Norway
Joined: Apr 18, 2014
Post Count: 257
Status:
Offline
3: Enhance and create effects to gifs or videos
Get the most out of a gif or video presentation – post-production in general with a focus on the TELDA film.
1. GIF POST PRODUCTION
Making gifs using an online service is easy, and understandably tempting. Dedicated post production software offers total control of the final result – a control we will not find in any online gif makers.
I will here use the free version of DaVInci Resolve to make a gif.
Basically all post production programs are similar when it comes to set-up, and that’s really the only thing we need to look at before starting to work.
A new word: Footage : The recorded, generated or created material we make our gif from.
So let’s assume that we have gathered the footage and are ready to start editing. Open DaVinci and start with the settings:
At the bottom of the start screen, there is this menu strip:
Click on the symbol far right. I have marked it blue.
A menu will appear.
This is the top of the menu. All that’s needed is to set the width and height of the composition. If it is for SH3D, like this demo, set the width to max 800 and the height to whatever fits the footage. Use only even numbers. (Settings can also be changed later.) Scroll down to the bottom of the menu and click save, and this window will appear:
This is where we import our footage. Press cmd + i, like it says. Any number of files can be imported at the same time. If there are many files, it can be smart to name them alphabetically. It makes it easier to place them in the correct order on the timeline. After import, look at the bottom of the screen and click on the edit symbol (marked):
This screen will appear:
(For this demo, I have borrowed “House by the Lake” from the demo pages of SH3D)
Grab the first clip and drag it down on the timeline, and it will appear in the preview window above. (Don’t worry if the left side of the screen don’t look exactly like this, I have some effects there.)
Note that there is a red loop symbol on the far right of the transport controls under the preview window. Looping can be activated and deactivated (toggled) by clicking it. There is, however, no need to use the transport controls, as the spacebar will play and stop the timeline. The red arrow on the left is the select-tool, and the magnet symbol represents is the snap tool. These two are the tools we need the most. (The function of each tool symbol will appear by hoovering over it.)
The left and right arrow on the keyboard moves the timeline cursor one frame back or forward. If we don’t want the cursor to jump back at stop, press number 0 to toggle jumping back on/off.
TIP: On a multiple display system, another monitor can be used for a clean preview:
Now drag the next clip to the timeline. If we select and push a clip upwards on the timeline, a new video track will appear as Video 2. There is no limit to the number of video tracks. Using two video tracks is recommended for creating transitions fast and easy. Use the selection tool to move a clip back and forth on the timeline. The duration of any clip is controlled by selecting the end of a clip, where a green line and a drag-symbol will appear. Make the clip longer or shorter by dragging the green line forward or backward.
Transitions ( How two join two film clips. Everything from a clean cut to an advanced animation.)
Adding effects is one place where a dedicated post-production program differs from an online gif-maker. The most used transition ( apart from a clean cut) is a crossfade. The easiest way to make a crossfade is to place clips on separate video tracks with an overlap. Here is how to do a crossfade:
The upper track will always cover the underlying track, unless the opacity is changed. Selecting a clip will reveal a white square in each top corner. Making a crossfade is very simple: Grab the white spot and move it until you have the right length. The longer the distance, the longer the crossfade. The overlap needs to be a s long as the crossfade.
Other, and more advanced transitions, can be found in the Effects menu.
Colour correction is fun, and can be opened from the bottom menu:
The colour correction works separately on each clip. Each clip will get a node where settings are stored. It’s easy to get lost. Use the reset option on the node to go back to normal and start over.
This is basically all you need to know to edit a gif.
For export, go to the menu strip at the bottom of the screen. Click on the delivery button (marked).
The settings page will open:
Click on “Add to render queue”.
And you will se this on the right side of the screen:
2. VIDEO POST PRODUCTION
I used Adobe After Effects for creating the effects – not because it is better than DaVinci Resolve, rather because I am more comfortable working with layers than with nodes.
Later I will go back to DaVinci Resolve for the assembly and the delivery of the final master.
These are some new words you might not know, that I will be using down the line:
Backdrop : Replacing the background with an image or a colour, Will be explained in detail shortly.
Keying : A digital technique we use to make backdrops invisible and to isolate special areas.
Tracking : Replacing one image with another by aligning it with the camera motion.
The magic of a BACKDROP and KEYING:
The word Backdrop comes from the theatre – a painted canvas or cloth that is lowered from the ceiling. The purpose is to create an illusion of a location, pyramids, snow covered mountains, maybe a beach, or the reception of a luxury hotel. The backdrop is changed from location to location.
We can also use the same kind of backdrops in film making, and we can use it in SH3D for creating the same sort of illusions as in a theatre. In filmmaking we normally use a backdrop to hide the location instead of providing it. To illustrate, I have created a typical small studio film-set:
A one-coloured backdrop can be used to change the location quite realistically. In this scene the backdrop is hanging behind the windows. It is completely green and is blocking the view from the windows for obvious reasons.
A backdrop must be lighted as evenly as possible, and we must avoid getting shadows on it. With today’s technology, the result can be monitored in real-time. Here is what the director can see on her monitor:
These days keying has become very easy.
