Does Studio make creating Technic MOCs easier?

[StudWorks]

I just started making MOCs and would like to make instructions someday. I already have Studio and currently I build everything IRL. Is it easier to build MOCs in Studio first? Especially with exteriors and chassis?

[Mr Jos]

I prefer to make it IRL first, and try untill satisfied. Then make the Studio file. Certainly for moving parts it's better to have it in your hand and test it.

[GerritvdG]

Also prefer to make IRL first. Works faster and you can test everything immediately. Lego is a very good tool voor rapid prototyping. Studio is great for testing other colours or pieces you don’t have or don’t have in the correct colour.

Finally if everything is really finished start with making instructions. I’m using my real life MOC to test what’s working when building and what’s not. 

 

 

[Seasider]

I often do a combination of physical build and stud.io.

I use physical building for testing mechanisms are working how I expect them to.

I use Stud.io to work out how to connect structures as you can chance them quickly.

But I also build my Stud.io creations as physical builds during the whole creation process to check the model connects together properly and is stiff enough. I’ve found a few times I’ve built floating structures in stud.io and I didn’t realise till I’d built it physically.

 

[astyanax]

Over the last year I've shifted to building in LDCad first, to make a full virtual mockup, testing only small tricky subassemblies as table scraps. Then I order parts (if needed) and build the physical model. Finally I update the virtual model with the improvements I came up with while building. 

But this really only works if you're very proficient with the software. With LDCad I know all the shortcut keys and so I'm very fast. I find it very convenient to quickly try alternative solutions in virtual reality, because you don't have to take apart the whole subassembly just to swap a buried pin or connector. I never managed to get a good workflow in stud.io. 

[Kostq]

Stud.io is a pain in the megablocks if you build something with lot of axles.
Also - you cannot easily test the rotation mechanics of meshed gears and/or levers.

Create physically if you can, then digitize it, then mod the physical, then re-do the digital model.
Then make instructions and reiterate until you hate yourself.
Then you're almost ready to release it to the world.

Good parts of stud.io ->

1. COPY-PASTING of identical parts - boy, does this save time!
   If you've made the instruction steps and copy the assembly - you've saved the time to re-do them too. Imagine a repetitive build with 8x sections.
    
2. You can use the forum to find LXF or IO or LDR files /or the other 3d formats it accepts/ and either compare your build to something or check compatibility.
   For instance - I'm currently making a 42106 MOC that aims to be compatible with MACK truck which I don't have. Yet I can import it and check the connection points and the scale EASILY.
    
3. If you're building a c-model - it shows you the remaining parts if you've imported the set's palette.
4. You can SEARCH for specific parts in a build - if you run out of spacers or specific connectors - you can find and re-do a section to "release" them.
5. You can get precise dimensions and weight and parts count for any moc imported in studio.
6. Re-coloring is fast
7. Built-in renderer. Rendering animations can save you buying camera gear for youtube or other demos.
8. Built-in instructions maker, which is decent at its job.

Bad parts of stud.io

1. Flexible parts are a pain in the gigablocks
2. It's RAM HUNGRY AF
3. You'll need a external versioning system - save often, and save as new files OFTEN too. Just in case.
4. Instructions take a lot of time to produce. They look good though.
5. Rotations are hard sometimes, you have to hide LOTS of parts so a single beam can rotate around a single pin. Beware
6. Axles are nightmare to position properly without the connect tool
7. Hinges may work, may not work - your luck.
8. You have to calculate gear mesh angles by hand. And the orientation of each meshed axle afterwards.
   3x16t gears cannot be placed one below the other if you don't rotate SOME of them. NOPE.

[Lipko]

For rapid virtual prototyping I'd toss Lego Digital Designer into a MOCer's toolkit. It's very quick for throwing together stuff, the biggest disadvantage is that many parts are missing from the inventory, though I believe there's a workaround for it. Never really needed it though. For more serious virtual design (for making animations or instructions), other tools are better of course. And I'm not falimilar with STUD.io, maybe it has pretty much the same features as LDD. If so, ignore my comment.

[langko]

I'm also one who prefers building in real life. I personally think of better ideas when the pieces are in my hands. It's easier for me to tell how strong a connection or assembly is as well. I spend a fair chunk of time on the computer at work, using solidworks and the like... So I really enjoy having a hobby where I can be creative whilst not sitting in front of a screen.

