League of Legends fan film, shot on a7S

Samuel H

Samuel H

Major Contributor
I get a lot more respect in these forums than I deserve. I know technical stuff and people assume I know how to shoot awesome videos as well. Let me prove I'm just a newbie learning his way by posting here the result of my biggest project to date. I'm very proud of it, I think I did a really good job, but it's far from what others in this forum can make. And no, the excuse that I shot this on a single day as a one-man-band with available lighting is not a full excuse, since I wouldn't know what to do with more time and actual lights on set (I did know how to scout the locations to find the time of day when we should be shooting at each place, to have reasonably good lighting without having to worry about it on-set).

Since we had a nice location, I went for a deep-DoF look for all the shots except the closeups. Very happy with the result regarding that.
On the day of the shoot, the sky turned out ugly white, fully overcast. So, nice diffused light, but I let the sky blow out in many shots because it was white anyway. It doesn't seem to have bothered anybody yet.

Shot on the a7S, internally, with vintage Leitz Elmarit-R 24mm f/2.8, Sigma 19mm f/2.8 and 30mm f/2.8, and vintage Leitz Summilux-R 50mm f/1.4. I used the 24mm in FF mode on a flycam nano, the 19 and 30mm in APS-C mode on a Nebula 4000, and the 50 on APS-C mode as well (most of those closeups are slowmotion) but on a MagicRig. Hoya ProND filters. Color corrected in Colorista 3, color graded in Looks 3 (I went for a super-saturated look because the short is based on a videogame and this look fits the theme). I did all the VFX myself as well. 60 shots. Some turned out great, others look a bit cheesy, but well, still learning :)

I planned the 19 and 30 mm in for a lot of shots because, even if they're not my best lenses, they have AF: I used prefocus-then-shoot for most of the shots because it's fast and relatively foolproof.

For future projects my head is struggling between "the a7S II looks awesome in paper" and "I'll probably benefit more from having more time to direct my friends-ahem-actors and a bit more control over the environment". Nothing new, I guess...

ANY COMMENTS WELCOME :)
(the target audience already knows who these characters are, who's friends with who, etc; it may be a bit more confusing to you...)

https://www.youtube.com/watch?v=7-P1HM5j1iU&index=1&list=PLXfQ_R9rAMS5K0qx8OOqpL-n_uRULcjB5

 
Last edited:
Great colors man! Style is very fitting. And as for the effects: they're certainly better than anything I could pull off, and seem to match the source material just fine.
If I had any suggestions it would be that there are some good editing tricks that help hits feel like they land harder (like cutting out 2 frames right before hits land).
Good work. (also your video in 2 days already has pretty much the same views as all of my videos on youtube! :) )
 
Last edited:
Haha, thanks :)
The views so far are mostly from my loyal followers. Whether it's a success or not in terms of views will depend on whether they share it and we manage to reach a wider audience.

I aggree that some of the punches feel weight-less. That's a problem in all my fight videos, actually. At least on this one I managed to make a few of them work (mostly the ones at the end).
I know about the cut-out-a-frame trick but I didn't use it here. In part because I don't mind the look of those punches too much in this case (all hits in the game are also weight-less), and in part because I was always feeling that it wouldn't work, given how I shot them in the first place. Maybe I could have tested it at least...
The definitive solution, in any case, would be to improve as a director and manage to get them performing better from the beginning (by the end of the day they were more confortable with what they were doing and I was better at making myself understood, that's probably why the last punches work better).
 
Fit's with the style! I don't know the CosPlay world at all, but it looks like fun, which is what i guess it's all about.

If you had time i'd have worked with the frames a bit more, darkening areas to draw attention and also the energy effects look like they've been 'screen'ed on (my bugbear about linear compositing) and some don't look tracked into place. And time remapping can help with editing energy. But then again, it's probably all in-line with the whole genre, right?

So is this based on a game then? (oops, it is, i just reread the post, doh)
cheers
Paul
 
Thanks! It was a blast, the cosplayers absolutely loved the day, and now love the result. Which is great because they'll want to work with me again (they're the best costume and makeup team I could ever afford).

