Here is the video shooting guide for KreadoAI Avatar & Voice Clone video shooting specifications.
This guide provides a comprehensive overview of the video shooting process for KreadoAI custom avatar. We encourage you to read it thoroughly in advance to understand specific requirements for equipment selection, location, model appearance, recording process, and the final video processing. This ensures that the customization outcome aligns with your expectations. KreadoAI currently supports green screen shooting and live shooting, please check the corresponding shooting guide according to your needs!
Submit entry:
1. Preparation for shooting (Choose Based on Your Preferences: Green Screen or Real Background)
1.1 Location:
Choose a noise-free shooting environment, paying particular attention to avoiding areas with planned construction activities on the day of the shoot.
1.2 Background
1.2.1 Green Screen
Ensure the green screen is smooth with even color and no wrinkles. The ideal distance between the model and the green screen is >2.5 meters. If shooting full-body, also cover the ground with the green screen.
When using a green screen background, avoid wearing clothing in colors close to the green screen (such as yellow or green) to prevent affecting the keying effect.
Minimize loose hair to prevent difficulties in keying the green screen.
Ensure uniform and adequate lighting with no shadows or overexposure.
1.2.2 Real Background
In post-production, it's not possible to change the on-location background. If there are no specific requirements during shooting, make sure the set is satisfactory before starting.
Ensure even lighting sources. "Three-point lighting" is recommended. If you want no shadows in the chin area, consider using reflectors or foam boards for additional lighting.
Mobile shooting is supported. Choose the highest quality (1080P or above) for shooting. For iPhone users, select the "Best Compatibility" format for shooting.
Ensure a stable background, free from real people or realistic images, to avoid impacting the final generation effect.
1.2 Equipment:
Select a smartphone or camera capable of recording videos with a resolution of 1080P or higher and a frame rate of 25FPS or higher.
Securely mount the smartphone or camera using a professional tripod or phone holder. Disable the autofocus function to ensure a stable, focused, and shake-free recording.
If using a camera, install a PAD prompter or professional prompter device. If using a smartphone, download prompter software. Ensure that the model maintains direct eye contact with the camera during recording.
Please note that background replacement is not supported for videos recorded with a smartphone.
1.3 Model:
Ensure the model's clothing is neat, using non-reflective fabrics, and avoid colors similar to the green screen to minimize the difficulty of background removal.
Avoid wearing pure white clothing.
Ensure the model's hair is well-groomed and does not obstruct facial features or the neck. Use hair gel or similar products (if necessary) to control stray hairs and prevent interference with green screen background removal.
Keep the model's makeup clean and free of oiliness. If glasses are worn, preferably choose invisible or small-frame glasses to avoid reflections during filming.
Models should thoroughly read the script (approximately 2000-2500 words) before the shoot, avoiding frequent pauses during recording. Models can use the provided sample script or their own familiar script. Ensure that the on-site script content is not repetitive and can support continuous speaking at a normal pace for over 7 minutes.
2. Shooting Process:
2.1 Test Shoot:
Before the final start, conduct a test shoot to ensure:
Frame: The model is appropriately positioned in the frame, with eye contact, no tilting or skewing, clear facial features, and any necessary hand movements stay within the frame.
Prompter: The prompter playback speed matches the model's speaking pace, keeping the prompter and camera in the same position to avoid wandering eyes.
Sound: Clear, noise-free audio synchronized with the video.
Expression: The model exhibits a formal and natural shooting state, with facial expressions and movements appearing natural, avoiding stiffness.
Note: If the model lacks extensive experience in on-camera delivery, hands can be naturally placed in front or on a tabletop for a more relaxed and natural appearance.
2.2 Final Shoot:
Ensure a quiet environment, and after clearing irrelevant personnel, proceed with the formal shoot. The estimated duration is 7-10 minutes, shot continuously without pauses or post-editing.
2.3 Sequence:
5-Second Silence: After powering on, the model remains silent for approximately 15 seconds, hands placed in front, maintaining direct eye contact. Blinking is allowed, but no mouth movement or other facial and hand gestures.
7-Minute Voiceover: Begin scrolling the prompter as the model records the scripted video with clear enunciation. Facial expressions, head movements, and natural hand gestures during the voiceover can be designed based on the final composite video requirements. Avoid excessive head movements and frequencies (side-to-side, up-and-down nods), and steer clear of specific gestures with explicit meanings (such as counting with fingers, waving, or shrugging). If hand movements are involved, ensure they do not obstruct the face and neck (keep them below the neck) and avoid actions like licking lips, sticking out the tongue, pouting, or exaggerated facial expressions. Keep lips closed during pauses in the voiceover. If certain movements feel awkward, placing hands naturally in front or on a tabletop is acceptable.
3. Video Submission:
Please provide a video file with audio (MP4 or MOV format supported). The video should have an aspect ratio of 16:9 for landscape or 9:16 for portrait, a resolution of >=1080P, a frame rate of >=25FPS, and a duration between 7 to 10 minutes.
Before submission, please perform a detailed check to ensure that the submitted video meets the requirements:
3.1 Video File - Visuals
The model should maintain eye contact with the camera throughout the entire video.
The model's facial expressions and posture should appear natural and relaxed, avoiding tension or stiffness.
No second person's face should appear in the video (including images or depictions of people on other objects).
The model's body and head should not make significant movements or turns.
The model's head and hand movements should remain within the frame, and hand actions should not obstruct the face or neck.
Even lighting on the model's face, clear facial features, and well-defined facial contours should be visible without hair or bangs covering them.
The model's gaze should consistently be directed towards the camera, with a normal blink frequency and no instances of erratic glances or side glances.
