This is amazing. I wish I knew about this earlier, I had been looking into how to extract 3D data from Google Maps and saw some attempts by others that involved taking tons of screenshots and running them through photogrammetry, but the results were poor. This is really something else.
Thanks a lot for this! After some troubleshooting (the inject into process feature is now hidden by default, you have to activate it in the renderdoc settings) I got it to work, and it's very useful and a lot of fun!
from all the francophone community of blender I want to thank you and congratulate you for the opportunities you offer us to cross your work and I really hope that your talent can be better known I managed to make everything work you are bravo hero thank you so much
"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" Did not work RenderDoc 1.6 an existing connection was forcibly closed by remote host
Wow thank you very much for this solution. After days of trying this approach I figured out some tips that might be helpful for you if you struggle (Some of those adviced are already given by the author): 1) Use different Chrome/Chromium versions. Make sure to actually use the old ones because Chrome tends to update automatically. 2) Use RenderDoc 1.2 portable version. In more recent versions I couldn´t catch the materials/geometry of the scene and I had problems importing everything to Blender 3) Keep injecting the chrome gpu processes: While running RenderDoc and opening chrome I had to constantly inject a new process given by the chrome window. Keep injecting the new given process until google has loaded completly. 4) When installing the blender addon install the release zip. Thanks!
Thanks for summarizing this! I just added a link in the description to a version of chrome for which it works (70.0). And indeed beware not to let it autoupdate!
I had a few issues but now its working ;) its important to close all chrome instances running ! and i had to use render Doc 1.10, awesome tool, thanks again !
Took a couple tries to get the timing down but OMG this worked so well! I was struggling with trying to make a 3d model of denver and vegas and the data sets for lidar either didnt load, were huge files, were off limits to me because of covid, or they had crazy spiky artifacts i had to manually fix. Plus i was limited to the areas that were already scanned. BUT THIS...THIS IS JUST AMAZING! IN 30 minutes after finding your video, i had a full 3d model of the Vegas strip (current 2020 as well) open in blender ready to be made printable on my 3d printer... you are a god.
@Sky Dome Cause Google has 3D building data for actually only a little proportion of the globe (mostly North America and Europe, then major cities in Asia but a lot of areas are left unscanned so far.
@@ElieMichelFr meu amigo nao ta indo fiz de tudo voce poderia me ajudar ele aparece o crome gpu mas no pograma nao ativa pode ser o navegador ele e bem recente
Me gustaria contactar contigo, ya que por tus apellidos creo que eres de España y necesito ayuda para poder descargarme la geometria de mi ciudad. Conseguiste importar la geometria a Blender? Y las texturas?
@@JMMedinaDev I put exactly what he said and it worked, just right click google chrome > properties > then replace target with this: C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" if you click google chrome after doing this it should give a pop up saying GPU starting pid: (whatever it is) Also capturing the 3d models is a bit hit and miss i tried like 8 captures on one of the models till i actually got a working model
@@jond532 yeh I figured It out , I copied It wrong from the description box. I fixed the screenshot , setting a delay on the RenderDoc and increasing the frame count
I can confirm that this works with latest Blender (3.0) and latest RenderDoc (v1.17). The only difference is the need to manually enable injecting in the RenderDoc options because ability to inject is hidden by default in the newer versions of RenderDoc.
Can confirm this does not work anymore. I've spent a lot of time trying and renderdoc will just not inject properly. What happens is that you inject after opening chrome as stated, but then you head to google maps and the page goes blank again, so you have to inject again because renderdoc has just disconnected itself. So you try to inject again but every attempt immediately kicks out back out. I think this is the creator of renderdocs doing.
thankyou so much! - Can confirm works with latest chrome and blender 2.93, RenderDoc 1.13 Enable injecting in the RenderDoc options Tools> Settings Target script modified for Chrome 32bit: C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files \Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" Run the modified chrome then Inject into "chrome GPU" 04:46 - import settings max blocks Thanks so much for this tutorial
Hi, thank you so much for such an amazing tool! I was able to get it working with RenderDoc 1.10, Chrome 89 (current latest), Blender 2.82 and maps models importer 0.3.5 after trying many many things and losing hope.. for those who are struggling, don't lose hope, it does work eventually!
Every time I hit “ok” the my api doesn’t connect and it goes back to the default screen saying “NOTE: Injection must happen immediately after process starts before any graphics API has been initialized”
Superbe travail Elie !! Ta solution marche à la perfection, ça a été assez complexe mais le résultat est super intéressant pour réaliser de l'intégration paysagère avec lumion par la suite . Encore merci !!!
Couldn't get injection working with any version of RenderDoc and Chrome / Chromium. Tested with latest version of Opera browser and injection is working.
does it work with google anymore? I can't inject, chrome just sart as usual, but with the pid number window and render doc is estabilishing the connection, but then fail to inject.
