For a project that took you only 24 hours this looks incredibly good. I can only imagine the amount of refinements that you could do if you spent a complete week on this. 🤔👏👍👍👍🔥🔥🔥
Depth of field effects and making the edges out of focus + adding dirt and scratches on the lens + adding a vignette + more erratic camera + motion blur + lens flares + a little bit of halation and grain would have took this to another level. but thank you for the in depth tutorial. This was really helpful to me.
depth of field does not make sense if you want this to look like its filmed with a real camera. erratic camera movement would look really out of place in a crane or helicopter type shot like this, same as dirt and scratches. going for a bit more of a vintage look could be done i guess, but all of this would be totally overdoing it in my opinion
trust me u dont wanna lower the aperture and blur the edge, it'll look like a miniature scale, and there is motion blur here, it's subtle, how much motion blur do u even want? ur comment is laughable 😂
Not to take away from what you managed to accomplish, because it is impressive, but the crowd software Weta designed actually had some basic A.I. that caused individuals to move toward enemies, and engage with them, etc. This crowd animation is more similar to what Disney did in Mulan, and originally for the background characters in the "Topsy-Turvy" number for The Hunchback of Notre-Dame, and it certainly surpasses that. Of course, if you'd added moths it would have brought everything to another level.
You know this guy is really good at what he does, iv fallen in love with blender after 500 hours , I'm not going to lie it's been one of the hardest tasks in my life to learn, I'm 38 now, and I honestly thought it was to late to learn I'm doing aabout 4 hours a day on average when I can get on, fantastic video pal, keep it up 😊
If it actually was done in 24 hours then by God this is amazing. If you spent a week or even a month you could actually make this look just as good as Return of the King.
Yes, It's possibile, he just download some stuff and put it togheter. (5 hrs of work) MAYBE he did the motion-capture by himself, the rest is just assembling a scene.
It actually looks really decent. For the volumetrics alternative setup, try using the camera's position information to drive a gradient that slowly makes the fog transparent as the camera approaches the fog planes instead of disappearing abruptly. This will add some depth to it
Everything here is so so good. I would echo other commenters in saying that randomizing flames would go a long way because the flame loop kinda stuck out in the final project. but i am so impressed, i could never pull this off
well done! it's truly amazing that anyone has criticism of the time and effort you put in to showing it can be done, and still there are people saying oh you should have done this or that etc. like how do people think at this stage it could be better, i remember not too long ago that people were saying it's not possible to have this scale in Blender, really you can't enjoy what is being demonstrated without complaining about something. Keep up the great work! 👏👏👏
It is INSANE that 1 dude can do in 24 hours what 1000 people took months to do 20 years ago. Granted, the Weta version still looks better, but considering their resources and time... this just blows my mind. I mean... it's unreal how good this turned out with 24 hours of work on open sourced software. The future is now. Amazing job. 🔥
Thats a commendable effort you have made and the color grading looks great. Very nice. I really want you to look on the randomization, as the motions sometimes looks too rhythmic.
Ngl at the start of the video I thought you were capping but bloody good work doin that in 24 hours you are one with the flow of productivity, locked in, respect
Amazing work, CG Geek, just a tip. For the ember particles, they seem to just disappear instantly once the lifetime is done. Maybe you could add a blend texture to your particles with strand/particles on, so that the particles sheink once their lifetime is done Love ur content btw❤❤
this looks awesome. However, the giants stood out too much, they doesn't seems to blend into the scene very well from the front view. Another nitpick I have is the armies should march around the giants instead of underneath them. Nevertheless, this is a great tutorial! Thanks for making it.
25 years ago 1500 employees made a professional and highly polished film production. Today one guy in his bedroom makes a comparison with an amateur production but still a good advertorial.
It was cool and all, but closer to previz than finished. You could easily see all the duplicated assets everywhere and there were several issues with the fire. Very fun though! I haven’t seen too many do crowd sims using Blender
That looks absolutely epic and I have the highest respect for that :,) But I think one thing, that would really give a nice touch to the scene would be some banners - They would add scale, verticallity and movement (there you could've used the cloth-animation) to the scene. And overall I think that banners just make armies look 10x as epic. But the scene looks great as it is!
You could have track the flames to an emptie and then track the emoty to the active camera. That way the flame would look to have more volume. Other trick is duolicate the plane and rotate it.
Amazing work for a 24 hour job! Although as some said Weta employed an AI system like Golem has for creating certain intelligence in the avatars created the scene you did shows how much technology has improved and using Free software! Some suggestions: I think using geometry nodes one can instruct the instances not only to move on terrain but also avoid obstacles or move towards an object (enemies). You could have also outputted a Zdepth pass and use DaVinci Resolve Fusion to add more spice like fog in post! The end colour grading the scene in DaVinci instead of doing a colour grade in Blender! This is my humble advice! Else Awesome work and kudos to you for sharing with us!
