[DOA5LR] Timmy's Private Stash, Tips & Tools - UPDATE(11/21/18): Kinky Overalls

[timmyc]

So I messed with some user defined ref files while testing 0.2.3a and made a LOD ref for Hitomi using my tool.   I think it turns out quite nicely.

 

Timmy LOD

Normal

 

If there's interest I can make some LOD ref models and share.

[minion88]

Awesome thread! Will definitely be using this as a guide as I learn how to mod.

[Guest]

So I messed with some user defined ref files while testing 0.2.3a and made a LOD ref for Hitomi using my tool.   I think it turns out quite nicely.

 

Timmy LOD

Hitomi_LOD_Test (1).jpg

Normal

Hitomi_LOD_Test (2).jpg

 

If there's interest I can make some LOD ref models and share.

 

I like your version better. If I wouldn't be busy with my Skyrim and Sims 3 projects, I would certainly play DOA and install your mods. You're doing excellent job.

 

[201193]

So I messed with some user defined ref files while testing 0.2.3a and made a LOD ref for Hitomi using my tool.   I think it turns out quite nicely.

 

Timmy LOD

Hitomi_LOD_Test (1).jpg

Normal

Hitomi_LOD_Test (2).jpg

 

If there's interest I can make some LOD ref models and share.

 

Make that UHLOD and so many dams will break at so many places.

 

I have been trying to do just that, but I always want a fuller pair and a bigger behind .

 

infact I like them even BIGGER than UHLOD proportions (greedy!!).

 

Hope you can do that if it is not too much trouble (original or modified UHLOD references)

 

Finally,  Keep Rockin' Timmy 

[timmyc]

Soo I noobed up again and sent out a debug build of 0.2.3a instead of the release build, lol.  Sorry for the unnecessary extra spam of debug messages I put into debug mode.  Please download the release version from the original tool post.

 

I've updated the link to indicate that it is the release build DOA5LRCharacterTransformerTool_V_0_2_3a_release.zip

 

 

Updated tool can be found here http://www.loverslab...lena/?p=1460252

 

Sorry for the massive debug message spam that you guys have to put up with in the wrong build   Release build is also slightly faster - although it's already plenty fast :bush:

[Mikkukan]

Very nice giftwraps for the girls, timmyc.  I'd like to see one for Lisa (maybe in different color) and Momiji as well.  Also, a wingless version for Nyo would be nice!

thank you for the xmas update! 

 

 

 

game 2016-02-22 01-41-29-87.jpg

 

 

I think she would look better without that big knot on the back of her head, or try make it smaller as a side garment.

[timmyc]

Okay I hope that was the last of my mistake, LOL.  I accidentally uploaded my own test version with a cap of SF500 (oops! It's a dangerously high number to leave smoothness at hahha).  The new link should be working now.

 

I was messing with smoothness level and found that most of the time, SF30 default seems to work just fine without any noticeable flaws.   Adaptive mode is great at preserving the shape of characters at the expense of slightly deformed edges, but most of the time regular smoothness mode gives you very very similar results which is essentially a fast approximation method.    It also runs very very fast with higher smoothness only giving it slightly longer run time overall.

 

 

Very nice giftwraps for the girls, timmyc.  I'd like to see one for Lisa (maybe in different color) and Momiji as well.  Also, a wingless version for Nyo would be nice!

 

I think she would look better without that big knot on the back of her head, or try make it smaller as a side garment.

 

I was hoping it would be enough since you could transform between them now, but I can see how the ribbons and wings can get in the way of enjoyment, lol.  Perhaps I can fix them up real quick.

 

[Guest GuachaLover23]

 

Sorry TimmyC u will always have my support, this re up limits is BS

 

Mr Duck

[timmyc]

Thanks!  I really appreciate it.

[Guest GuachaLover23]

Thanks!  I really appreciate it.

 

No problem pal.

 

Mr Duck

[LGMODS]

Talking about SF, if I can give a test that I made :

For my kinky Nyotengu I used edited Phase 4 body for transformation and tried Adaptative SF30, SF50 and SF100 but I was only happy with Adaptative SF300, it's high but it gave a better result for this mod from Phase 4 to Nyotengu.