Hi, this is amazing! Thank you. I have some quiestions. I am struggling with adapting the second line to the location of your Chrome installation. I never get to the point when i have my number of GPU. It saids "Windows cannot find the file" Could you help me with that?
When I press OK in the Chrome dialog box, RenderDoc stops injecting and returns to the selection screen for injection. In Chrome there's nothing at the top to indicate the process it still happening, tried a few methods and it still doesn't continue to inject after pressing OK?
To fix this, I adjusted the text to copy into the target-box in the chrome shortcut as follows so it looks like this (copy everything after the brackets, you'll need to adjust the filepaths to the ones on your pc)C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" --use-angle=gl
Maybe the same thing is happening to you as it did with me. I thought this method would work on both Google Earth and Google Maps but apparently it only works with the 3D mode of google maps!
@@IbizanHound2 I searched on google map but there are no 3d mode. Then I could find the function of globe but It's not 3d mode at all at the moment. The 3D mode is still working on your pc?
@@josephkim7831 Yes. Here is a quick video tutorial on how to do it. we.tl/t-le4j8MLMb6 Have you downloaded the version 70 of Chrome like Elie is saying from the link in the description of the video?
@@IbizanHound2 Thank you for your answer and sharing your video :) I installed Version 70.0.3538.77 (Official Build) (64-bit) I checked it on chrome setting I installed canary , But it doesn't work too ! I did those below 1. Enable hardware acceleration. Settings --> Advanced --> Enable hardware acceleration. 2. Enable two things at chrome://flags. One of them is "WebGL Draft Extensions" (halfway down) the other is "Override software rendering list" (at the top). 3. Close Chrome and make sure it is not running in the background. Re-open. Then, It didn't work still. Actually, I am from Korea. So, I changed website link "google.co.kr/maps/" to "www.google.com/maps/" I can't still see the 3D mode on google map. On my desktop, Intell(R) Core(TM) i5-6400 CPU "2.70GHz RAM 16.0 GB NVIDIA GeForce GTX 960 What about you?
Hi Elie, do you maybe know what to do when your chrome extension does not work? In my case when I make the shortcut and paste the target path chrome still opens normally without that extra window with the update. Do you maybe have a solution? Thanks :)
hi I used all the good versions according to github but when I put it on blender I have a very small strange shape. and I don't know why but my api is marked d3d11 I'm not opengl I don't know how to change it. if someone can help me maybe it's not the api the problem?
Hi Elie, this looks amazing. I'm just trying to get it to work but when I click on the OK on the PID window the tab in Render Doc just closes. I've tried lots of version of Render Doc 1.13, 1.2, 1.11 and latest but same result.......any idea what's causing the issue. Many thanks
Heyyy, you've made a great tutorial. But, i cannot import my file into Blender. It just says "Invalid RDC Capture File" And i have to use Google Earth instead. Any help please?
Get blender 2.83 LTS and install the google maps importer addon there. It hasn't been updated for 2.91 yet. Or just use google earth, if it didn't work
Whenever I try to run the Chrome shortcut, it gives me the PID popup, but it also loads the Chrome tab, not the white screen. I am pretty sure this is what is causing it as I don't ever see just the blank white screen and I dont get the "debug" window. Any help?
This is the best Tutorial on this topic it's like it was a hidden secret, i wonder why they made it so difficult to understand. Thanks # Élie #Michel you are the Best....
Sebastian Schimper hi, can you help me with this? I cannot get it working. Could you get me a map and I pay you for it? I need ‘blythe valley business park’ in Solihull, UK. Email me on nathbrady789@hotmail.com I can pay cash, or with a short animation intro or something.
Hi, Great video! I am following AltArch version (taken from yours) When I import the .rdc capture to blender, it is just a blank flat solid. Not even buildings are shown. No contours, no textures, nothing. Just a flat solid. I have tried uninstalling everything, and also with different versions of everything. Everything works until I import into blender. I also changed the box limit to various numbers and still nothing. 200, 500, etc. It is a small area I captured tho, like 0.4 km2. The captures themselves look ok by themselves. Any help? Regards, M
Check out the very new release: github.com/eliemichel/MapsModelsImporter/releases/tag/v0.2.0-rc1 Your issue is among the one this vew version intends to tackle!
Here my try. Software: - RenderDoc 1.2 portable - Vivaldi latest version 2.8.1664.40 - Blender 2.8 - Maps Models Importer v0.1.2 Hardware: - Nvidia gtx 960m 1) Started nvidia control center (found into old windows control panel) -> manage 3d settings -> (tab) program settings -> added vivaldi -> selected Nvidia gpu 2) Addeded --disable-gpu-sandbox --gpu-startup-dialog --use-angle=gl to vivaldi launcher as showed for chrome in the video Followed the video istructions worked fine until blender import. Restarting windows helped, everything works fine, the capture is perfectly loaded. Thanks!