I take note about the secondaries. I did apply a bunch of them, but I haven't been using them for a long time and tend to fall on the short side. Thanks.
The effects are indeed screened on. It's precisely how I composited them. In this case there's no way out of this problem because that's all I know so far. Still learning my way around VFX...
 
Looks good, especially for a one day shoot. A few things to consider.
1. Swish track over the punches, some seem to be missing something, a grunt pass when the people get hit. Also throw punches is combinations, it seems as if most attacks are a single punch followed by a single block.
2. More coverage, Extreme close ups, hands on weapons,
3. Using a longer focal length with compress the image and give it a closer feel.
4. Faster cuts would help the pacing, it seems a little slow. Sometimes it works well to cut out a frame or two when a punch makes contact, speeds it up and makes it more intense.
5. The main score seems to get redundant, maybe mix it with something for different scenes.

Overall it was decent, I think you have a lot of potential, just keep shooting stuff like this and each new project will be better than the last.
 
^ SUPER USEFUL THANKS A LOT

Together with not-enough-secondaries, not-enough-coverage-and-variety-of-shots is something I have noticed I do, but it's also something I find difficult to get rid of. I'll try to fight harder on these fronts.
 
Samuel H said:
Thanks! It was a blast, the cosplayers absolutely loved the day, and now love the result.
Click to expand...


And this is what matters in the end on a shoot like this!

As people have said some of the punches feel weightless but stunt fighting and can be tough unless the actors are experienced (plus I could imagine everyone wanted to not destroy their cosplay as I have friends who do it and it takes a ton of work). My only other comment to look at was the pacing. There were a few points (mostly the end stare down) where it felt like it was dragging a little bit. However I dont know how you would fix that since you wanted to show off all the characters and there were just so many (which makes sense because its league) so in this situation it was probably fine.

And lastly (im a big offender of this ) was the sound design feeling empty at times, but come on sound design is a whole job in its self haha. Great job was great to see some of your work by the way.
 
Hey Sam - Awesome film! The comments above about editing the fight scenes brought to mind Tony Zhou's piece on Jackie Chan who says, "the editing is the most important part." Interesting comparison of HK and Hollywood editing styles at 5:00, https://vimeo.com/113439313.

Best,
Andrew
 
@ReelFilmGeek
Yes, we did this just for fun, so it was a success from the start. Then again, I want to get better, so, thanks everybody for all the feedback!!

@baklava
I love that piece about Jackie Chan. Not easy to imitate, though, because of the "he gets hurt" part!! RackaRacka get that perfectly, but I couldn't shoot like that, I'm a "safety first" kind of guy.
About the rithm, completely right, it's not working smoothly in my peace. I'll try to shoot a longer fight scene at some point, to see if I can get better at this.
 
THANK YOU EVERYBODY

This is what I take so far as the main points that need improvement:
* punches need more weight: more elaborate choreography with combinations of punches, use longer lenses, better angles, take frames out (and I add: spend more time on set practicing the choreography)
* more coverage and variety of shots: extreme close ups, hands on weapons, etc
* rithm in the edit, in general (feels a bit slow, specially in the final stand-off)
* rithm in the edit, in fights (also helps with the "punches need more weight" issue)
* work more on secondaries: darken areas to draw attention
* work more on VFX: some look "screened"
* work more on sound design: it feels a bit empty at times
* make it shorter so music doesn't start to feel repetitive (or use two songs, which I know would be more difficult for me)

I'll also add "work more on lighting", because I know it's my biggest shortcoming (with such tight schedule, here it wasn't even an option).

I haven't set my mind completely yet, but I'm thinking about the following: I'm going to put my a7S on ebay, and also the N4000 (I'm already over the weight limit and new a7 bodies are heavier); I'm going to get an RX100 IV, and a smaller gimbal, and I'm going to shoot with those for 12 to 15 months. This would allow me to focus on these more important things and put my mind away from changing lenses and toying with gear, as I patiently wait for the next generation to arrive (most of my projects have a lot of slow motion, so having 4K at 24p but not at 60p makes it a lot less useful). All this is easier said than done, but for what I have planned for the next 6-8 months an RX100 IV would be enough, and who knows what the following 6 will bring. I'll definitely miss working with my lovely vintage Leitz glass, but I'll reduce the depreciation hit, which is a nice bonus.
 
