019-12-05, 10:03)Magnus Forsberg Wrote: [ -> ]I'm planing to do a rework of these subfiles tonight.

Done now.
I have eliminated the need of the primitives 1-4cylih and 1-4cyloh, and replaced them with a fourth subfile 4716s04.
I still get a incorrect version of 4716s01 when I download this part from the PT. Something is strange....
(2019-12-05, 8:30)Philippe Hurbain Wrote: [ -> ]42088 - s\4716s03.dat should be s\42088 -4716s03.dat (and referred to as such).
Attached the model with correct embedding. I made embedding with LDCad, first removed "42088 - " prefix using file->cleanup detect. Then File -> cleanup -> Check "embed unofficial part AND add prefix -> use OMR.. button and key in 42088 set number.

My version now works. I don't get any warnings nor errors from MPDCenter. The only differences with my file and yours is in the axl4hol3.dat part. Is my version correct?
(2019-12-05, 22:57)Magnus Forsberg Wrote: [ -> ]I still get a incorrect version of 4716s01 when I download this part from the PT. Something is strange....
Cache issue? part works for me.
Yes it is. That's because I updated that primitive in the meantime!
(2019-12-05, 22:46)Roland Melkert Wrote: [ -> ]But if it already has the wrong prefix this approach will see it as part of the folder name. Which imho is the correct way to handle it. (crap in crap out Smile )
Yes. The underlying idea was to use the detect prefix mechanism to propose that prefix to the user, but indeed it doesn't work very well if you try to detect prefix and there is none...