Maybe it's because of my edited body (I mean maybe an original body won't need a SF300) but the result was looking very smooth and beautiful, then SF500 could maybe make even more since all lower was not to my taste I guess (for this mod only).

 

Anyway this option is great, depending of the character it's really cool to test different SF, especially on Marie Rose.

 

 

[timmyc]

Talking about SF, if I can give a test that I made :

 

For my kinky Nyotengu I used edited Phase 4 body for transformation and tried Adaptative SF30, SF50 and SF100 but I was only happy with Adaptative SF300, it's high but it gave a better result for this mod from Phase 4 to Nyotengu.

Maybe it's because of my edited body (I mean maybe an original body won't need a SF300) but the result was looking very smooth and beautiful, then SF500 could maybe make even more since all lower was not to my taste I guess (for this mod only).

 

Anyway this option is great, depending of the character it's really cool to test different SF, especially on Marie Rose.

 

Yeah, i tested 500 briefly and found some turn out okay and some turn weird.  I can relax it to 500, but I am afraid that most people are going to see that 500 > 100 and just use 500 all the time.  However, the higher does not always mean the better, and they see crappy outcome and say this thing sucks, lol.  

 

What you are doing is the best actually - trying out many values to see what is to your liking.  I am trying to take a bit of a controlled approach to guide most users to what I think is the most optimal outcome, which is why I left the adaptive mode unchecked and smoothness at 30 which works pretty well for most transformations based on my own tests.

 

If people want 500 I can push an update out with a cap of 500.

[LGMODS]

Yes I understand about 500 > 100, you are right I think not everyone want to test every SF, this is more for thorough persons and can confuse others.

If you think this is not a good idea then I understand that you don't want add it, but know that I could be interested for some specific mods, I will also wait to know if you will do it before make transformation of my OT-LOD for give some more test.

 

Giving a little guide about this SF option is a good idea, as you said higher SF don't always mean better result and the choice to keep SF 30 per default is a good idea, keep it, since it's the most approximative best result for most of transformations.

 

 

[Edit] About SF500, you could release two versions of your tool maybe, a normal (something like SF100 max for people who want quick transformation) and an advanced one with all options.

[pringles89]

Yes I understand about 500 > 100, you are right I think not everyone want to test every SF, this is more for thorough persons and can confuse others.

If you think this is not a good idea then I understand that you don't want add it, but know that I could be interested for some specific mods, I will also wait to know if you will do it before make transformation of my OT-LOD for give some more test.

 

Giving a little guide about this SF option is a good idea, as you said higher SF don't always mean better result and the choice to keep SF 30 per default is a good idea, keep it, since it's the most approximative best result for most of transformations.

 

 

[Edit] About SF500, you could release two versions of your tool maybe, a normal (something like SF100 max for people who want quick transformation) and an advanced one with all options.

please give Lola what she wants so she can make mods even faster (for me)! j/k about the last part, mostly

 

So I messed with some user defined ref files while testing 0.2.3a and made a LOD ref for Hitomi using my tool.   I think it turns out quite nicely.

 

Timmy LOD

Hitomi_LOD_Test (1).jpg

Normal

Hitomi_LOD_Test (2).jpg

 

If there's interest I can make some LOD ref models and share.

LOD ref models would be great (H-LOD would be even better!)

[timmyc]

Sorry everyone.  After uploading an update to the 0.2.3a release build yesterday. I had forgotten to change the share policy on my media fire for the file to public.   Please try again and let me know if you have any issues downloading it.

[zoidnoid11]

I tried out the latest ABC tool today, just because it's been a while since I used it and it's been updated about a hundred times since. Not going to go into too much detail, but from my own testing I can safely say your tool is noticeably superior. Processing time has always been better, but the end result with each version keeps improving, whereas the other tool seems to be introducing bugs as it tries to add features (though I admit that may also be down to me not quite understanding what those options are supposed to be for ).

 

A couple quick questions.

 

Just to reconfirm, there's no need to run Skeleton patcher if AutoBC works, right? I'm just surprised it continues to be included in the package.