Thank you for an amazing tutorial. I am struggling with the scale of the model. It seems like every capture is a different scale and has to be visually scaled to an aerial image. Even if i use unitless scale throughout the process there's no logic that gets it to a true 1:1 scale. When i heard your french accent, i thought metric, but that seems to just complicate the scaling issue. Any idea of what unit or scale RenderDoc captures from Google maps or is that completely arbitrary depending on your zoom factor? For testing purposes off course!
This is great, but i am working on it since yesterday and still doesn't work... I've tried Chrome, opera and vivaldi on every version,main recurring problem, after i use the line: chrome.exe --disable-gpu-sandbox --gpu-startup-dialog --use-angle=gl --disable-extensions --disable-plugins the "OpenGLES, F12, PrtScrn" thing is not present in Chrome and also, in RenderDoc the API line says: "OpenGLES (Not Presenting)" What's wrong? Ismy graphic card unsupported ? ive a GTX 960 M
in this point: Make sure both RenderDoc and Chrome run on the same GPU (if you have a laptop with integrated + dedicated graphics devices) how can i do this possible . i have a desktop with dedicated graphics , not sure how to make they both run in the same GPU. the main problem is i have edited the google chrome path and i have the GPU notification but i dont have the chrome GPU api nor the OpenGLes black dialog box up top.
@@ElieMichelFr I put C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" on the link but nothing works in API
This method used to work with windows 10 but since I've installed 11 it's does allow to make the changes on the properties of chrome. Does anyone knows how to solve this problem?
Hello, I tried everything you said but it still give me an error and I am sure that it has to to with the capturing process in renderdoc. I tried your samples and it works fine. Is there a way I can email you, so you can take a look at what's wrong with my capture file.
Not sure if you're still facing an issue - try using v1.25 of renderdoc, or try different versions - as long as you see the black header in google chrome, you're good and as long as the API isn't none then everything should work. Also if your files aren't rendering, or coming up blank when loading onto blender, try downloading larger areas, the bigger files downloaded and worked quite well for me, all my rdc files are >80mb and once loaded onto blender, you can filter out the tiles you don't want. Initially, I loaded 17/18mb files and they didn't load at all, but I checked some thread on the internets and someone mentioned it worked for them after using a much larger area. Anyways - goodluck!
Hi, I installed renderdoc 1.2 because with the portable i coundn't inject and i am using vivaldi with a nvidia 1070. Now I am stuck in the capture since i am not being able to capture anything even when everythings seems okay. Any suggest ? Ty for the video :)
Mine too. I was working for 4 hours straight on this and I cannot seem to get chrome 70x and renderdoc 1.2, 1.6 or 1.8 to work. Nor can I get other browsers to work. Is there a more brute force way to get this done? Like the link you shared with more information in your blog? I reallly want this to work but ithe capture frames pops up for 2 500 milliseconds and goes away immediately, nothing happens even if I race to click it. I have to close out at least 5 different GPU /PID (whatever it is called) diaglog boxes for chrome. Any ideas?
I am using Blender 2.82 / Renderdoc 1.6 After injecting I open Chrome browser but no pop up appears or Chrome Gpu in renderdoc, Do you know why? Thank you
upvote for this guy. i've heard api change's on chrome's side broke capturing, giving invalid rdc files unable to import, however this older release seems to be working. thanks
OK here was my process whcih worked eventually: Needed Blender 2.8 ( the beta) not 2.79 . (sorry i didnt listen!) Needed RenderDoc 1.2 portable. (older version) . Had disable AV cos of abp.exe it didnt like. Also had to click on the Vulkan is missing, then it adds it I made a mistake of putting the flags inside the quotes should be outside "C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --disable-gpu-sandbox --gpu-startup-dialog --use-angle=gl --disable-extensions --disable-plugins In fact the --disable-gpu-sandbox was needed - even though chrome says it was an unrecognised flag - otherwise PrntScr didnt work
hey, i found it! go to tools > settings > general -> at the bottom u have to select "enable process injection..." and then you restart your RenderDoc and once u restart it it will show up on the top left
holy shit at first i couldnt get it to work but i read through the github faq at the bottom and got it working! I love this. Going to use it to make an awesome game
I have some problem whiles changing my google chrome target , after i changed and I try to open it and it came windows cannot find 'C:/ etc .. Any help ples
First of all Élie, thank you for all your work! I have used the photogrammetry approach before and was playing with this since last year, awesome and the LilyCaptureMerger will make it even easier. Is there a way to get rid of the edges/borders of each tiles while capturing? What is the exact (real world) size of the tiles - any tile size. I cannot seem to find info anywhere, or did I just missed something. Thank you.
