Meet Meline : 8 Years later vs. Agent 327 (and other animated shorts)

General Forums Blender and CG Discussions
1

First off let me say I really liked Agent 327, I only hope one day to have that quality of work. The entire team did a great job!

I do not want this to sound like I am bashing any animated short, these are serious questions.

What are the technical differences between Meet Meline and Agent 327? (i.e. Meet Meline didn’t use SSS, etc… !?)

How were two 10+ year old computers able to render out a film twice as long as Agent 327 and still look equally impressive?

I am not saying one film is better than the other, they both are beautiful in there unique ways, but one is almost a decade old made by 2 people (other than music).

How did they do this technically? It took a team of talented people and I am sure a render farm for Agent 327 how do 2 people create something that is on par with Agent 327 with computers over 10 years old? what "cheating"or techniques did they do to save so much time rendering?

I dont know how much Agent 327 cost to make but $18 million for a feature seems a little much, considering 2 people did Meet Meline for practically free (other than the cost of 2 years of their time.) Is this $18 million including a marketing budget?

I guess I am asking, what are all the technical differences between Agent 327 and Meet Meline?

Again sorry if it seems I am bashing on Agent 327 (its just because its new and got me thinking.)

I think the movie is gorgeous and very well done!

I saw the behind the scenes of Meet Meline, but it didn’t answer all my questions.

Meet Meline https://www.youtube.com/watch?v=Y1vdZIIrY6w

Agent 327 https://agent327.com/

2

Thats because they look equally impressive to you. :wink:
Very subjective.

In my opinion they are worlds apart to begin with.
For an animation done in 2010, Meline is absolutely fantastic ! - graphically (scene dressing, textures…).
A tremendous effort for 2 people !

But there is much more to an animation short then just the graphics.
Screenwriting would be one…

Meline might be the most meaningless set of (awkward) animations I have seen in a while. :slight_smile:
But the renders are nice.

3

Come on man use some common sense nothing is for free. That Meline shot absolutely cost money to make you’re at just bad at following the $$$ and seeing the hidden costs. There is no such thing as a free lunch.

4

Meet meline used no modern rendering tech, just straightforward old school lighting. Plus they rendered it at 480p. That’s how they rendered it on 2 computers.

As for 2 people getting it done, Meet Meline was much smaller scale than Agent 327. It has fantastic art direction and design, but only one character, simple layout, and a small (but detailed) set.
It’s a good example of keeping tight control of the scope of your film, so it can be achieved with minimal resources.
Agent 327 is also a good example of this, just with a larger scope.

5

It also look like Meet Meline is heavily composited / graded to give a good look and feel but also to maybe to get rid of the clean look of CG.
I guess if you look at raw renders they are way below what’s done on Agent 327. I guess also if you turn these renders to HD you’ll see a big difference with the quality in Agent 237.
The camera work is also good and give a “real” feeling to the movie.
I also agree with rbx775 that the animation is not on part with the image quality.

6

I understand there are some cost involved, but it is negligible.

From the website "“Meet Meline” is a 3D animated short film created independently and without any budget by Virginie Goyons and Sebastien Laban "

7

Ahh!! That was a big part of it. I just assumed an HD version was floating around, but nope it was 480p originally.

8

Gah, where do I start?!..

Back then we upscaled and roto’d to get higher rez. We could quadruple the rez and then just spot fix some of the resulting artifacts. There were a few tools also to automate the process. So render resolution is not an issue or relevant. I still upscale almost all the time now days from 720p to 1080p. Only when the content requires 1080p do I generate it natively to begin with. Bottom line, there’s always tricks that the DCC peeps use and it’s driven by what the producers want and who their audience is. So that’s a “moo” point as Joey would say because you are just trading one artifact for another when you put render rez under the microscope and stare at it for longer than a viewer ever would or even care about. Also we used compositing heavily back 10 years ago because of limited resources. Purists now days and Blender official projects tend to be more purist and do full frame rendering rather than partial with a final composite. I’ve brought this up in person with the people involved on a few different occasions and in IRC several times. This is a matter of choice and if you can handle the organization and workflow of layering, partial renders and compositing.