 

Is there a technical reason the transformation is automatically 'adaptive' past SF100? Only reason I ask is that Sarah's one of my favorite characters to play, but she's the least attractive girl (not saying much, the model's still prettier than any Tekken, MK, etc models), and I'd like to see if we could improve the situation by giving her someone else's proportions by upping smoothness. Not a big deal either way, her weird shoulder bone positioning is probably ultimately to blame.

[timmyc]

I tried out the latest ABC tool today, just because it's been a while since I used it and it's been updated about a hundred times since. Not going to go into too much detail, but from my own testing I can safely say your tool is noticeably superior. Processing time has always been better, but the end result with each version keeps improving, whereas the other tool seems to be introducing bugs as it tries to add features (though I admit that may also be down to me not quite understanding what those options are supposed to be for ).

 

A couple quick questions.

 

Just to reconfirm, there's no need to run Skeleton patcher if AutoBC works, right? I'm just surprised it continues to be included in the package.

 

Is there a technical reason the transformation is automatically 'adaptive' past SF100? Only reason I ask is that Sarah's one of my favorite characters to play, but she's the least attractive girl (not saying much, the model's still prettier than any Tekken, MK, etc models), and I'd like to see if we could improve the situation by giving her someone else's proportions by upping smoothness. Not a big deal either way, her weird shoulder bone positioning is probably ultimately to blame.

 

Hey thanks for the candid review!   I really appreciate it.

 

The tool started out as a pet project as I had pretty much lost all interest in DOA5 modding prior to making this one.   To me,  speed is of the essence, especially in graphics applications where workloads are easily distributable among multiple cores/threads.     Another one of my proprieties is to make sure that the tool is simple enough to understand and use, which is why I opted to just have one number for people to play with (and I really hate pop up windows when I am working on mods, lol).  With the inclusion of adaptive mode, it complicates things a little, but I made sure to add a tooltip to it (if you mouse over my stuff you can see some explanations as to what they do) to avoid confusion.

 

As for the reason why I still include the now legacy Skeleton Patcher (I still think it's one of the most reliable tools - it simply works), I included it because a few transformations I tested (eq. Xmas In July Nyotengu/Helena outfits) don't work very well with my bone stuff, so it is there as a backup and the reason why the bone stuff is toggleable.    In these cases I have also tried to use bone tool to no avail, so skeleton patcher is the perfect fall back method.  However, you should only use either the built-in auto bone/collision transfer or Harry's tool.  Not both of them.

 

The cap of SF100 for normal smoothing mode is a design decision I made since anything over that is going to be very deformed as shown below -

Ayane to Marie @SF500 regular

 

I don't think it is a good user experience, so I decided to cap the normal mode at 100.   I'd like the experience to be a somewhat controlled one, so I set the default smoothing mode to regular at 30, which I think, more often than not, gives the most "optimal" outcome.

 

I have been considering relaxing the constraint a bit on regular mode, but I am afraid that some people might think that the higher is better and just use maximum for everything and think this stuff sucks if the outcome is not what they expect

 

As for what the smoothness value is the best - it really depends.  It depends on the kind of model you have, so I strongly encourage a bit of experiment with the different values to see what turns out best for your model.

 

I haven't tried many Sarah transformations, but I can see where you are going with high smoothness.  I think making it very high would probably fix the stiff shoulder problem, lol.  100SF seems pretty decent going from Ayane to Sarah.

 

[Verhoven69]

Feedback:

 

The transform tool you made is really helpful. Compared to the ABC tool, yours managed to fix the seam in the buttcrack which the former couldn't fix.

[timmyc]

So I figured out one of the last issues with the transformer which is, once again, in regards to transformation from pretty much anyone to Marie, lol.   This time, however, it appears that the fix benefits other characters as well to some degree.

 

This is update brings some slight quality improvement to the adaptive mode.

 

Change Log
V0.2.4
- Further enhanced adaptive smoothness mode to correct some odd normal vectors around chest area (especially apparent on large girls such as Helena/Lisa to Marie transformation)
- Slightly tweaked the adaptive smoothness mode to increase the performance by a little bit
- Added new start-up message to advise users to use either the built-in bone/collision transfer or a third party tool.
 

 

The focus of this update is to get rid of anomalies on the chest area after transformation using adaptive mode.