Turned out to be pretty amazing result . But you can enhance it a lot i think the background sky and mountains are too bright and you can add shadow plane in fromt blocking out the big characters light and enhance them by adding some rim light they are looking flat now not much contrast is there .Just my opinion
I am absoluteness dumbfounded Blender is playing the animation with thousands of orcs in real time 🤩 I know they have simplified rigs, but still, can we get a round of applause for the Blender developers? Wow!
Hey steve, what if just for fun, you try to do this in unreal engine 5, i mean with real time rendering, that would be crazy. And the suit would be +++++
It's absolutely epic and in 24 hours is nuts ofc. I just wonder if you could've given all the orcs slightly ofset animations from eachother through geometry nodes. That's the one thing that I think would've been the biggest improvement.
I'm not that experienced but it seems to me it should be easy to offset the torch flame animations. They play at the same time in the render and it's quite distracting. Overall, sick job. So cool what an individual can do nowadays.
I would suggest you take even more time to get a better end result. While it is impressive to do that much work in such a short amount of time the result is pretty janky and honestly not that good. In any case thank you for sharing.
I understand your criticism for sure. But I think his point was to see what he could do with very limited means in a very limited time frame, and see how it compares, if at all. Yes, of course more detail can be added, more time spent, more expensive computer(s) bought to process more data faster for all the extra detail. But I really don't think he's saying it's as good as a multi-million dollar production which took thousands of people a couple years. But it's pretty freakin awesome, all things considered. Being uncompromising with quality and detail, I would be interested in seeing one of your projects. Would you consider posting on, or sharing a link?
So cool! although the Weta shot had a lower angle for the camera. That helps sell the large scale. Of course the color grading also matters. Looks awesome though!
![[MV] A leap of faith...ความเชื่อใจครั้งสุดท้าย (From "Petrichor the series")](http://i.ytimg.com/vi/U1piZH2CNXA/mqdefault.jpg)
if you randomize the flames, it will greatly improve it. They animate exactly the same together and effects the illusion.
Could you randomize the animations of each uruk-hai?
@@johnleoridyep, just change the starting point to be randomised for each. You could use a driver for it
Hundred percent
randomize animation starting point
For a project that took you only 24 hours this looks incredibly good.
I can only imagine the amount of refinements that you could do if you spent a complete week on this. 🤔👏👍👍👍🔥🔥🔥
Depth of field effects and making the edges out of focus + adding dirt and scratches on the lens + adding a vignette + more erratic camera + motion blur + lens flares + a little bit of halation and grain would have took this to another level. but thank you for the in depth tutorial. This was really helpful to me.
ok Christopher nolan.
depth of field does not make sense if you want this to look like its filmed with a real camera. erratic camera movement would look really out of place in a crane or helicopter type shot like this, same as dirt and scratches. going for a bit more of a vintage look could be done i guess, but all of this would be totally overdoing it in my opinion
trust me u dont wanna lower the aperture and blur the edge, it'll look like a miniature scale, and there is motion blur here, it's subtle, how much motion blur do u even want? ur comment is laughable 😂
here's a challenge, you do it; you replicate this, disney or Weta; with Blender in 24hrs and show us all how you would implement it!
Depth of field would make the scene look small, you can't have depth of field on aerial shots unless you fake it
We need more of these large scale renders!
Not to take away from what you managed to accomplish, because it is impressive, but the crowd software Weta designed actually had some basic A.I. that caused individuals to move toward enemies, and engage with them, etc. This crowd animation is more similar to what Disney did in Mulan, and originally for the background characters in the "Topsy-Turvy" number for The Hunchback of Notre-Dame, and it certainly surpasses that. Of course, if you'd added moths it would have brought everything to another level.
Moths are life
Agreed. Maybe should have grabbed Unreal 5 to render this one...
@@ghosteclipz7007 **proceed to add moth on a space scene**
here's a challenge, you do it; you replicate this, disney or Weta; with Blender in 24hrs and show us all how you would implement it!
@@GaryParris Alright, and here's a challenge for you: re-read my comment, and this time try to comprehend it.
You know this guy is really good at what he does, iv fallen in love with blender after 500 hours , I'm not going to lie it's been one of the hardest tasks in my life to learn, I'm 38 now, and I honestly thought it was to late to learn I'm doing aabout 4 hours a day on average when I can get on, fantastic video pal, keep it up 😊
keep it up, its worth it, even if its just for personal growth!
Dude, amazing for 24 hours. Nice to find all those assets.