Now, in a computer-generated video-reality, we don’t really have to worry about shadows or reflections. We simply use a luminous backdrop.
“A luminous backdrop?” Yes, and this is how to make one:
Open SH3D - new window. Draw an L-shaped wall with approx 30 meters sides, joined by a small 90 degrees arc. Set the hight to 5000 cm and the thickness to 1 - 5 cm. Then apply the same green colour to both sides of the wall.
Select the walls and export the selection as Green-screen.obj. Open Green-screen.mtl with Textedit or Wordpad (or what have you), and change the illum value. Illum 1 or 2 must change to illum 0. There might be one or two material groups in the mtl-file. If two, they should both read like this:
newmtl wall_1_1 illum 0 Ka 0.07058824 0.7490196 0.0 Kd 0.07058824 0.7490196 0.0 Ks 0.0 0.0 0.0 Ns 1.0 Ni 1 d 1.0
The Ka and Kd will most probably not read the same as mine, but it’s of no practical consequence. Save the changes in the .mtl and import Green-screen.obj to SH3D. On import you must adjust the orientation - (that means; fiddle with the arrows, until it looks like this.) :
You can of course adjust the angles in the modify furniture window, but then re-sizing can become a bit troublesome.
As shown here; no shadows can bite on a luminous surface:
Create a new level for the backdrop, and make the elevation zero minus the thickness of the backdrop.
TIP: Using a white luminous backdrop and a textured white sky, rendering with moon symbol and “Enable sky lighting”, will create an evenly lighted image without ground shadows.
Okay:
The settings are quite similar in all video editors. This is my setup in After Effects:
The following process will be similar with every layer based post production software.
I have now loaded my files and placed the background on the timeline. Now I need to place the foreground, and use keying to get rid of the green stuff:
And then I will place the “Middelground”, which is the tram.
That did not work, so I must render a new video with a different coloured backdrop. I make some changes to the tram and render a new video, this time in Q4. I want to see how the tram blends inn after I have taken it to pieces and changed and added textures and decorations.
Every video editing software will be able to freeze a frame. There’s an option in Aftereffects called: “Freeze on last frame”. The frozen image will stay until cut.
I think the tram blends in well.
The next word is TRACKING
As you can see in the Telda-film, images on the monitors are changing while the camera is in motion. One way to do this is to create multiple video renderings with different images on the screens, and then make transitions between the takes.
The other, and faster, method is called Tracking The tracking procedure is similar in most video editors.
Open tracking from the Animation menu.
Drag the track points in place.
Apply Analyse, as shown in the menu:
Then it can result in this:
Neither SH3D nor After Effects can create walking and talking people. I have checked if there are any free software that can do that, and in addition to Blender, there is a free program called DAZ. DAZ can make people walk and talk, but it’s a rather complex task. (So is animation in Blender!) (But everything gets easier once you get the hang of it…. they say)
The big brother of DAZ is POSER, and it is not free. It starts at $ 249.95. (It has a free trial period.) Poser needs some training for things like walking and talking, but it can do a lot more. I have used Poser for many years, and I still feel I am not really mastering it. Today there are several AI-programs that can make more realistic characters than Poser, but more realistic is also more synthetic in a way - a bit like the images created by ChatGPT. I like the style of Poser, and the fact that it does not pretend to be anything but animation software. I don’t want to loose the animatic look.
Well… Integrating SH3D renderings with characters from Poser can present some challenges. Creating walks is one of them, especially when someone is walking around corners, coming out to the street from a building or coming into a room from an open door. Over time I have learned that “markers” can be a great help. Markers can be an obj-file of relevant models from the SH3D-project, or just plain boxes with the exact same size and placement. These boxes are quick to make in SH3D, and they do not weigh more than a few kb, making them easy to import and handle in Poser. I prefer using boxes as markers for two reasons: One: They help me create good walk paths a lot faster. Two: They can be used by Poser as invisible shadow catchers, which is equally important. Shadow catchers can be rendered with an alpha channel, meaning I can adjust the softness and strength of the shadows on separate layers in any post production program. Most of the shadows in Telda are created with Poser, but shadows from buildings and bigger objects, like the tram, are created with SH3D.
Here is a demo from another Telda-project:
For simple walks in straight lines, I skip the markers, and just use a top-view image from SH3D. It’s always best to create walks from a top-view camera:
I don’t want to zoom her to invisible, so instead I make the walk path much longer than needed. That way I have something to go on for the next step.
Perspective is another challenge. SH3D is using “Field of view”, where Poser is using millimetres, and direct converting does not seem to correspond. I have to rely on what I call “helpers”.
Dimensions. How can we be sure that the dimensions of our 3D models are correct? I need to compare to a known object, like my “helper” - a business man who is 1.80 high. To help getting the perspective right, I spread helpers at all relevant places, and include them in my background image for Poser. Then I can compare to them when I place my walking characters.
I admit that the tiled texture on the platform is chosen deliberately to make perspective alignment easier. When I get the guides from Poser aligned with my platform tiles, I’m half way done.