If the goal was to make instructions I would personally rebuild the model digitally after I was happy with it in real life.

Edited July 28, 2021 by langko

[Seasider]

48 minutes ago, Lipko said:

For rapid virtual prototyping I'd toss Lego Digital Designer into a MOCer's toolkit. It's very quick for throwing together stuff, the biggest disadvantage is that many parts are missing from the inventory, though I believe there's a workaround for it. Never really needed it though. For more serious virtual design (for making animations or instructions), other tools are better of course. And I'm not falimilar with STUD.io, maybe it has pretty much the same features as LDD. If so, ignore my comment.

Unfortunately I’m Mac based and found LDD no longer worked on my Mac so switched to Stud.io. 
 

it’s very similar in lots of ways and I’ve got a lot more proficient with time. One tip is to use temporary pins and axle pins to connect/position parts and add the axles as the last step

[syclone]

As my tag suggests, LDD is my preference. Tried Stud.io - a pain in the a$$ to work in, overcomplicated and barely has any advantages now that new parts can be added into LDD in a matter of few minutes (check digital forum) and developer options can be enabled.

Recently also started using LDCad+LDpub, but ONLY for instructions, and it is A LOT better than stud.io for part positioning and flexaxles, although the tutorials don't explain very well what to do and assume some stuff for known, but once you get over it, stuff makes sense and the program becomes easier to use with hotkeys.

My parts inventory has always been limited, so LDD allowed me to build a lot of stuff I didn't have parts for. So far ~95% of my builds started in LDD, then I test the model and progress forwards either physically or digitally. It's very user-friendly, so it feels natural to build unlike the professional CAD nature of LDRAW-based tools. Stud.io seems to have tried to combine LDD and these LDRAW tools, but in my eyes it has a horrible UI (feels like usng an Apple computer) and general building procedure (aforementioned problems by others), so it never caught me in. 

Having short patience, not having to dismantle tens of parts just to change something small also is very helpful.

Another bonus of digital building is the ease of backup - I still have .lxf files from when I started using LDD in 2012, but a lot of photos/videos got deleted over time to save space.

Now, is it something essential? depends on the person. It has been discussed countless times before, and some absolutely despise any sort of digital building, others only use digital tools... whatever fits you better.

 

Edited July 28, 2021 by syclone

[Lipko]

I have to add that in my opinion, the prerequisite of working with any virtual design environment efficiently is that you are proefficient with the bulding techniques (mechanisms, gearboxes and structures). Either by building lots of sets or by building MOCs in real life. I, for one, rarely test-build virtual assemblies, the reason I regularly switch between virtual and real building is that in many cases it helps me (I don't fully understand how) to overcome impasses. I can't remember too many occasions when the real build didn't behave as I expected from the virtual model.
Truss designs (when working with non-perpendicular connections so cannot utilize perpendicular beams) is one example that needs real life test building. Especially because of twisting.

Designing only with real building has a big negative effect on me: I tend to stick to sub-optimal solutions because it would be too tedious to fix in real life. Change a deeply-dug beam in a real build vs in a virtual environment. You get the idea. This can be overcome if you can afford to copy and rebuild stuff and have 2-3-4 prototypes built at the same time, but this required a fairly big inventory, especially if special parts are involved.

EDIT: this is just me and some other builders. There are many builders, some of them are among the best builders out there who doesn't use any virtual building environments during development.

Edited July 28, 2021 by Lipko

[Jundis]

Coming back to your question:

If you just start MOCing and don't have lots of parts in your collection, I would really suggest you use stud.io as it is linked to bricklink and you can easily order missing parts and also virtually try different configurations without the need of a big collection.

Markus Kossmann once said that the Lego designers only build in RL, as they have the complete pallet in front of them. And (as many others here stated), a RL build really shows the mechanism in working condition. For my first MOCs, I had many instances that I completed a MOC virtually, ordered parts and then saw that something didn't work out. This can be even some impossible combinations which can't be assembled in RL.