If it actually was done in 24 hours then by God this is amazing. If you spent a week or even a month you could actually make this look just as good as Return of the King.
not a chance
I say, let him do it. Heck I’d even subscribe to his channel
Yes, It's possibile, he just download some stuff and put it togheter. (5 hrs of work)
MAYBE he did the motion-capture by himself, the rest is just assembling a scene.
"You can do this for free!"
*grabs a $5000 motion capture set*
It's about half of that if you qualify for the Indie Bundle he talks about but fair enough 😅
U can still do it for free
I mean you can use Mixamo instead of using expensive motion capture. He just had to do it because of sponsor.
@@RokokoMotionor Rokoko video!
@@RokokoMotion or Animate it yourself(dreaming about Rokoko suit)
It actually looks really decent. For the volumetrics alternative setup, try using the camera's position information to drive a gradient that slowly makes the fog transparent as the camera approaches the fog planes instead of disappearing abruptly. This will add some depth to it
I did this in a sense, by using the Mist pass to add some fog in the compositor as well.
@@CGGeek Noice, I did notice it. Great work
Everything here is so so good. I would echo other commenters in saying that randomizing flames would go a long way because the flame loop kinda stuck out in the final project. but i am so impressed, i could never pull this off
well done! it's truly amazing that anyone has criticism of the time and effort you put in to showing it can be done, and still there are people saying oh you should have done this or that etc. like how do people think at this stage it could be better, i remember not too long ago that people were saying it's not possible to have this scale in Blender, really you can't enjoy what is being demonstrated without complaining about something. Keep up the great work! 👏👏👏
Great video Steve.
Thanks mate! appreciate it
These guys from Oasis even know Blender. Respect!
11:15 we can see your animation of the guy tripping in the middle of the battlefield.
Wow!!! What a Creative Process!!! Thanks for Sharing!!! 👍😎
Amazing to think you can do this in little more than a couple of days and relatively basic tools nowadays.
Super cool job for only 24 hours! Would love to see it again when you can spend more time on it
this video literally amazed me its so impressive keep it up . like if u give week or months to this output will be really high budget periodic movie
Color grading left the chat☠️ ,all the work was fantastic 🔥
It is INSANE that 1 dude can do in 24 hours what 1000 people took months to do 20 years ago. Granted, the Weta version still looks better, but considering their resources and time... this just blows my mind. I mean... it's unreal how good this turned out with 24 hours of work on open sourced software. The future is now. Amazing job. 🔥
Bro, you could have give that Troll some love on the left arm :D
This is so awesome dude
Thats a commendable effort you have made and the color grading looks great. Very nice.
I really want you to look on the randomization, as the motions sometimes looks too rhythmic.
This is bonkers. Good job!
Ngl at the start of the video I thought you were capping but bloody good work doin that in 24 hours you are one with the flow of productivity, locked in, respect
You’re genius man
nice for 1 day's work.. needed to vary the torches animation.
spotted the uruk hai next to the troll at 11:14
The geonode technic is so smart, good job
That is really impressive! It doesn't look as good as LOTR, but you're just one dude!! Keep up the good work!
Should do bunch of variations of trolls, cuz they are big and easier to notice the single variation of their movements
can you do more of this kind of stuff love the lotr CGI effects and large scale armies
CGgeek is such a god... the GOAT even...
Thanks so much mate! We all know Ian Hubs is the GOAT ;) but appreciate the compliment! Glad you enjoyed the vid
This was sick
Amazing work, CG Geek, just a tip.
For the ember particles, they seem to just disappear instantly once the lifetime is done. Maybe you could add a blend texture to your particles with strand/particles on, so that the particles sheink once their lifetime is done
Love ur content btw❤❤
Liked the video. When you said you put twelve hours in I figured that's what making good clips is - time and knowledge. I'm very interested.
dude thats amazing, its magic
Really nice... I feel like it would have come out better if you made some offset to the flame touch and brought them a bit closer
Final shot looks epic!
this looks awesome. However, the giants stood out too much, they doesn't seems to blend into the scene very well from the front view. Another nitpick I have is the armies should march around the giants instead of underneath them. Nevertheless, this is a great tutorial! Thanks for making it.
@CGGeek Thanks mate, totally awesome content. The free assets on the pateron make it even better.
Keep up the grind bro 💯
Your videos are theraputic. It's crazy that you did it in 24 hours, amazing 🧿
thx so much! great to hear
great job dude. Very impressive.
Really cool exercise, very well done!
Amazing job!!
nice work
Impressive work!
Very cool project und result.
ayo that was amazing. i would not be able to make this in a month or so. damn bro. amazing work
Great breakdown!
Looks great.
