Q&A of POP 2 Hardware Improvements & Software Parameters Adjustment

Dear Users,

We noticed some questions you may concerned about, and want to clarify our answers here.

(1)Has the POP 2 scanning hardware been improved? as im wondering if the POP2 uses the same scanning hardware with POP but just tweaks the settings in the config file like people have done with the POP1.

The hardware of POP 2 has been fully upgraded from chip to sensor (see table below for details).

Thanks to the overall improvement of POP 2 hardware, its scanning effect also has the following improvements:


Both POP and POP 2 are high-precision scanners, and the accuracy of POP 2 is up to 0.1mm, better than POP’s 0.3mm. This subtle difference can be measured with very sophisticated instruments and specific measurement methods, and you can find the obvious differences easily when scanning objects with rich small details, however, it is difficult to visually recognize when it comes to scan objects lack of details.

Scan Speed and Texture:

The scan speed of POP 2 has significantly improved from 8 fps to 10 fps, and model texture performs more vivid. Refer to beta tester’s models at our forum - the texture result is much better than POP’s.

All above improvements are closely related to hardware improvements. POP 2 will surprise you with significant improvement in both scanning speed and accuracy, especially when scanning objects with rich details.

2. Ivan Calvin Hobbes only changed some of the software parameters, which greatly improved the model details. Why didn’t Revopoint do setting like this?

It’s glad to see that we have so many professional users like Ivan.

Firstly we’d like to clarify that the hardware has been improved to make POP 2 reach a higher accuracy of up to 0.1mm, a better texture result and more fluency scanning experience at a speed of 10 frames per second.

Secondly, we explain “Why didn’t Revopoint do setting like this?”:

1) Software Algorithm

From single frame data to a completed 3D model, there are very complex algorithm calculations that will affect many key indicators.

POP or POP 2 is as an entry-level high-precision 3D scanner. When the POP series 3D scanners are designed, the key indicators we consider orderly are robustness (the most important), accuracy , fluency , calculation speed , and then, model details .

Adjusting the alignment parameters of POP and POP 2 for a specific model (such as Agrippa), one could get a better result than that of the official settings, but it could also cause bad issues, such as more frequently miss track appearing, more noise, longer calculating time (above 10 mins perhaps) and others. Without lots of tests, you can’t ensure what results you will get.

2) Open Mesh Parameters

Since the first post of Ivan was released, we have been discussing how to enable professional users to achieve better results through independent adjustment of parameters. In the new version of Handy Scan (underdevelopment), we will open mesh parameters. The following is the crude UI:

(Only for reference. It’s not the final version)

As POP & POP 2 are consumer-grade product, we contribute to bringing our users a good scanning experience and don’t want to make the product very complicated.

Believe us, the mesh parameters opening is a start, we will gradually open more parameters to professional users.

3) Fast Scan & High Accuracy Scan

As you may know, in the beta test software of POP 2, we have two scan modes, fast scan & high accuracy scan, to meet various users’ need. In future, we are willing to design more modes according to the needs of users (especially modes for professional users).

What objects do you want to scan most with POP or POP 2? What is your computer configuration? Welcome to leave a message for our reference to introduce more models to you.


Thanks for your feedback :wink: !! Nice news.
Agree with you, more post processing needs when we tweek the software for more details , but for advanced user, it will be very useful :wink:


What is the resolution of the Depth map of the POP versus the POP 2?

Thank you for this reassuring message on opening the parameters to users who want them and for the details concerning the hardware improvements.


It is good to see this response . I have generally had a good experience with my POP, and it has been far superior to my previous scanners. However, the numerous improvements being promised with the POP2, look as though they will address the few issues I have had with the scanner.


Consumer seems to imply “dumb”. Your product is purchased by professionals who have more important places to put their money than in a scanner they use once or twice a year.

Go ahead and make it complicated as an optional set of settings. Quality scan results are worth any optional complexity


Great news! Especially the option of choosing the mesh density will be very helpful in my case, since sometimes I need a fast, not very detailed scan, and in other cases I want to get as many details as possible, and willing to spend more time to get it.

Colour, speed,easier tracking and alignment and a press button are all very welcome of cause.
Maybe I will (in some cases) use the scanner to scan feet directly instead of casting the foot first and than scan the cast, my current workflow to create made-to-measure footwear.


How will this parameter rollout be introduced? Will users be able to easily tweak their existing devices? Or will we be locked out with encryption as appears to have happened with POP1 & need to buy new devices in order to have updated parameters?