Hi Tamas, what do you mean by edges of each tiles? If you want to merge them together, you can Ctrl+J them, then Alt+M > By Distance to fuse common points. The real world scale unfortunately depends on the capture, but 1. LilyCaptureMerger can handle this 2. afaik once you measured it once you can multiply it by 2 or 4 or ... or 1/2, 1/4 if it is different for another one.
@@ElieMichelFr Just bought LilyCaptureMerger and it works very well with the newest Blender and Chrome browser. Still I have the "borders". All the rectangular areas - no matter what size/height it was captured from - if you zoom in in Blender, have a thin border, like if you imagine they were of paper and the borders were bent down. Unfortunately I cannot post screenshots here...
My renderdoc always goes back to the pre-inject page when I click "ok" on the chrome tab, then chrome just generates another code. I can't get past this point, I believe I've used the recommended versions, and have tried different versions as well, always with the same result. Any idea what I've got wrong?
In blender 3.6 i have this Error: Failed to load the RenderDoc Module. Please report to MapsModelsImporter developers providing the full console log with debug information. Any solution for this problem?
Bonjour ELie et merci pour ce formidable outil. Superbe pont de Cheviré :) Je suis donc arrivé grâce à toi à ma première étape en important le modèle 3D de Google vers Blender. Comment augmenter le niveau de détails (LOD) lors de la capture avec RenderDoc ? Quelle procédure recommandes-tu pour en faire ensuite l'export vers MSFS2020 ? A te lire.
Pour augmenter le niveau de détail tu peux déjà dézoomer dans le navigateur (ctrl+molette) pour faire croire à google maps que tu as un plus grand écran. Ensuite si ce n'est pas assez tu peux utiliser LilyCaptureMerger eliemichel.gumroad.com/l/KSvXuu pourb fusionner plusieurs captures
hello Elie, thanks a lot for the great workflow. It appears that maybe Google now is blocking the injection. Tried it on all different browser and my guess is there is a safeguard in the maps itself. Do you have any ideas on how to work around it? Thank you for looking into this!
@@KashishSharma93 shift + esc on chorme brings up taskmanager for chrome specifically. There you can find the PID for gpu. But there is some code(I gather) in new version of chrome especially to block the process.
This is amazing. I wish I knew about this earlier, I had been looking into how to extract 3D data from Google Maps and saw some attempts by others that involved taking tons of screenshots and running them through photogrammetry, but the results were poor. This is really something else.
Thanks a lot for this! After some troubleshooting (the inject into process feature is now hidden by default, you have to activate it in the renderdoc settings) I got it to work, and it's very useful and a lot of fun!
Thanks dude. That was my quetion
wish i could upvote this! very helpful reply w/ the new version of renderdoc
how can I activate it? please help
@@shidrokhsedaghat6946 Tools>Settings>Enable process injection
@@shidrokhsedaghat6946go to tools --- settings --- general tab --- click enable process injection
from all the francophone community of blender I want to thank you and congratulate you for the opportunities you offer us to cross your work and I really hope that your talent can be better known I managed to make everything work you are bravo hero thank you so much
This is one of the craziest hacks I've seen
@Alvin HOW
@Alvin thanks I will try it!
"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" Did not work RenderDoc 1.6 an existing connection was forcibly closed by remote host
Died laughing when i saw him using RenderDoc for this =))
Genius!
This is quite helpfu. I used Blender 2.82 and I used it for a downtown of Tulsa of my job. You did a great job with this.
Tulsa? Are you making a Watchmen series 😂?
Élie, thank you so much! this is by far the best/ clearest tutorial for such workflow I have seen so far.
An extremely helpful tutorial! Thank you very much for taking the time to share it with everyone. Great job!!
I'm trying to accomplish this for the last 5 years!! Thanks a lot, dude.
Wow thank you very much for this solution. After days of trying this approach I figured out some tips that might be helpful for you if you struggle (Some of those adviced are already given by the author): 1) Use different Chrome/Chromium versions. Make sure to actually use the old ones because Chrome tends to update automatically. 2) Use RenderDoc 1.2 portable version. In more recent versions I couldn´t catch the materials/geometry of the scene and I had problems importing everything to Blender 3) Keep injecting the chrome gpu processes: While running RenderDoc and opening chrome I had to constantly inject a new process given by the chrome window. Keep injecting the new given process until google has loaded completly. 4) When installing the blender addon install the release zip.
Thanks!
Thanks for summarizing this! I just added a link in the description to a version of chrome for which it works (70.0). And indeed beware not to let it autoupdate!
Whats a portable version of render doc? The one that does require installation ?
Thanks for making this tool. I'm still trying to get it to work but it is encouraging that there are resources like this and your github
It is working for me now. This is an amazing tool, thank you.
In case this helps anybody:
Blender 3.6.2
RenderDoc 1.25
MapsModelsImporter 0.6.2
2:22 My chrome window is not getting the debug overlay but I followed all your steps. Using Renderdoc 1.9 and the latest version of chrome
Did you resolve this? I have the same issue and cannot see a fix.
