Creality Sermoon S1 3D scanner review - Part 2: Using Creality Scan 4 software with Intel Graphics

Earlier this month, we checked the specifications and did an unboxing of the Sermoon S1 high-end 3D scanner with support for blue laser and infrared scanning. I had some concerns about my host machines due to the fairly high minimum requirements: Intel Core i7 Gen7 CPU, NVIDIA GPU (6GB VRAM), 16GB RAM, Windows 10/11 (64-bit). I had planned to use an ASUS Vivobook 16 laptop with a 13th Gen Intel Core i5-13500H SoC with a 12-core Raptor Lake processor and integrated Intel Iris Xe Graphics, as well as 16GB of RAM to run Creality Scan 4 software. After talking with Creality, it was confirmed to me that a “dedicated graphics card” was required.

I have some good and bad news. The good news is that I might get a mini PC with an NVIDIA RTX 5060 Ti (16GB) external GPU for review. The bad news is that I have to wait until the end of December or the beginning of January to get the kit. So I decided to go with a three-part review of the Sermoon S1 scanner. After the unboxing post, I’ll now test the 3D scanner with my underspec’ed laptop with integrated Intel Graphics, and review it with more objects and scanning methods once I get a machine with a proper GPU. The key takeaway is that you can use the Sermoon S1 3D scanner with Intel graphics, but there are caveats, and I would not recommend it.

Creality Scan 4 installation and initial setup

The first step is to connect the USB-C Y cable to the Sermoon S1 3D scanner and secure it with the two built-in screws. Then you can connect the other ends of the USB-C Y cable to the power supply (DC power jack) and a spare USB 3.0 port on the host computer. Note: I used a 5 Gbps USB port through a USB-C dock connected to my laptop.

I then booted into Windows 11 (Linux is not supported) and installed Creality Scan 4 software, also written as CrealityScan 4 without spaces. When launching the software, we are asked to select the scanner model, although we can see the Sermoon S1 was already connected in the background.

I selected Creality Sermoon S1, and after clicking the View button, I was brought through an 8-step Quick Start wizard. It mostly explains how to use the scanner. On important part is the differences between Blue laser mode for high-accuracy, high-detail scans that may require markers (some black and white stickers) and the Infrared mode for fast scanning, and tracking of geometry (e.g., human faces and bodies, sculptures….), texture (e.g., a porcelain vase with drawings), and markers for objects with insufficient geometry and texture.

The right scanning distance is also important, since you may have under-exposure or over-exposure otherwise, as explained in the screenshot below.

Performance test and PC requirements

After the Quick Start is over, the software will run a performance test, and I didn’t like my laptop claiming “Poor PC performance”. I can expect a scanning frame rate of up to 9 fps in blue laser mode and up to 16 fps in infrared mode. Spoiler: In practice, it will be quite lower than that.

We’re told we can Learn More, so let’s do that. The important bit:

For 3D scanning software, NVIDIA graphics cards are usually a must and often the only choice. AMD or Intel graphics cards may not work properly.

The table below lists the requirements of the host for optimal user experience with the Sermoon S1 and other Creality 3D scanners.

So my laptop is indeed underpowered for this use case. 64GB of memory is especially recommended for Blue line scanning, and it will also depend on the size and complexity of the scanned objects:

For infrared structured light scanners like Ferret and Otter series, 16GB is currently the recommended starting point for a smooth experience. For blue line laser scanners such as Raptor and Sermoon series, which generate more point cloud data during laser point cloud scanning, it is recommended to increase RAM to 32GB and 64GB.

You’ll want to get a $1,500+ computer or (gaming) laptop with your $2,500 3D scanner if you don’t have one already.

Firmware update and Sermoon S1 calibration

Never mind, let’s carry on with my underpowered $750 laptop :). The Creality Scan 4 software then detected a new firmware for the Sermoon S1.

The first time I did that, the software hung during the initial Parameter backup step. I waited for one hour before I decided to kill the software. After a short “is my device bricked?” scare, I restarted the software, and it completed the firmware update successfully.

After that, I clicked the Calibrate button, as the CrealitScan 4 recommended. I was asked to scan the QR code on the back of the calibration board with the 3D scanner. The machine detected it, and I was asked to place it on my desktop in the same orientation as shown in the program.

After clicking OK, a 10-step calibration process started. I was asked to move the 3D scanner around as shown in the program, either straight up or at an angle with the scanner’s camera and sensors always facing the board. This is done for Near Mode…

… and Far Mode.

Once the 10th step was complete, the program performed calculations for calibration, and after a minute or two, the calibration was completed. I got a 95.36 score. I understand anything above 90 is fine. It’s a fairly straightforward process once you understand how to move the scanner in space.

First scans with Creality Sermoon S1, model creation, and 3D print

At this point, we are ready to create a new project and start our first scan.

