Very timely! With displacement work so well for nanite, AND quixel ORD channels having displacement, I'm doing all kinds of heavily displaced materials. This is a super-useful (and pretty quick) technique for breaking up repetition at distance while keeping crazy detail up close. Thanks again!
Awesome!! IMO, definitely one of the BEST Texture Bombing methods I've seen so far! Question: Based on your experience, I would be interested in knowing which Texture Bombing Method (say the 5 most common?) works best with which types of Textures. For example, I think I read that Cell Bombing works best with granular-type textures (fine stone, dirt, cement, etc??.) and others work best for more defined textures (Ground Twigs/Leaves textures), etc. Thanks again for the video!
I use cell bombing for any organic texture. rock, grass, ground with twigs, or anything. But in case of textures that have pattern, like brick or tiled wood, the bombing methods don't work correctly.
Алексей, привет! Прикольно встретить здесь своего приятеля. Такого рода инструменты есть много где , в короне например, но для анриала урок весьма полезный.
Thank you for posting the video! I have a question. Is there a way to reduce Texture Repetition for textures with patterns (e.g. bricks or tiled wood)? If there is, I would like to ask if you can make a video!
I didn't exactly understand the question, but I think what you're talking about. If you increase the displacement magnitude, some areas may look bad, especially in the black areas of the alpha channel of the cell texture. I hide it by decreasing the displacement magnitude.
In short: very little, the Texture Variation uses noise instead of cell & it blends the "visible lines" based on height rather than the same texture alone. The cost is similar but the result is cell can give more variation (without looking as broken). There is a little more to it, but that's the basics.
Any notes on the performance? I noticed it samples the texture at least twice as much but I would be interested in the precise calculations and performance hits on a big terrain. It also needs to sample that voronoi texture but that doesnt have to have a huge resolution, right?
The most important thing is the increasing sample count, I think. I haven't gone deep into the exact performance calculations though. When using with landscape material, I'd use it only on layers where the repetition is really obvious and can't be hidden by foliage and stuff.
You add an offset from the Voronoi texture, which has a fixed size, so I think it would repeat if the map is larger than the area covered by the Voronoi texture.
@@gone41674 the voronoi texture is also tiling. so it can go on forever. just make sure the relation between the texture size and the cell size is reasonable. like I mentioned in the considerations section of the video
@@HojDee Yes, I’m aware that the Voronoi texture is tiled, which is the problem. Tiling causes repetition, since the UV offset is read from the Voronoi texture, the final texture will also repeat. This repetition will occur if the world size exceeds the coverage of the Voronoi texture.
@@tomastoegel4973 in the end, I made it a material function. It can be used in any material inside unreal engine. For 3ds max, the principles are the same I guess. If you're familiar with its material system you should be able to create it
hi! get it here: www.hojdee.com/courses-details/free-downloads and here's the tutorial on how to get it, if you run into any issues: ruclips.net/video/8yRxmuY9If4/видео.htmlsi=770df6CkfG-jx-Cj
6:27 Why divide? Multiplication could be used to achieve the same result but with much cheaper instruction cost. Divide should be avoided where it's possible.
@@SirRebonack multiply is cheaper, you're correct. But with modern hardware, the difference is not that significant. If you wanna make it really optimized though, you should go with multiply for sure.
@@nikoheino3927 The shader compiler cannot optimize this division because it is based on dynamic data (sampled texture). Division is fine to use on parameter values, which are able to be optimized. Division is significantly more expensive than multiplication.
Very timely! With displacement work so well for nanite, AND quixel ORD channels having displacement, I'm doing all kinds of heavily displaced materials. This is a super-useful (and pretty quick) technique for breaking up repetition at distance while keeping crazy detail up close. Thanks again!
@@inwardocean glad to help :)
Be careful of doing this too much, as it will start to affect performance
Bro this is incredible. You have no idea for how long i have been trying to find this.
glad I could help
You are absolutely killing it bruh, I learned a lot from you
glad to hear that man :)
best tutorial by far
thank you.
Fantastic work! Recommending your channel to my discord.
@@rileyb3d thanks man! I appreciate it :)
FINALLY! I’ve been trying to do something like this for AGES!
@@loganasaurusrex8246 glad I could help :)
Thank you so much for making this video!
Your content is great and the way you explain things is so understandable.
@@galravid-tal8517 glad you liked it :)
Awesome! Will keep it when I create a new environment :)
great :)
Excellent bit of knowledge. Thanks a lot for sharing it!
@@alltradesjack6594 glad you liked it man :)
Thanks! This is a great explanation. Looking forward to de-tile my landscapes :)
glad it helped :)
Awesome!! IMO, definitely one of the BEST Texture Bombing methods I've seen so far!
Question: Based on your experience, I would be interested in knowing which Texture Bombing Method (say the 5 most common?) works best with which types of Textures.
For example, I think I read that Cell Bombing works best with granular-type textures (fine stone, dirt, cement, etc??.) and others work best for more defined textures (Ground Twigs/Leaves textures), etc.
Thanks again for the video!
I use cell bombing for any organic texture. rock, grass, ground with twigs, or anything.
But in case of textures that have pattern, like brick or tiled wood, the bombing methods don't work correctly.
@@HojDee - thanks. Would be nice if Epic eventually makes it a built-in Node!
quality tutorial man...
thanks!
Very cool, I use similar techniq in redshift. Thank you!