@@rickyleaburke Oh I gave up 4 months ago lol
I'll be recreating my father's old apartment in Queens and desperately needed the background imagery from the 15th floor. This is amazing.
Eu estou modelando um apartamento também, isso é ótimo!
this is amazing, i was actually doing some photogrammetry on-site and this blender-add-on helped me position it.
Awesome!!!!
It took some cursing and a couple of hours trying browsers but eventually worked like a charm.
Thank you very very much for this!
Which browser did you end up using?
Can confirm this works 11/21/2019. My setup:
- GTX 1060
- latest Windows
- Chrome 70.0.3538.77
- RenderDoc 1.2 Portable
- Blender 2.80
- MapsModelsImporter 0.1.2
Khannor VG i have a 2080Ti SLI
16gb graphic cards
I had a few issues but now its working ;) its important to close all chrome instances running ! and i had to use render Doc 1.10, awesome tool, thanks again !
Took a couple tries to get the timing down but OMG this worked so well! I was struggling with trying to make a 3d model of denver and vegas and the data sets for lidar either didnt load, were huge files, were off limits to me because of covid, or they had crazy spiky artifacts i had to manually fix. Plus i was limited to the areas that were already scanned.
BUT THIS...THIS IS JUST AMAZING! IN 30 minutes after finding your video, i had a full 3d model of the Vegas strip (current 2020 as well) open in blender ready to be made printable on my 3d printer... you are a god.
how did you get it printable?!?! i get a mess that wont slice...
Now Imagine using this to make a 3D printed miniature of Egypt. too bad there's no 3d data for Egypt ;;
@Sky Dome Cause Google has 3D building data for actually only a little proportion of the globe (mostly North America and Europe, then major cities in Asia but a lot of areas are left unscanned so far.
@@ElieMichelFr meu amigo nao ta indo fiz de tudo voce poderia me ajudar ele aparece o crome gpu mas no pograma nao ativa pode ser o navegador ele e bem recente
my inject into the process does not work. Once I click on OK on pid popup, it goes back to the same page to add a new pid. What do I do?
Use an older version, use 1.9 of renderdoc and see if it works
@@mahuba2553 I have the same question. I installed 1.9 of renderdoc but I could not open it. Maybe it does not support windows11?
This is probably the most useful hack I have ever come across. Perfect for using it to replicate a section of a city ♥
Thank you very much for the addon and the explanation! Amazing!
Me gustaria contactar contigo, ya que por tus apellidos creo que eres de España y necesito ayuda para poder descargarme la geometria de mi ciudad.
Conseguiste importar la geometria a Blender?
Y las texturas?
I just want to say this is a very cool idea and i'm impressed that it actually works
I jumped from my sofa when a realized that I live 500m away from that bridge XD
Lmao
I followed exactly as you said using the lastest version and all worked perfectly thankyou!! Got exactly what i wanted in less than 20 minutes!
Dont you have problems adding the target to the chrome shortcut? what did you exactly put in?
@@JMMedinaDev I put exactly what he said and it worked, just right click google chrome > properties > then replace target with this:
C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog"
if you click google chrome after doing this it should give a pop up saying GPU starting pid: (whatever it is)
Also capturing the 3d models is a bit hit and miss i tried like 8 captures on one of the models till i actually got a working model
@@jond532 yeh I figured It out , I copied It wrong from the description box. I fixed the screenshot , setting a delay on the RenderDoc and increasing the frame count
I can confirm that this works with latest Blender (3.0) and latest RenderDoc (v1.17). The only difference is the need to manually enable injecting in the RenderDoc options because ability to inject is hidden by default in the newer versions of RenderDoc.
Lifesaver!
Was just searching for that option
Which version of MapModels Importer do you use?
Which version of MapModels Importer do you use?
@@crazyblaster8195 0.4.0
@@mariokotlar303 i can't install the addon ? you know how can i get it ?
Can confirm this does not work anymore. I've spent a lot of time trying and renderdoc will just not inject properly. What happens is that you inject after opening chrome as stated, but then you head to google maps and the page goes blank again, so you have to inject again because renderdoc has just disconnected itself. So you try to inject again but every attempt immediately kicks out back out.
I think this is the creator of renderdocs doing.
thankyou so much! - Can confirm works with latest chrome and blender 2.93, RenderDoc 1.13
Enable injecting in the RenderDoc options Tools> Settings
Target script modified for Chrome 32bit:
C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files \Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog"
Run the modified chrome then Inject into "chrome GPU"
04:46 - import settings max blocks
Thanks so much for this tutorial
thanks!
thank you so much
Hi, thank you so much for such an amazing tool!