Fantastic work Bro
This is AWESOME!!!!
25 years ago 1500 employees made a professional and highly polished film production. Today one guy in his bedroom makes a comparison with an amateur production but still a good advertorial.
Thumbnail is epic. ❤
It was cool and all, but closer to previz than finished. You could easily see all the duplicated assets everywhere and there were several issues with the fire. Very fun though! I haven’t seen too many do crowd sims using Blender
11:14 , DOING THE WORM CLOSE TO THE CAVE TROLL ON THE LEFT
That looks absolutely epic and I have the highest respect for that :,) But I think one thing, that would really give a nice touch to the scene would be some banners - They would add scale, verticallity and movement (there you could've used the cloth-animation) to the scene. And overall I think that banners just make armies look 10x as epic. But the scene looks great as it is!
Awesome dude. You're a blender legend.
I love that you pronounce Rokoko the same way you'd pronounce Sudoku despite it not having any Us in it
I think there was like one time he said it right.
thanks to rukuku for sponsoring this video
You could have track the flames to an emptie and then track the emoty to the active camera. That way the flame would look to have more volume. Other trick is duolicate the plane and rotate it.
Very nice man I love it.
I see new CG Geek Video, I click CG Geek Video
Amazing work for a 24 hour job! Although as some said Weta employed an AI system like Golem has for creating certain intelligence in the avatars created the scene you did shows how much technology has improved and using Free software! Some suggestions: I think using geometry nodes one can instruct the instances not only to move on terrain but also avoid obstacles or move towards an object (enemies). You could have also outputted a Zdepth pass and use DaVinci Resolve Fusion to add more spice like fog in post! The end colour grading the scene in DaVinci instead of doing a colour grade in Blender! This is my humble advice! Else Awesome work and kudos to you for sharing with us!
Wd nailed it, they deserve every worth of Penny,
great video
for 24 hours this was awesome dude!!
Turned out to be pretty amazing result . But you can enhance it a lot i think the background sky and mountains are too bright and you can add shadow plane in fromt blocking out the big characters light and enhance them by adding some rim light they are looking flat now not much contrast is there .Just my opinion
Awesome tutorial, keep up the amazing content
Man what a fantastic job you've done... how bad I want to learn this kind of stuff 🥺🤩
I don't understand why TH-cam doesn't recommend your latest uploads. I'm subscribed and my notifications are on.
Crowd simulation can be someone's niche expertise in the 3d animation industry. I'd love to see a demo reel of that guy.
I am absoluteness dumbfounded Blender is playing the animation with thousands of orcs in real time 🤩
I know they have simplified rigs, but still, can we get a round of applause for the Blender developers? Wow!
Great work
ggreat workkkk, just wow. A one man team
WELL DONE
the fact that we've come so far that we can do mocap at home now 😆
amazing bro
Would well as an advanced pre-viz!
Emitter of fire 🔥 sprites is the same time loop in final shoot, do it in random to more realistic
U are amazing 🔥
Amazing Bro!!!!
the torches are all in sync it takes away from the final result
Adding a tiny bit of blur and grain in post, it'll get more photorealistic.
Just.....thank you 😊
Good work!
The things that actually missing on TH-cam is two armies fighting. Like actually running up to each other and start fighting.
For 24 hours by one person this is insanity
Nice work.
Hey steve, what if just for fun, you try to do this in unreal engine 5, i mean with real time rendering, that would be crazy. And the suit would be +++++
It's absolutely epic and in 24 hours is nuts ofc. I just wonder if you could've given all the orcs slightly ofset animations from eachother through geometry nodes. That's the one thing that I think would've been the biggest improvement.
I think it's awesome.
sooo much work.. super fun
I'm not that experienced but it seems to me it should be easy to offset the torch flame animations. They play at the same time in the render and it's quite distracting. Overall, sick job. So cool what an individual can do nowadays.
This is amazing
Very good video 👏
I would suggest you take even more time to get a better end result. While it is impressive to do that much work in such a short amount of time the result is pretty janky and honestly not that good. In any case thank you for sharing.
I understand your criticism for sure. But I think his point was to see what he could do with very limited means in a very limited time frame, and see how it compares, if at all. Yes, of course more detail can be added, more time spent, more expensive computer(s) bought to process more data faster for all the extra detail. But I really don't think he's saying it's as good as a multi-million dollar production which took thousands of people a couple years. But it's pretty freakin awesome, all things considered. Being uncompromising with quality and detail, I would be interested in seeing one of your projects. Would you consider posting on, or sharing a link?
Make a longer in depth video on this please!
So cool! although the Weta shot had a lower angle for the camera. That helps sell the large scale. Of course the color grading also matters. Looks awesome though!