Samuel H said:
I haven't set my mind completely yet, but I'm thinking about the following: I'm going to put my a7S on ebay, and also the N4000 (I'm already over the weight limit and new a7 bodies are heavier); I'm going to get an RX100 IV, and a smaller gimbal, and I'm going to shoot with those for 12 to 15 months. This would allow me to focus on these more important things and put my mind away from changing lenses and toying with gear, as I patiently wait for the next generation to arrive (most of my projects have a lot of slow motion, so having 4K at 24p but not at 60p makes it a lot less useful). All this is easier said than done, but for what I have planned for the next 6-8 months an RX100 IV would be enough, and who knows what the following 6 will bring. I'll definitely miss working with my lovely vintage Leitz glass, but I'll reduce the depreciation hit, which is a nice bonus.
Click to expand...

Really? The RX100 nice though it is is super noisy from the samples i've seen. I'm sure there's an A7sII with your name on it :)

Will most of your shoots be outdoors like that? If so then light modifiers would be a useful addition. I use frames from Lastolite which is a UK company and may be cheaper than matthews in the US. They do 1x1m and 2x1 frames that break down and join together. You can get diffusion, reflection (white only - never metallic because it will reflect specular) and also black. Don't underestimate the power of the dark side (of frames). Sucking light out is just as important as pushing it in. Just using modifiers can be a great way to hone those skills too.

If you're looking for lights be very careful if you're going down the LED route. Some are awful but the last generation has made great leaps. The quality of Aladdin lights and also the new LightStorm ones seem very good. Remote Phosphor is where it's at but that's really expensive still.

And Leitz glass doesn't depreciate. If anything the reverse is true, i'd hold onto glass like that if i were you.

As for VFX. That's all about workflow and pipeline. Light is very simple in essence, but getting your footage into scene linear in order to comp properly is the challenge. Shooting log is ideal. And then a reverse process at the end of the chain to bring your working footage back into delivery range.

When working in scene linear, adding light is just adding math. Motion blur works correctly and DOF looks correct. Working in any other way makes comping just look wrong, or a lot harder to look right...

cheers
Paul
 
* The RX100 IV is nowhere close to my a7S, but over these 12-15 months the a7S will depreciate, the a7S II will depreciate even more, the RX100 IV also but it's cheaper so in dollar terms it won't take such a big hit, probably. Plus the RX100 is a different thing, something I'd probably like to have even if I had an a7S II to begin with. Because it's tiny, and because of the frame rates.
* I'm not selling my glass, that's for sure. It would only be the body and the stabilizer, the glass would be stored and maybe used from time to time on the NEX-5N that I still keep (bought it cheap, it got even cheaper, it didn't make much sense sellling it).
* I have a set of tungsten fresnels, and I'm making my own LED fresnels, using 100W CREE bulbs with CRI95 and optical-grade glass (tis LoL video put that project on hold for four months, I'll try to retake it this autumn). But on this short I was shooting outdoors and alone, I couldn't even use the reflectors and diffusers I also have!
* I always work in a linear space, long-time Stu Maschwitz follower :)
 
(My camera is on Ebay, with a high starting price. If it's sold, I'll run this "learning the important stuff experiment" and I'll shoot with an RX100 IV for the next 12-18 months, which is my expectation of "until the a7S II can be bought for $2K, or until an a7 body comes out with clean 4K60 and slog3".)
 
Samuel H said:
* I always work in a linear space, long-time Stu Maschwitz follower :)
Click to expand...

What workflow are you using? Out of curiosity...?

Samuel H said:
(My camera is on Ebay, with a high starting price. If it's sold, I'll run this "learning the important stuff experiment" and I'll shoot with an RX100 IV for the next 12-18 months, which is my expectation of "until the a7S II can be bought for $2K, or until an a7 body comes out with clean 4K60 and slog3".)
Click to expand...