I was able to get it working with RenderDoc 1.10, Chrome 89 (current latest), Blender 2.82 and maps models importer 0.3.5 after trying many many things and losing hope.. for those who are struggling, don't lose hope, it does work eventually!
Every time I hit “ok” the my api doesn’t connect and it goes back to the default screen saying “NOTE: Injection must happen immediately after process starts before any graphics API has been initialized”
@Lollo yeah go to your task manager and close every chrome tab that pops up, for some reason there is like 5-6 that run in the background
Superbe travail Elie !! Ta solution marche à la perfection, ça a été assez complexe mais le résultat est super intéressant pour réaliser de l'intégration paysagère avec lumion par la suite . Encore merci !!!
Couldn't get injection working with any version of RenderDoc and Chrome / Chromium. Tested with latest version of Opera browser and injection is working.
When i inject theres no api in render doc and no overlay in google chrome...
does it work with google anymore? I can't inject, chrome just sart as usual, but with the pid number window and render doc is estabilishing the connection, but then fail to inject.
Hi, this is amazing! Thank you. I have some quiestions.
I am struggling with adapting the second line to the location of your Chrome installation. I never get to the point when i have my number of GPU. It saids "Windows cannot find the file" Could you help me with that?
When I press OK in the Chrome dialog box, RenderDoc stops injecting and returns to the selection screen for injection. In Chrome there's nothing at the top to indicate the process it still happening, tried a few methods and it still doesn't continue to inject after pressing OK?
To fix this, I adjusted the text to copy into the target-box in the chrome shortcut as follows so it looks like this (copy everything after the brackets, you'll need to adjust the filepaths to the ones on your pc)C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" --use-angle=gl
This is excellent and has a huge amount of potential for its application.
4:50 when i try to import "rcd file" (45 MB), nothing happend in Blender 2.8.
same here :(
Maybe the same thing is happening to you as it did with me. I thought this method would work on both Google Earth and Google Maps but apparently it only works with the 3D mode of google maps!
@@IbizanHound2 I searched on google map but there are no 3d mode. Then I could find the function of globe but It's not 3d mode at all at the moment. The 3D mode is still working on your pc?
@@josephkim7831 Yes. Here is a quick video tutorial on how to do it. we.tl/t-le4j8MLMb6
Have you downloaded the version 70 of Chrome like Elie is saying from the link in the description of the video?
@@IbizanHound2 Thank you for your answer and sharing your video :)
I installed Version 70.0.3538.77 (Official Build) (64-bit)
I checked it on chrome setting
I installed canary ,
But it doesn't work too !
I did those below
1. Enable hardware acceleration. Settings --> Advanced --> Enable hardware acceleration.
2. Enable two things at chrome://flags. One of them is "WebGL Draft Extensions" (halfway down) the other is "Override software rendering list" (at the top).
3. Close Chrome and make sure it is not running in the background. Re-open.
Then, It didn't work still.
Actually, I am from Korea. So, I changed website link
"google.co.kr/maps/" to "www.google.com/maps/"
I can't still see the 3D mode on google map.
On my desktop,
Intell(R) Core(TM) i5-6400 CPU "2.70GHz
RAM 16.0 GB
NVIDIA GeForce GTX 960
What about you?
Thank you for this! As many of the others on this thread i was looking for years for a method like this one!
Hi Elie, do you maybe know what to do when your chrome extension does not work? In my case when I make the shortcut and paste the target path chrome still opens normally without that extra window with the update. Do you maybe have a solution? Thanks :)
I was looking for something like this for years. Thank you so much!
hi I used all the good versions according to github but when I put it on blender I have a very small strange shape. and I don't know why but my api is marked d3d11 I'm not opengl I don't know how to change it. if someone can help me maybe it's not the api the problem?
Guy this is fix I just change the language in blender to French to English
This is officially wholesome! And thank you for the merger!
Hi Elie, this looks amazing. I'm just trying to get it to work but when I click on the OK on the PID window the tab in Render Doc just closes. I've tried lots of version of Render Doc 1.13, 1.2, 1.11 and latest but same result.......any idea what's causing the issue. Many thanks
me too...
same
You must close all other chrome pages that are open except for the one you are working on!
the chrome shortcut isnt working for me lol
Heyyy, you've made a great tutorial. But, i cannot import my file into Blender. It just says "Invalid RDC Capture File" And i have to use Google Earth instead. Any help please?
Get blender 2.83 LTS and install the google maps importer addon there. It hasn't been updated for 2.91 yet. Or just use google earth, if it didn't work
Followed the instructions, but I don't get a PID popup. Is my Chrome to new ? Or is there another problem (I use RenderDoc 1.31) ?
Whenever I try to run the Chrome shortcut, it gives me the PID popup, but it also loads the Chrome tab, not the white screen. I am pretty sure this is what is causing it as I don't ever see just the blank white screen and I dont get the "debug" window. Any help?