For example -

Without the enhancement, old logic would make parts around chest somewhat dark and unnatural looking.  Very different from her natural lighting.  This update aims to address this issue.   The challenge is to figure out where the breasts start and end, lol, and i was able to leverage a certain geographic feature of the breasts, since they kind of stick out :bush:,  to pin point their location dynamically on the fly - no hard coded coordinates required

 

 

When I was drafting the update, I thought about just putting the "recompute normal vectors" functionality into this transformer as well, but I figured that if you guys want the perfect (or rather, better) normal vectors on the transformed models, you could always just use my converter for this purpose.  No sense in spending time on redundant feature and reinventing the wheel, so to speak.   With this recompute normal vector functionality, the normal vectors are recalculated based on the information of the model in real time and all the triangles have their ends smoothed just like what a regular editor such as 3ds Max would do.  The result is a very smooth looking model.  However, this also exposes issues in the model if fake normal vectors are used to mask certain rough edges - recomputing normal vectors in this case would expose these problem areas.

 

I will take a little time here to review how to use my converter for pretty much the same purpose here.  My converter and related tutorials can be found here http://www.loverslab.com/topic/50997-doa5lr-timmys-private-stash-updated-101015/?p=1346825

 

So if the model ends up having weird normal vectors like the shots from V0.2.3a pictured above, my other tool can deal with this issue very easily.

Drop the TMC you wish to modify onto the tool.

 

Make sure you have these settings selected and uncheck everything else and hit transform   For more on what the seam smoothing value does, please refer to my tutorial here http://www.loverslab.com/topic/50997-doa5lr-timmys-private-stash-tips-tools-updated-022316-transformer-v023a-converter-v040a-updated-giftwrap-set-for-nyohelena/?p=1359441

The model should now be repaired!

 

Comparison shots

 

Some in-game shots

The new normal vector correction on adaptive mode appears to work fine on non-Marie target transformation as well.  These are done @SF100 adaptive.

 

I've also included a bit more information on the bone stuff in the start-up messages.

 

I have also included some test versions of Timmy LOD reference models I quickly put together for you guys to play with.  Just drop them into the "UserDefined" folder and you can transform between the regular ref models and these LOD models.  For instance, Ayane can be transformed to her bigger self - UD_Ayane_LOD:

 

I hope she is visibly bigger, lol!

 

Updated tool & sample ref files can be found here http://www.loverslab...lena/?p=1460252

 

[Guest]

So I figured out one of the last issues with the transformer which is, once again, in regards to transformation from pretty much anyone to Marie, lol.   This time, however, it appears that the fix benefits other characters as well to some degree.

 

This is update brings some slight quality improvement to the adaptive mode.

Some in-game shots

0_2_4_tests (1).jpg0_2_4_tests (2).jpg0_2_4_tests (3).jpg

Most brilliant.... Marie Rose is divine.. looks perfect. Thank you.

[timmyc]

I fear this has become a pattern lol.  A few hours after releasing an update, today I tried it again to make sure all is well and ... I found a bug already.

 

Sorry for such frequent update spams, but this one is necessary to uphold my title as buttcrack meister

 

I only focused on the front side yesterday and had completely forgotten to check the backside, and today I did - I was horrified by what I saw!

 

Change Log
V0.2.4a
- Corrected an issue where the top of the butt crack would have their normal vectors overly smoothed in adaptive mode.

 

 

Hopefully there will be no more game breaking issues like this.   Please let me know if you run into any issues and I will try my best to address them. 

 

Updated tool & sample ref files can be found here http://www.loverslab...lena/?p=1460252

[timmyc]

Feedback:

 

The transform tool you made is really helpful. Compared to the ABC tool, yours managed to fix the seam in the buttcrack which the former couldn't fix.

 

I'd like to thank you for reminding me to check the characters' backside after transformation, and thanks for the feedback!

[timmyc]

If people like the LOD models I can spend more time on finishing them (and polishing them up) and release more of them.  The sample pack is just to test the waters.

[201193]

If you can make the option for UHLOD I will be grateful. (bushfully that is.)

 

Hope it is not much of a hassle.

 

Again, Keep up the great work timmy.

 

Rock On!!

 

 

[zoidnoid11]

All hail the buttcrack meister.