The best camera is the one you have with you, can't argue with that!

cheers
Paul
 
paulcurtis said:
What workflow are you using? Out of curiosity...?
Click to expand...

Right now, with how good color work is directly in Premiere and how well dynamic link works...
- Cut in Premiere, get my edit lock there
- Any shot that has VFX gets a "right click - substitute by AE composition" treatment, then I work in AE on a 32-bit linear workspace, when it's done I save the file and that's it, work goes on in Premiere
- (for this project I had 60 VFX shots, so I actually have six AE files with 10 shots each; having 60 files would be silly, andputting too much stuff into one single file gets confusing and can bring trouble)
- Color correction for each clip in Premiere using Curves for exposure, Fast Color Corrector for white balance, and Colorista III for saturation and secondary corrections
- Color grade for the whole clip in Premiere using an adjustment layer with Colorista III and Looks 3 (often requires tweaking the color correction step for some individual clips)
- When everything is done, export from Premiere as H.264 with very high bitrate
- Re-encode to normal-bitrate H.264 in MKV wrapper that plays absolutely anywhere using Handbrake (reasoning: "pirates surely know how to deal with this > research > wow this actually works great, it does play everywhere")

Stu has also s*%tfed back from the "no mastering from your NLE", Premiere is good enough for most work right now. The 32-bit linear workspace thing remains, though (unless you use that all the overlayed stuff tends to look weird).
 
Samuel H said:
Right now, with how good color work is directly in Premiere and how well dynamic link works...
- Cut in Premiere, get my edit lock there
- Any shot that has VFX gets a "right click - substitute by AE composition" treatment, then I work in AE on a 32-bit linear workspace, when it's done I save the file and that's it, work goes on in Premiere
- (for this project I had 60 VFX shots, so I actually have six AE files with 10 shots each; having 60 files would be silly, andputting too much stuff into one single file gets confusing and can bring trouble)
- Color correction for each clip in Premiere using Curves for exposure, Fast Color Corrector for white balance, and Colorista III for saturation and secondary corrections
- Color grade for the whole clip in Premiere using an adjustment layer with Colorista III and Looks 3 (often requires tweaking the color correction step for some individual clips)
- When everything is done, export from Premiere as H.264 with very high bitrate
- Re-encode to normal-bitrate H.264 in MKV wrapper that plays absolutely anywhere using Handbrake (reasoning: "pirates surely know how to deal with this > research > wow this actually works great, it does play everywhere")

Stu has also s*%tfed back from the "no mastering from your NLE", Premiere is good enough for most work right now. The 32-bit linear workspace thing remains, though (unless you use that all the overlayed stuff tends to look weird).
Click to expand...

What format is the source footage in? (is it log, 709 or something else)

PPro internally is 32 bit float, it is capable of great results but IMHO it is restricted when getting footage in and out. The next version (announced) supports HDR and OpenEXR and this is a good step, in the right direction but i've always felt that we, the users, need more control over specifying what kind of footage we are giving it.

cheers
Paul
 
These are Premiere presets, they don't work in AE. I work with unaltered log footage in AE.

There are two sets of curves:
- Some presets to adjust exposure: carefully adjusted to add or take one third of a stop of light, or two thirds, or a stop, etc. That's the effect in the midtones, the clipping points in the shadows and highlights don't change, so the effect there is smaller. They work great and are super fast way to fix exposure variations, see example here.
- Some presets to de-log the footage, adding contrast, again without clipping anything that wasn't clipped already. They push the mid point, add contrast in the mids, and roll the highlights and shadows.

This gives a nice non-log starting point for color grading. Or at least one that I like and which I find easy to work with.

If a clip was shot with correct exposure (i.e. overexposed one stop with respect to what the Sony would say) and has correct white balance, just adding one of these contrast curves and then a MB Looks preset is enough for a simple project.
 
You must log in or register to reply here.
Back
Top