I have the same problem :(
dude when i do the copy and paste thing it doesnt show the pid and the location is exactly the same too
This is the best Tutorial on this topic it's like it was a hidden secret, i wonder why they made it so difficult to understand. Thanks #
Élie #Michel you are the Best....
Worked great with my laptop, I'm so happy, great work dude!
Sorry, what kind of GPU and version of Renderdoc, GGchrome do you use? Thank you!
Hey Élie Michel, this is awesome work! Thanks so much!
Sebastian Schimper hi, can you help me with this? I cannot get it working. Could you get me a map and I pay you for it? I need ‘blythe valley business park’ in Solihull, UK.
Email me on nathbrady789@hotmail.com
I can pay cash, or with a short animation intro or something.
Hi, I cant get this working. can I pay you to get a 3d model of a specific location for me?
Hi,
Great video! I am following AltArch version (taken from yours)
When I import the .rdc capture to blender, it is just a blank flat solid. Not even buildings are shown. No contours, no textures, nothing. Just a flat solid.
I have tried uninstalling everything, and also with different versions of everything. Everything works until I import into blender.
I also changed the box limit to various numbers and still nothing. 200, 500, etc.
It is a small area I captured tho, like 0.4 km2.
The captures themselves look ok by themselves.
Any help?
Regards,
M
After Inject in RenderDoc it says
Connection Status: Established
API: OpenGLES (Not Presenting)
Help.
Check out the very new release: github.com/eliemichel/MapsModelsImporter/releases/tag/v0.2.0-rc1 Your issue is among the one this vew version intends to tackle!
the Chrome shortcut won't work for me. Chrome starts but I don't get the "pause"-button with the PID
not gpu process starting on chrome with asus strix 1070, not working
Either the PID number changes, the geometry in-browser is corrupted, or the capture immediately button doesn't work
Here my try.
Software:
- RenderDoc 1.2 portable
- Vivaldi latest version 2.8.1664.40
- Blender 2.8
- Maps Models Importer v0.1.2
Hardware:
- Nvidia gtx 960m
1) Started nvidia control center (found into old windows control panel) -> manage 3d settings -> (tab) program settings -> added vivaldi -> selected Nvidia gpu
2) Addeded --disable-gpu-sandbox --gpu-startup-dialog --use-angle=gl to vivaldi launcher as showed for chrome in the video
Followed the video istructions worked fine until blender import.
Restarting windows helped, everything works fine, the capture is perfectly loaded.
Thanks!
I was going crazy trying to get this to work with the latest versions, and going back to the versions you posted worked for me. Thank you!!
@@FelipeTorrents thanks, only vivaldi work, i've tried many chrome version that didn't worked.
when i press ok the injection fails
Did you solved the problem?
No
you are a modern Prometheus man!!!
la carpeta de --C:\Program Files (x86)\Google\Chrome\Application\chrome.exe-- no existe =(((
Thank you for an amazing tutorial. I am struggling with the scale of the model. It seems like every capture is a different scale and has to be visually scaled to an aerial image. Even if i use unitless scale throughout the process there's no logic that gets it to a true 1:1 scale. When i heard your french accent, i thought metric, but that seems to just complicate the scaling issue. Any idea of what unit or scale RenderDoc captures from Google maps or is that completely arbitrary depending on your zoom factor? For testing purposes off course!
i dont get a pop-up with the tag added in the shortcut so I can't get the PID
This is great, but i am working on it since yesterday and still doesn't work...
I've tried Chrome, opera and vivaldi on every version,main recurring problem, after i use the line:
chrome.exe --disable-gpu-sandbox --gpu-startup-dialog --use-angle=gl --disable-extensions --disable-plugins
the "OpenGLES, F12, PrtScrn" thing is not present in Chrome and also, in RenderDoc the API line says: "OpenGLES (Not Presenting)"
What's wrong?
Ismy graphic card unsupported ? ive a GTX 960 M
It works great! Love it. Thank you so much for sharing
thank you so much! question about scale: when importing the 3d model into blender, is there a way to know at which scale the model is imported?
in this point: Make sure both RenderDoc and Chrome run on the same GPU (if you have a laptop with integrated + dedicated graphics devices) how can i do this possible . i have a desktop with dedicated graphics , not sure how to make they both run in the same GPU. the main problem is i have edited the google chrome path and i have the GPU notification but i dont have the chrome GPU api nor the OpenGLes black dialog box up top.
Hi, I clicked "OK" for GPU code but after I got API: None (Capture Frame not functioning)
How should I do ?
Using the right shortcut for starting chrome? Ensuring no chrome process was running?
@@ElieMichelFr I put C:\Windows\System32\cmd.exe /c "SET RENDERDOC_HOOK_EGL=0 && START "" ^"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe^" --disable-gpu-sandbox --gpu-startup-dialog" on the link but nothing works in API
@@kennystewart7901 me too. Would the path have changed?
did you both ensure that chrome was completely stopped before doing so?