Now, my usual approach is:

1. Start with a stud.io-built till I think "This and that main assembly seems to work"
2. Order some parts, if needed and built assemblies separately, like you would built a real Lego model
3. Test it and improve it
4. Repeat steps 1 to 3 till your model is finished :D

[Jurss]

Mostly I use studio to develope some part, where during building (developing) process lot of disassembling is involved.

Sometimes I check for some parts which I don't have, if that would be usable.

Sometimes I do some scaling.

But mostly I build for real.

[Gray Gear]

I prefer to work with real bricks, and only do the digital model afterwards for the building instructions. That way I can see right away if something works or is rigid enough, and don't keep building on an uncertain foundation.

I also like using my hands and how the bricks feel while building, it is something I don't want to lose.

[StudWorks]

That’s really excellent feedback thank you everybody!

I have never made instructions for a real life build before, any suggestions on how I can make  instructions after I have already built the MOC?

[Lipko]

40 minutes ago, StudWorks said:

That’s really excellent feedback thank you everybody!

I have never made instructions for a real life build before, any suggestions on how I can make  instructions after I have already built the MOC?

It is a very vague question. How would you approach a problem/task in general?
For making instructions, it's pretty simple, to be honest. This will be lengthy but only bcause of the details.

1. You make a full virtual model of the real model. For even big and complex models it's not that hard, you don't necessarily have to tear don't the whole model. But that's also an option: Tearing down the model gradially making photos after every step. This makes sure that the model will be assembleable. Using the photos, you can recreade the virtual model. This will be risky, you may forget to take photos or a photo doesn't show all changes.
2. export the 3D to some software that's capable of instruction making, for example MlCad. It's still not the final document maker program, you'll need it to make the steps. But maybe you built the model with this program. In that case, ignore this step.
3. Make the steps. The 3D file is actually just a huge list of the parts with their color, position and orientation. In some random order. No connectivity info, so is NOT usable as steps. In reality, the MLCad file is just a plain text file you can also view and edit in any text editors.
   So you make the steps by rearranging the parts in the list. The way this can be implemented in the User Interface will vary. In MLCad for instance, you have a rendered image on which you can clink and select parts, and there's also a window with the list and the selected parts are highlighted. Arranging can be done by dragging on the list (I'm not sure about this) or cut+paste the list entries.
   The steps themselves are just special text entries (maybe STEP in MLcad) in the list. You can (it's a good practice) make subassemblies (I honestly don't remember how that's implemented in MLCad file, maybe sumething like "sub name_of_subassembly" part list  "end_sub")
4. Make the actual instruction document itself in the program (or in a separate program, like LPub). It renders the text file as an actually instruction-looking series of images and all the visual stuff you'll find in an instruction, like parts list per step, length annotations on axles, and so on. You can place the rendered step images as you wish, you can set subassemblies to be rendered as callouts (little bubble with sub steps), and so on.
   By the way, in case of MLCad the file you are working on in LPub is still the same text file but with a tons of meta-commands added by LPub under the hood. These meta-commands will be ignored by MLCad commands but still be visible in the text list view. You have to be careful when jump back and forth between MLCad and LPub, some things might get screwed)
5. Export the document in the program to something that can be published. Like a series of images or a PDF file.
6. Some post processing if you like. I, for example, find it much easier to add arrows, highlighting circles and such at this point. It can be done in MLCad too in a very incovenient way (arrow is added as a part that has to be hidden in later steps, I don't even remember how).

 

Edited July 28, 2021 by Lipko

[GerritvdG]

2 hours ago, StudWorks said:

I have never made instructions for a real life build before, any suggestions on how I can make  instructions after I have already built the MOC?

Another approach is using studio for everything. Nice tutorials can be found here https://youtube.com/playlist?list=PLaEL-gouFOoHg5Okoua-cIGxjyrPOOUql

[Berthil]

9 hours ago, astyanax said:

Over the last year I've shifted to building in LDCad first, to make a full virtual mockup, testing only small tricky subassemblies as table scraps. Then I order parts (if needed) and build the physical model. Finally I update the virtual model with the improvements I came up with while building. 

I also do it like this for my GBC machines but with Stud.io. It requires IRL building experience.