So that’s a segue to what angle are we looking at here for “quality” or whatever your preferred word is. From an artist or technical standpoint or from a viewer standpoint different assessments of quality can be barfed out by anyone. It’s still a “moo” point because perspective is subjective and an endless fractal conversation. (Lol, I just made that up in context and I like it. Inspired by BA concern trolls! I think I’ll keep that descriptor! ; )

Now from strictly a viewer standpoint and in consideration of my perception of the initial question, yes, blender projects are very expensive. ToS cost ~300k euros including developer costs IIRC and it was considered incomplete from what I remember hearing from some of the people directly involved. The following year blender was used in feature length film as the primary VFX software and that whole film cost $500k to make with a couple Hollywood B list stars. This was something regularly mentioned here in Hollywood and was considered a factor by many here that I regularly talk to in not supporting Cosmos when they did donate to ToS. “I donated to ToS and look what we got in comparison to what we produce here in town so I’m not throwing my money away on Cosmos.” Of course this is very biased by the fact it’s from my circle of peeps.

But then again, who cares? Either you like supporting blender and it’s primary org on their terms or you don’t. This was hotly debated here on Troll Town, BA when Cosmos funding versus game feature funding was discussed in an endless fractal conversation thread. ; ) So considering the two productions this thread is about, return on investment(ROI) is definitely dramatically different in the two production. It could be argued that you have less investment in the old animation and almost as much of a return as the heavily invested Blender Studio(not an open source/CC project) and request for future investment. The key here I think is what is the return you want on your investment. Blender open movies and projects have a more complicated return that is expected and it is not just the resulting media production. So to simplify the ROI of an official Blender project is to be in error in the first place IMO and from my perspective. In other words, as has been stated many times before, you are paying for a lot more than just the end result when you fund an official blender project.

Meanwhile back at the ranch with Lassie because the barn is burning, there is a movement to provide a more Meet Meline ROI in general blender production by independent teams and studios. More on that another time in the near future. Just keep watching Blender Nation and in the not to distant future you will see exactly what some of us have been working on that has a “lean production” and higher ROI as I like to call it.

9

So you´re saying that the time they put into making the short wasn´t worth anything? They might not have taken out any pay, but they either spent a lot of time they could have spent on doing paid work, or they´d already saved up money to be able t work on it.

10

I agree 100% that was not the point. I was responding to another member with that quote. I was talking strictly MONEY! In my above post I even mention the time it takes. They both had full time jobs and worked on Meline on their spare time for 2 years. Of course its worth a whole bunch.

11

Interesting, I will keep an eye out.

12

This just goes to show that many people actually can’t tell the quality difference between path-traced physically-based rendering and the “old workflow” (i.e. tons of hacks). Kinda sad, but there we go.

The difference in computation time can be entirely attributed to the use of path tracing. It has been around since the seventies, but it has been prohibitively expensive up until recently. It’s an extremely inefficient approach and it doesn’t necessarily give better quality than other approaches. What it does give is predictable quality and instant previews. You don’t get splotchy artifacts that jump from frame-to-frame or annoying pre-processing steps. It works with arbitrary geometry - and large amounts of it. It’s all about the workflow.

Another thing to keep in mind is Parkinson’s Law:

“Work expands so as to fill the time available for its completion”

The whole Agent 327 short could’ve been done in half the time, as well. It wouldn’t have been as good, but again, a lot of people wouldn’t even notice. The artists do notice, however - and they’re the ones doing the work.

13

No it’s not negligible I repeat again you are not seeing the hidden costs. Every day when you are at work you are generating an income for your self, your hourly rate, when you are not working you are burning through money, your personal burn rate. Take how many hours they spent working on this project and multiply it by their burn rate and you will get how much it cost to make this movie.

If they didn’t save up any money for this movie or had an external source of financing like a kickstarter then you bet your ass they had to keep working at another job to ensure that they could work on this movie.

Nobody is going to be making a short movie when the don’t have a roof over their head, food in the kitchen, electricity, clothes on their back.

For something like Agent 327 or a Pixar movie what is basically happening is the employer is indirectly paying for the artists living expenses in exchange for their labour.