Files from Poser are very easy to import to any video-editor, since they need no keying.
I see no relevance in describing the process of creating sound and music in this post, but sound and music has a natural place in all film and video production.
DaVinci Resolve has a semi-professional audio editor plus a free sound library that can be downloaded. Not very easy to use, though. My preferred free program is called Audacity (https://www.audacityteam.org). It runs on Windows, Mac and Linux, and has all that’s needed to produce a decent soundtrack for any video project. Avid Pro Tools has a free but limited intro version that runs on Windows and Mac. Top quality, but steep learning curve. Cakewalk is a free and very capable program, but runs only on Windows. Garageband is a good alternative for Mac, as it is part of the MacOs.
I am now back at DaVinci Resolve for the final master creation.
Above we have the delivery options for DaVinci Resolve. I use DaVinci Resolve to synchronise my film and the soundtrack, and to export the final video master.
The assembly/editing is no different for film/video than for gifs, as described at the start of this post. The only difference is the audio, which is not available in gifs. If the video has a soundtrack, it will automatically be placed on an audio track right under the video.
About audio: There is one thing I know from experience that needs to be mentioned before I go on:
“Scrubbing”. Scrubbing means that the audio playback will follow the cursor movements, and scrub the audio forwards or backwards as we move the cursor back and forth on the timeline. This function is what frustrates beginners the most, and this is how to turn it off:
Standard video format for social media streaming is MP4. Clicking on a link to an MP4-file, the video will open and play in a new browser tab. So an MP4-file is what I am making.
DaVinci Resolve offers multiple presets for delivery to the most common platforms, like Youtube, Vimeo, Tik-Tok, Facebook and X, and it even offers direct uploads.
For export, go to the menu strip at the bottom of the screen. Click on the delivery button (marked).
The settings page will open:
If your video has audio, there is one more important thing to do: With the render menu open, click on the Audio tab, and these settings will appear:
Click on Add to Render Queue:
The witch decorating the tram, is a tribute to Ceciliabr who has kindly donated me her entire archive of 3D models.
Telda’s distinctly tasteless and unbecoming costume, is simply a matter of making her a recognisable person – a person we will immediately notice in a crowded place – visible without being spectacular – just easy to spot. The reason for that: She will appear in another film where this is essential.
Dr. Dusan Calvin is named after Isac Asimov’s fictional robot psychologist; Dr. Susan Calvin.
Café Social is named after a lunch bar in Frankfurt, where I was generously let in a quarter before they opened, and was served a big mug of coffee after having driven non-stop from Oslo to pickup a girl with an ear infection that prevented her from flying.
I often use names from real places. Constructed names have no history.
Attention guiding – to visually attract attention to a particular part of the screen.
At the end of the main scene of the film, when our man Dr. Dusan Calvin is leaving after the Telda has vanished, he declares a wish to have a look at the café. He then exits camera left. In the next scene, with the overly noisy café in the background, he is coming towards us from a camera right position. When I tried this transition, I got a feeling that the cut was jumping. I wanted viewers to see him already at the first frame of the café scene, so after he has exited on the left side of the frame, I have inserted some giffy birds flapping and flying across the picture from the centre towards the right, where they become invisible when passing the opening of the dark garage. At exactly the point when the birds are at the same place on the screen as our Dr. Calvin will appear in the next cut, I do the transition – and the cut is no longer jumping. Of course this is more noticeable on a big screen than on a laptop.
I do the same with the next transition: As our man Calvin exits camera left in the Café Social scene, an uncategorised bird of questionable origin comes diving down to feast on a hot-dog some overfed schmuck has dumped on the sidewalk. It’s supposed to have the same effect as the birds in the scene before. I don’t know if that works.
The girl on the phone is based on a conversation I overheard in the reception of Hotel Europa in Prague 2001. She was right: The Hotel Europa WAS creepy!
Making the Telda video has inspired me to write a script for a 20 minutes animation, based on an answer Dr. Susan Calvin gave to a reporter in a fictional interview:
Reporter : “Are robots so different from men?” Susan Calvin: Worlds different! Robots are essentially decent!
Netherlands
Joined: Apr 8, 2022
Post Count: 1001
Status:
Offline
Re: 3: Enhance and create effects to gifs or videos
Wow! Again a great tutorial. This will take some serious study time. Thank you!
---------------------------------------- Dodecagon.nl 700+ 3D models, manuals, and projects
Russia
Joined: Aug 13, 2022
Post Count: 352
Status:
Offline
Re: 3: Enhance and create effects to gifs or videos
Thanks! I'm still afraid))) This is a completely different level, professional. Thank you again for sharing such important secrets of the profession. We need to study everything carefully...
France
Joined: Mar 3, 2015
Post Count: 875
Status:
Offline
Re: 3: Enhance and create effects to gifs or videos
Whaouh! Thank you so much, Xiste, for revealing all your secrets! It's exciting, very educational. I'll try to find the time to experience some of these tricks (I did a try, some times ago, with the tramway moving. Very rudimentary result!)