Never start with making submodels, building steps and instruction until the model is finished IRL and tested. Otherwise it easily could be double the amount of work to create instructions and making instructions is already a LOT of work.
For making a modular building I make a 100% digital design in Studio first, get the missing parts, build it and adjust file where needed and also create instructions afterwards.
The renders out of the box from Stud.io are realistic enough to judge a digital modular design for me.

Edited July 28, 2021 by Berthil

[howitzer]

I don't use Studio because I don't like the UI, but I use LeoCAD instead. Usually when building something more complex, I'll model it on computer while building it also IRL, as that seems to minimize the amount of taking apart and rebuilding models. I like building without digital tools more but with Technic it's just really hard to think ahead enough to avoid constant disassembly-reassembly loop, which takes the fun out of building. Smaller models (like my TC20 entry) I've built without digital tools though.

[Mr Jos]

16 hours ago, StudWorks said:

That’s really excellent feedback thank you everybody!

I have never made instructions for a real life build before, any suggestions on how I can make  instructions after I have already built the MOC?

I have made all my machines as a real life build first. I use Studio for everything, no other programs needed (only for my software but that's not needed for most normal technic builds).

I take it if possible in sub-assemblies and make that part in Studio. What I do is already start 'drawing' in the order I think it should be build step by step (this makes the process for making instructions later SOOOO easy). No problem if you're not sure, you can change and add steps later anyway, but remember, DO NOT start making instructions untill the 3D is finished, sometimes the program likes to mess up the instructions if you change something to much in the 3D build. Just changing a pin to another step because you added it to soon, in floating air is no problem though.

If you finish a sub-assembly, just select every piece of it, and make it in Studio into a 'subassembly' like here in my example, every arm has 1item in the main 3D steps.

You can then hide it completely to make the next sub, or hide partially to connect the next part to some of the beams, and later hide it.

 

How big are your builds? Because I notice Studio gets greedy of the RAM and does not like my 5000+ pieces builds to show everything. 2000 pieces works perfect.

[Aleh]

On 7/28/2021 at 1:14 AM, Mr Jos said:

I prefer to make it IRL first, and try untill satisfied. Then make the Studio file. Certainly for moving parts it's better to have it in your hand and test it.

Totally agree! I like to touch parts, to experiment and not looking in the desktop of my PC. 

[Berthil]

21 hours ago, Berthil said:

Never start with making submodels, building steps and instruction until the model is finished IRL and tested. Otherwise it easily could be double the amount of work to create instructions and making instructions is already a LOT of work.

 

9 hours ago, Mr Jos said:

DO NOT start making instructions until the 3D is finished, sometimes the program likes to mess up the instructions if you change something too much in the 3D build. Just changing a pin to another step because you added it too soon, in floating air is no problem though.

 

Double warning and more when reading the other comments! Stud.io will mess up all your (submodel) pages in the page layout of your instructions when changing too much in the original 3D model and that's even apart from bugs in the Stud.io that can do that anyway. ALWAYS make a backup of your work when you are satisfied with it before making changes the next day!

Edited July 29, 2021 by Berthil

[Aleh]

Yeah, I build my latest MOC first in real life, and after it was finished I stared digital version and was sure there won't be any changes.

[Thirdwigg]

This is a helpful topic, and one right up my alley right now.

I really jumped into Studio in the last six months. I had trouble with LDD and MLCad, and it was time for me to improve my instructions.

I build all of my models IRL first, and then build it in Studio. I build in Studio with the instructions steps in mind, so create steps as I build. Then I produce the instructions, which takes some time, but the result looks good.

I agree with others that the flexible parts are hard to do well, and some parts (like the pneumatic valve BB0874) are absent, and I am not clear as to how to add parts.

Maybe I'll build a MOC in Studio first at some point, but it will take some time before I am ready to do that.

 

 

[Berthil]

36 minutes ago, Thirdwigg said:

I agree with others that the flexible parts are hard to do well, and some parts (like the pneumatic valve BB0874) are absent, and I am not clear as to how to add parts.

It's possible to import LDraw parts (with needed subparts) into Stud.io as custom parts (until they are added to Stud.io).
This makes it possible to position pneumatic elements the right way in your digital build but needs adjustment of parts list afterwards.

It is cumbersome to position flexible elements in Stud.io but temporary fixings/click points can make it easier. 

Edited July 29, 2021 by Berthil