As i scan for making 3D asset is quite annoying to have poor texture, photogrammetry take time to set up and process, so revopop is perfect to quickly get something. The only things revopop is missing at this point is nice and clean texture. I think of cross polarized workflow to avoid specular and get albedo, i see some new possibility with double smartphone camera but this is not interresting for the revo at this point(High-Res Facial Appearance Capture from Polarized Smartphone Images - YouTube) So may be a dual mount rig (revo + dslr) with some cross polarisation filtering … but doing structured light process + photogrammetry process kill all the advantage to get things done quick.
I think about it since i baked the range with cross polarisation DSLR capture.
I did some test of this technique at home and it’s pretty easy to get very interesting result on reflective surfaces.
I don’t know if the future revopoint API will permit to get transformation matrix and timestamp to authorize automatic dslr camera placement in space. If so it should be pretty straightforward to use photogrametry software to re-project the images.
Else the “easyest” way should be to made a photogrammetry of the model just to align cameras…
If the revopoint developers could think about this use case for the API design we could have a very interesting tool to get very high quality texture on revopoint scans 
That what I requested since POP2…possible ? Yes very much but not at the same time while scanning .
The Scanner produces 12-18 photos per second using the build in RGB camera that changing the resolution will not improve anything , since the quality of color is based on the amount of the pixels (points) what right now is no more than at maximum 7x7 K
What would be possible is having an dongle that could capture the needed photos from the DSLR at 35 degrees each but after scanning .
Technically it would be impossible to replace the RGB with DSLR in real time scanning .
It would drag everything down and enormous data files will be produced .
Right now it can reach even 5GB so imagine with 4000 of DSLR pictures at minimal 35Megapixels on top of it .
And if I have to take pictures for photogrammetry, I would really not bother with 3D scanner at all , double job waste of time , makes really no sense to me .
But maybe the only solution for people that can’t texture their models other way for any reason .
I do texture my models manually in Zbrush for my work
And the RGB color per vertex the scanner produces is great for reference while doing so since Zbrush use RGB color per vertex ( poly paint ) for texturing process .
1 Like
There is effectively no way to do it that way.
But texture reconstruction doesn’t need thousand of dslr photos.
I think more about a system where, using the API we can get the current frame number during the scanning process every time we take a photo. Then, after the reconstruction process find the range matrix for each dslr shot.
Using some calibration tool and with a well built rig we can then find the dslr transformation matrix in object space
Using the best of the two worlds, structured light for volume recovery and big sensor and optic for high def textures.
Future improvement could be computer driven dslr shot to enable the range to take picture himself at the best moment… but that’s far future 
1 Like
I do that for 25 years on a daily basic , no it don’t .
12 pictures per 360 degree rotation would be already perfect for projection mode .
There is a lot of stuff that can be done to make it better , however a better optics on the scanner is not one of them . The texture mapping codes are not good enough to profit from anything else at this moment .
Here’s my video review of the Revopoint Range!
I included scans not shown earlier in this thread (including more humans), as well as 3D prints.
In a nutshell:
Nice product, structured light scanning really has come a long way. As someone who is a long-time photogrammetry user (about 10 years at this point!) it is REALLY refreshing to be able to preview scans in real time and process them in minutes, not hours.
If you plan to scan humans, the Turntable really is a must-have.
Textures are quite underwhelming (especially as a photogrammetry user), hopefully software upgrades will improve this. At the moment, you will want to do like @PUTV does and create textures yourself.
Video is in French, but English subtitles are available!
And as a bonus, there’s a Sketchfab link in the video description, if you want to check out the scans ! I’ll upload more in the future.
5 Likes
Excellent review Benoit, CC worked on spot …
Thanks for the props 
kind of you …
1 Like
This is actually really useful for me! Thanks 
1 Like