So tell me again how the expenses on this movie were negligible?

14

Well theres way different technic evolved.

  • the low Resolution with heavy post process treatment help meline, despite the low quality of some animation, textures.

  • the artistic direction… if the overall quality is cartoonish, the face is aiming at more realism ( it is the eyes and reflection on it who bring this “realistic touch” ( largely inspired by Final Fantasy CG it seems ). Now, 327 is based on a comic book, and all is aspect is coming from it.

  • Again the artistic direction and approach are so different… Need to say that Meline artists was really good too, it was really well realized, well written, artistic line was spot on. But in term of technical we are nowhere near.
    In fact, i ask me how will look Meline if it was done now.

15

I the girl talked then it would take 4 years to make.

How were two 10+ year old computers able to render out a film twice as long as Agent 327 and still look equally impressive?

No. It’s just in your head. It’s day and night between the two. But the art direction could be equally impressive.

16

https://www.youtube.com/watch?v=BI-dzAdHHAA is real time even. As Beer Baron said, it’s about workflow. Several high end GPUs or CPUs are much cheaper than artists time.

17

I didn’t say I don’t see a difference. I said they are equally impressive. I wouldn’t say one is “better” than the other, just different. I was mainly asking what technically are the differences between the two (which you answered). As someone new to the 3D world I am trying to develop an “eye” to specifically what I am seeing.

What makes Agent 327 look so clean and eye pleasing? is it Filmic, HDRI’s, SSS, Raytracing, all of the above? (even compared to Cosmos Laundromat (which now watching it after 327 looks harsh and not as eye pleasing, but still very good)).

I am just trying to understand the technical differences between all the approaches used in these films.

18

The creator of the animated short film “Pidgeon: Impossible” did a very good video on what he calls ‘the production triangle’

Meet Meline might have been done on no significant budget and it does look great so my question is, how long did it take them to make it? Probably a couple of years at least which is where the “hidden costs” that @tyrant monkey talks about.

19

There is also a huge difference, between those 2 movies.

Agent 327 is an existing comic, its a 2D comic not 3D.
From an artistic standpoint, it is ‘production’ art,
Its known how the character should behave and look, and his personality (ea funny agresive anxious etc)
But it is not known how that would look like in 3D or how to animate that… thats where animators en 3d artist get in.
They try to give an experience of that 2D figure… and yes i think they nailed it.

Unlike the other movie where everything could be original, and everything would be good.
A to fat 327 would not work, or to skinny or to worried, or etc etc…

So agent 327 is great in that respect, a great example of what blender can do for artist.
This time it does not seams to be all around some new kind of feature in our blender program.
This is what animators, and creators of meshes can make, and it feels all correct.

Meet Meline looks good, but it had no style to look alike.
Also staying at a style is easier if it is your own style.
But it would be harder if you had to create something in someone’s else style.

Meet Meline isnt bad either, but comparing them is like apple and peers.

Though maybe the only thing is that perhaps it could have been done earlier agent 327…
Cause this time i think it depends less on new features.

(maybe there is some new future for clothing used in agent 327 but im not sure about it, and well that could have animated to in older version.).

20

The “project triangle” is bullshit, it only exists because project managers really like to draw things.

In most projects, you can’t arbitrarily trade off time for cost. You can pay your workers a premium for overtime (crunch) and in effect they’ll be more tired and less productive. To a first approximation, time spent equals cost. You can and must trade off quality for time/cost, because of finite budgets and diminishing returns.

These personal projects play by slightly different rules. Without a financier setting boundaries on a budget, they can go on almost indefinitely. Most of these projects never finish, but then you also never hear about them. When such a project does finish, it is actually remarkable.

What makes Agent 327 look so clean and eye pleasing? is it Filmic, HDRI’s, SSS, Raytracing, all of the above? (even compared to Cosmos Laundromat (which now watching it after 327 looks harsh and not as eye pleasing, but still very good)).

95% of it is the actual artist’s work. The technology may enable or ease some things, but it doesn’t work all by itself. If Cosmos Laundromat looks more harsh to you, some of that may be down to Filmic tonemapping in Agent 327, which is more muted.