Hi! So, is the accuracy of up to 0.1mm independent of distance from object? Or is that only at closest distance (150mm)?

Indeed these developments are welcome. I understand that some maybe even most users would use the POP in order to 3D print their subject so texture quality is not that important for them. For me and for some others recording an object in 3D with high detail in the texture is the goal (not for printing), especially in the cultural heritage sector where I work. I would really like to see a higher megapixel RGB camera included in the POP2 or a variant. On chip processing is a good step forward but it does not record finer surface detail.

I was able to get very similar results to the POP2 texture from my POP1 using the third party Topaz Sharpen AI and Gigapixel AI as I demonstrated by processing the texture file of your original dinosaur scan on Sketchfab.com. This is your original textured model Dinosaur Texture by POP (obj) - Download Free 3D model by Revopoint3d (@Revopoint3d) [f6009a7] - Sketchfab and this my AI modified textured model Dinosaur Texture reprocessed with Topaz AI - 3D model by HeritageNI.com (@heritageNI) [e5e6825] - Sketchfab. This is good up to a point but not a substitute for higher megapixel and therefore more real detail.

Perhaps a larger sensor could be fitted without to much other internal modifications? If the cost was a bit more then a variant product could be maybe be considered?


I strongly agree with this statement made by HeritageNI. The 2 mega pixel camera on the POP 1 & 2 do not provide sufficient texture quality. The texture maps are not sharp enough.

More details can be obtained if the textures are of higher quality, giving sharper, clearer images. I use zBrush to make a bump map from the photo textures of scans,this way, more detail can be added to the final model. A plain model can be given fine details in the actual geometry and 3D printed. Please see the example images. Revopoint, please consider making the camera higher than 2 megapixels in the future. Would that be possible for the POP 2? Thank you.

Revopoint POP 1 scan from sketchfab. Base model with more details applied using the texture.

Another 3D scan with high quality textures, used to create nice details on the smooth model


Hi, you forget that the RGB camera takes 8fps at the resolution of 1280 x 800 of each frame . So you scan 100 frames , it makes 100x at 1280x800 , and so far it is the best resolution 3D camera can offer , because POP1 or POP2 is actually a 3D camera not really a 3D scan , 3D scans needs calibrations, 3D cameras don’t.
I would suggest to not generate the textures in the software , but instant export the obj in ply format , it would include the per pixel color vertex info so in Zbrush you can project the color data into you clean UVs model. You need to install the ply plugin for zbrush what is free .
My workflow is as follow:
Scan the model , generate the OBJ from the cloud or use Meshlab to do that, export the obj in ply format, import it to Zbrush , under subtools create a duplicate , create clean UVs for the scan and after that project the color data from the original to the final OBj with new UVs , it would give you higher resolution and much better way to edit the textures or touch up .
Sometimes I edit the color data on the original OBJ direct in Zbrush before projecting .

Adding higher resolution for the RGB sensor will not make dramatic changes , it is not so easy as it sounds. There are still limitations of how fast Pop1 or 2 can process the data and higher resolution will affect everything in the process and the costs would be so much higher to produce it and it will be no more consumer friendly and definitely you can’t use phone with it , remember Pop1-2 is a mini AI computer on its own already .

1 Like

I’m looking forward to modes for professional users.


Hi @Erked ,

Thank you for your reply.

You can download the latest software version from our official website. These software adjustments have been realized on the new Revo Scan.

If you have any further suggestions, please feel free to let us know. We will keep our eyes on improving our software.

Best Regards

1 Like

Overall you guys did a great job with the scanners. Major thing I would like to see improved is the frame rate. 20 to 30 fps would help with tracking. The last major improvement would be the ability to scan without overscanning. The last one is difficult and and costly if I understand it correctly. Thanks all at Revopoint for your hard work.

Hi Eric , I don’t think you would be able to get 30FPS using WiFi, phones or tablets , you also shutting additional color frames so that would limit it only to windows PC like Einstar at this moment .
There are limitations how fast you can go on that budget.

12-18FPS that Range offers are amazing in comparison with just 10FPS , tracking issue is virtually not existent even on featureless objects .

Also the big FOV is also a key in better tracking .
And of course the difference between a phone or a PC that can handle the speed and not chocking on it .

1 Like

Oh, okay. 20fps would be fine with me. I see what you are saying in regards to performance vs cost. I’m so excited to see what it can do.

1 Like