@@LexFomin you're welcome :)
Yeah, it's math stuff. It can be done in any engine I think
Алексей, привет! Прикольно встретить здесь своего приятеля. Такого рода инструменты есть много где , в короне например, но для анриала урок весьма полезный.
4:10 you don't need saturate node. The texture's Alpha output is already in the range 0.0 to 1.0
yeah. I just run everything I want to connect to the alpha of a lerp through a saturate. It's a habit I guess
Thank you for posting the video! I have a question.
Is there a way to reduce Texture Repetition for textures with patterns (e.g. bricks or tiled wood)?
If there is, I would like to ask if you can make a video!
good job
thanks
god bless youuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuuu 😭
thanks man
Thanks!!!
you're welcome man :)
Hi, Thanks for tut, I have problem when use displacement appear both texture cell , how to only make displacement for main texture?
I didn't exactly understand the question, but I think what you're talking about. If you increase the displacement magnitude, some areas may look bad, especially in the black areas of the alpha channel of the cell texture. I hide it by decreasing the displacement magnitude.
TextureVariation node is also useful... oO
yeah. I'll make a video about it in the future
Thanks a lot!
@@ilyachernobrov you're welcome man :)
what's the difference with using texture variation nodes?
I was about to ask the same. That node does the trick
In short: very little, the Texture Variation uses noise instead of cell & it blends the "visible lines" based on height rather than the same texture alone. The cost is similar but the result is cell can give more variation (without looking as broken). There is a little more to it, but that's the basics.
@@DamienRamirez thanks for the explanation
내가 학생 때 이 영상을 봤었다면 지금 더 좋은곳에서 일하고 있었을 거야..
I cant find the voronoi texture in the files...
there was some issue with the file. I've fixed it. download the textures file
Any notes on the performance? I noticed it samples the texture at least twice as much but I would be interested in the precise calculations and performance hits on a big terrain. It also needs to sample that voronoi texture but that doesnt have to have a huge resolution, right?
The most important thing is the increasing sample count, I think. I haven't gone deep into the exact performance calculations though. When using with landscape material, I'd use it only on layers where the repetition is really obvious and can't be hidden by foliage and stuff.
Does it work for an infinite world size? For example, if you set Size - Main Texture to 0.01, would it still avoid repetition?
You add an offset from the Voronoi texture, which has a fixed size, so I think it would repeat if the map is larger than the area covered by the Voronoi texture.
@@gone41674 the voronoi texture is also tiling. so it can go on forever. just make sure the relation between the texture size and the cell size is reasonable. like I mentioned in the considerations section of the video
@@HojDee Yes, I’m aware that the Voronoi texture is tiled, which is the problem. Tiling causes repetition, since the UV offset is read from the Voronoi texture, the final texture will also repeat. This repetition will occur if the world size exceeds the coverage of the Voronoi texture.
is there a way with simpler nodes to make random rotation? its to heavy the way you show it...
you can skip the rotation. or add a switch to turn it on or off. it works great with just the offset and the scale. or even just the offset
Ok, definitelly the way but should be as one material like a "plugin" with all parametres in it. I wait for something similar in 3DS max.
@@tomastoegel4973 in the end, I made it a material function. It can be used in any material inside unreal engine.
For 3ds max, the principles are the same I guess. If you're familiar with its material system you should be able to create it
I have a problem when I choose the modeling option, there are no tools
maybe the plugin is not enabled
Does this have some impact on the performance?
Yes, a bit, because of the indirect texture call that may invalidate the texture cache.
and also the increasing sampler count. the impact is not that much. but you gotta watch out for it. especially if developing for mobile
@@thutmosisSeven Thanks for the info!
@@HojDee Appreciate it!
I think the most important question is, how expensive is this method on the GPU?
it increases the texture sample count. makes it twice. so that's the real bottleneck, I think. it's not optimized in that sense
What is difference between cell bombing and texture bombing?
@@simonb1009 the method is different. cell bombing is cheaper and works better in my opinion
@@simonb1009 watch this video to learn more about texture bombing
ruclips.net/video/SfJY3e-3wwY/видео.htmlsi=v9Vs-QccW6OJlX7l
Hi! can t find cell texture. Can you give direct link? thanks
hi! get it here: www.hojdee.com/courses-details/free-downloads
and here's the tutorial on how to get it, if you run into any issues: ruclips.net/video/8yRxmuY9If4/видео.htmlsi=770df6CkfG-jx-Cj
Yeap, me neither, can you provide the link for your website ?
@@Cyberex there was some issue with the file. I've fixed it. download the textures file
Is there any way to download or buy your cell bombing MF file?
@@indigochoiart I've put it on my patreon. The link is in the description
@@indigochoiart it's also available in the epic's city sample project, if you have it.
@@HojDee thanks for the good work
@@indigochoiart glad to help :)
6:27 Why divide? Multiplication could be used to achieve the same result but with much cheaper instruction cost. Divide should be avoided where it's possible.
@@SirRebonack multiply is cheaper, you're correct. But with modern hardware, the difference is not that significant. If you wanna make it really optimized though, you should go with multiply for sure.
not sure how its handled here, but atleast in most programming languages, the compiler optimized most divisions into multiplications.
@@nikoheino3927 The shader compiler cannot optimize this division because it is based on dynamic data (sampled texture). Division is fine to use on parameter values, which are able to be optimized. Division is significantly more expensive than multiplication.
@@SirRebonack oh okay, good to know.
I think going through all the material is very time consuming. epic should make auto texture edit plugin.
@@anotherreality7951 you could just make it once and reuse it anywhere you want