I first had the idea to scan a LEGO figurine, but I didn’t get anything in either Blue light or infrared scanning. This is despite making sure the distance was in the optimal range and the object was indeed captured by the RGB camera. Then I remembered to read the specs again, and the minimal object size is 5x5x5mm with Blue line scanning, or 150x150x150mm for infrared scanning. So only the former would be an option for a small figurine. Anyway, I tried something larger with a plastic lantern to which I added markers. While markers were detected in the software, I could not see any point cloud, and I noticed some serious lag.

I eventually decided to scan my own face using the Infrared mode. Hooray! I finally got something. When scanning, it’s important to check the distance for the right exposure, either using the LED on the “Play” button of the unit or the colorful bar in Creality Scan 4. When scanning another subject, I ended up mostly checking the monitor since I also needed to check whether the target was in frame with the RGB camera. You’ll also notice the rather low 5 FPS frame rate during capture.

This was the first result of my first scan. Good progress. However, I failed to scan my entire head myself.

So I asked somebody else to be the scan subject. I managed to capture the head and part of the bust, but the lack of support for the Intel GPU results in a low 4-5 FPS, and it took me about fifteen minutes to properly scan the head without holes in the point cloud. I found the back of the head/hair harder to capture, so I had to do many passes there. Also, I initially forgot to probably scan the chin and the top of the neck, so there was a hole in the capture, but we can always resume the scan as needed. It would have probably taken me 2 to 3 minutes with an NVIDIA GPU at 30-60 FPS, better lighting, and more experience with the 3D scanner.

At the end of the scan with have raw data that is still imperfect.

We can now click on the Fusion icon to get a cleaned-up result.

I had some issues with storage capacity during that step. I have a 512GB SSD with Windows 11 (97.39 GB) and Ubuntu 24.04 (376.95 GB) partitions. The C: drive only had a few GB free, and Fusion processing takes several GB, and it failed to complete due to a lack of storage space a few times. I cleaned the drive, deleted a few programs, and even compressed the C: drive before CrealityScan 4 could complete the task.

This results in a cleaner 3D shape, but I noticed holes on the back of the head and neck. This can be addressed by resuming the scan or having the program handle that part in the Mesh Processing before exporting the data as an STL file for 3D printing or further editing in Blender or other programs. I went with Meshing, selecting the option to fill small holes with a 20mm perimeter and also requesting a closed design.

I had no issues with storage this time around, but the memory capacity was. I had to close all programs to complete the conversion to Mesh, so 16GB of RAM is definitely tight here. That’s with the Infrared mode, and the Blue line mode requires even more memory.

At this point, I exported the capture as an STL file (62MB in size). I also noticed a Head AI Repair option. So I downloaded the plugin (about 200 MB)…

… and selected one of the models closest to the capture I had made. Something happened with new hair and a proper bust, but the scanned face was gone for good. I think “Test” may mean it’s still at the Beta stage..
I clicked Reset, so cancel it. One thing that bothered me was that I could not find a Save project option in the program. So I closed the CrealityScan 4, reopened it, and I could still see my project listed. That means everything is automatically saved.

So I have an STL file now. Good. But can I actually use it? To find out, I wanted to print it with the Ender-3 S1 Pro 3D printer connected to a Klipper controller. I used Cura to generate the GCODE file after scaling the design to about 17% and automatically adding support underneath.

I used PLA filament at 200°C. Here’s the result taken out of the 3D printer with minimal cleaning.

The front does not look great, but the back and hair are not too bad, except for a vertical line… Two reasons for the result. First, Cura generated supports on the T-shirt. Not sure why, and I missed it before exporting the G-code. I further cleaned it up, and that part looks a bit better now.

I can’t fix the face, and after asking Grok, it seems the most likely reasons are either a dirty nozzle or the print speed was too high for the face. I’m going for the latter since I’ve been using this printer intermittently for 3 years, and never changed or cleaned the nozzle. The scanned data could be better, too. See STL file loaded in Blender.

It’s probably because the subject moved a little (or even talked!) during scanning. I suspect it would improve with an NVIDIA GPU as it’s easier to ask somebody to stay still for 2 or 3 minutes, rather than 15 minutes…

That’s all for today. My main takeaway is that it’s possible to use the Creality Scan 4 software with the Sermoon S1 scanner on a laptop with a 13th Gen Intel Core i5 CPU with Intel Iris Xe graphics and only 16GB of RAM, but you will have limitations to what you can scan (large objects and Blue laser scanning will be an issue), and the framerate is rather low resulting in wasted time or suboptimal results. A more powerful machine with 32GB or 64GB RAM and an NVIDIA graphics card with 6GB/8GB RAM is a must to make full use of the scanner’s capability. Note that Intel and AMD graphics cards are not supported by the software. Hopefully, I’ll get a Windows machine with more memory and an NVIDIA GPU soon, and give it a try with more objects.

I’d like to thank Creality for sending the Sermoon S1 3D scanner for review. It can be purchased for $2,399 on the company’s store in the configuration reviewed here, as well as on Amazon for the same price.

Jean-Luc Aufranc (CNXSoft)

Jean-Luc started CNX Software in 2010 as a part-time endeavor, before quitting his job as a software engineering manager, and starting to write daily news, and reviews full time later in 2011.