During the first 15 seconds of silence, the model's lips should be closed, maintaining a natural facial expression.
The model should speak without stuttering, with clear pronunciation, natural lip movement, and visible teeth. Lips should be closed during pauses.
No edited or spliced scenes, and no frame-skipping situations.
3.2 Video File - Audio
Clear audio with no significant echo or reverberation, low noise, and synchronization with visuals.
No disruptive sounds like coughing or throat clearing from the model, and no secondary voices.
Standard pronunciation without the use of dialects.
Natural recording state, normal conversational expression, natural pause rhythm, and the ability to convey emotions moderately without excessive exaggeration.
It is recommended to use a microphone-equipped, high-performance smartphone, or recording device.
4. Reference Characters for Filming:
Male Business Attire | |
Female Business Attire | |
Male Casual Attire | |
Female Casual Attire | |
Male Sitting Posture | |
Female Sitting Posture |
5. Sample Text:
When I was only six years old, in a book called "True Stories" about the old forest, I saw a wonderful illustration of a boa constrictor devouring a large beast. At the top of the page was a facsimile of the painting.
"The boa constrictors swallow their prey whole without chewing, and are then immobilized; they digest it during a long six-month sleep," the book says.
At the time, I was thinking a lot about my adventures in the jungle, so I also made my first drawings with colored pencils. My number one work. It goes like this: I show this masterpiece of mine to grown-ups, and I ask them if my painting frightens them.
They answered me: "What's so terrible about a hat?"
What I drew was not a hat, but a boa constrictor digesting an elephant. So I drew the situation in the python's stomach again so that adults could understand it.
These grown-ups always need explanations. My second number works like this:
The grown-ups advised me to put aside these pictures of boa constrictors with open or closed bellies, and to focus on geography, history, arithmetic, grammar. In this way, at the age of six, I gave up the wonderful profession of being a painter. The failure of my No. 1 and No. 2 works discouraged me. These grown-ups don't understand anything on their own, and they have to be constantly explaining things to them. It really bores the kids.
Later, I had no choice but to choose another profession. I learned to fly a plane and have flown almost all over the world. Indeed, geography has helped me a great deal. I can tell China from Arizona at a glance. This is useful if you get lost at night.
Thus, in my life I have had many contacts with many serious people. I've lived a long time among grown-ups. I have observed them carefully, but this has not changed my opinion of them much.
When I meet a grown-up who seems to be a little clearer, I take out my No. 1 work that I have always kept to test him. I wonder if he really has the ability to understand. But the answer was always: "It's a hat." I didn't talk to him about pythons, primeval forests, or stars. I had to accommodate them and talk to them about bridge, golf, politics, ties, and so on. The grown-ups were then very glad to know such a reasonable person as me.
I was living alone, with no one I could really talk to, until that breakdown in the Sahara six years ago. Something is broken in my engine. Since I had neither a mechanic nor a passenger with me, I attempted to perform this difficult maintenance work alone. It's a matter of life and death for me. The water I carry with me is only enough for a week.
On the first night, I slept on this big desert far away from human fireworks. I am lonelier than a shipwreck on a raft in the middle of the sea. And you can imagine my surprise when a strange little voice woke me up at dawn the next day. The little voice said:
"Draw me a sheep, please?" "Ah!" "Draw me a sheep..."
I stood up all of a sudden, as if I was being bombarded by thunder. I rubbed my eyes vigorously and looked carefully. I saw a very strange little fellow staring at me seriously. This is the best portrait I ever made of him. But, of course, my picture is much inferior to his own appearance. It's not my fault. At the age of six, grown-ups discouraged me from my career as a painter, and I never learned to draw, except boa constrictors with open bellies and closed bellies.
My eyes widened in amazement at the sudden appearance of the little guy. Don't you forget, I was in a place thousands of miles away from people. And this little guy gave me the impression that he neither looks lost, nor does he look tired, hungry, or scared. He didn't look like a child lost in the middle of nowhere in the desert. When at last I was able to speak again in my astonishment, I said to him:
"Oh, what are you doing here?" But he calmly, as if he had something important to do, repeated to me:
"Please... draw me a sheep..."
When a mysterious thing overwhelms you, you dare not disobey it. In this uninhabited desert, facing the danger of death, although such a behavior makes me feel very absurd , I still took out a piece of paper and a pen. At this time, I remembered that I had only studied geography, history, arithmetic and grammar, so I told the little guy that I couldn't draw. He answered me:
"It's okay, just draw me a sheep!"
Since I have never drawn a sheep, I repainted the boa constrictor with its belly closed in the two paintings I only know how to draw.
"No, no! I don't want a boa constrictor. It has an elephant in its stomach."
I was dumbfounded when I heard his words. He went on: "The python is too dangerous and the elephant takes up too much space. I live in a very small place and I need a sheep. Draw me a sheep."
I just drew him.
He watched intently, and then said:
"I don't want it. The sheep is very sick. Draw me a new one."
I started drawing again.
My friend smiled innocently and cutely, and politely refused: "Look, what you drew is not a lamb, but a ram with horns."
So I drew another one.
This painting, like the previous ones, was rejected again.
"This one is too old. I want a sheep that will live."
I'm getting impatient. As I was in a hurry to overhaul the engine, I sketched the picture and said to him hastily:
"This is a box, and the sheep you want is inside."
At this moment I was very surprised to see my little judge beaming with joy. He said:
"That's exactly what I want, . . . You say this sheep needs a lot of grass?"
"Why do you ask that?"
"Because my place is very small..."
"What I drew for you is a very small lamb, and the place is small enough to feed it."
He put his head close to the picture.
"It's not as small as you say... Look! It's asleep..."
In this way, I met the little prince.