@@ElieMichelFr I ended up pulling it off. As it turns out, you need to accept the prompt to get it started
This method used to work with windows 10 but since I've installed 11 it's does allow to make the changes on the properties of chrome. Does anyone knows how to solve this problem?
This works great!!!! thanks to all who helped to create this. One question, does anyone know what the real world size/scale is of a block?
a Minecraft block is equivalent to 1 cubic meter
Hello, I tried everything you said but it still give me an error and I am sure that it has to to with the capturing process in renderdoc. I tried your samples and it works fine. Is there a way I can email you, so you can take a look at what's wrong with my capture file.
Not sure if you're still facing an issue - try using v1.25 of renderdoc, or try different versions - as long as you see the black header in google chrome, you're good and as long as the API isn't none then everything should work. Also if your files aren't rendering, or coming up blank when loading onto blender, try downloading larger areas, the bigger files downloaded and worked quite well for me, all my rdc files are >80mb and once loaded onto blender, you can filter out the tiles you don't want. Initially, I loaded 17/18mb files and they didn't load at all, but I checked some thread on the internets and someone mentioned it worked for them after using a much larger area. Anyways - goodluck!
Hi, I installed renderdoc 1.2 because with the portable i coundn't inject and i am using vivaldi with a nvidia 1070. Now I am stuck in the capture since i am not being able to capture anything even when everythings seems okay. Any suggest ? Ty for the video :)
I'm stuck with the same problem. It hangs up a bit like in the video but no capture is saved and is still stuck at "0 captures saved"
Same problem
renderdoc only has "Attach to running instance" there is no "inject into process"
My window with the "capture frames" also closes when i want to do it
Mine too. I was working for 4 hours straight on this and I cannot seem to get chrome 70x and renderdoc 1.2, 1.6 or 1.8 to work. Nor can I get other browsers to work. Is there a more brute force way to get this done? Like the link you shared with more information in your blog? I reallly want this to work but ithe capture frames pops up for 2 500 milliseconds and goes away immediately, nothing happens even if I race to click it. I have to close out at least 5 different GPU /PID (whatever it is called) diaglog boxes for chrome. Any ideas?
5:30 chicken? :D
I am using Blender 2.82 / Renderdoc 1.6 After injecting I open Chrome browser but no pop up appears or Chrome Gpu in renderdoc, Do you know why?
Thank you
Try to run google as an adminstrator
I had the same issue. This video helped me to solve it.
ua-cam.com/video/F_XsmoZJmG8/v-deo.html
You have truly helped me out with this! Thanks a lot!
For Download of RenderDoc: renderdoc.org/stable/1.7/RenderDoc_1.7_64.msi
upvote for this guy. i've heard api change's on chrome's side broke capturing, giving invalid rdc files unable to import, however this older release seems to be working. thanks
@@johnline welcome
I dont see "inject process" under File menu :(
Le pont de Cheviré à Nantes! (FRANCE) J'ai immédiatement reconnu! ^_^
Wow. This is really amazing. Can't wait to 3D print my neighborhood !
OK here was my process whcih worked eventually:
Needed Blender 2.8 ( the beta) not 2.79 . (sorry i didnt listen!)
Needed RenderDoc 1.2 portable. (older version) . Had disable AV cos of abp.exe it didnt like. Also had to click on the Vulkan is missing, then it adds it
I made a mistake of putting the flags inside the quotes
should be outside
"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe" --disable-gpu-sandbox --gpu-startup-dialog --use-angle=gl --disable-extensions --disable-plugins
In fact the --disable-gpu-sandbox was needed - even though chrome says it was an unrecognised flag - otherwise PrntScr didnt work
Thanks for this detailed feedback :) Those are mistakes others might check not to do!
my api says D3D11 (active) i do not see open GL Please Help
Man, there isnt Injet into process in RenderDoc 1.31
hey, i found it! go to tools > settings > general -> at the bottom u have to select "enable process injection..." and then you restart your RenderDoc and once u restart it it will show up on the top left
Print Screen does not work :/
Press "Capture Frames Immediately"
hi, is the same versions of renderdoc and blender still working till now? I m receiving an error when I try to put the capture in blender
Such accessibility is game changer :)
thanks for sharing quick and efficient
The connection status is established but the API states: "none"
in which version appears the inyect into process, i been triying in many versions butit doesn´t appear
same
holy shit at first i couldnt get it to work but i read through the github faq at the bottom and got it working! I love this. Going to use it to make an awesome game
I have some problem whiles changing my google chrome target , after i changed and I try to open it and it came windows cannot find 'C:/ etc .. Any help ples
First of all Élie, thank you for all your work! I have used the photogrammetry approach before and was playing with this since last year, awesome and the LilyCaptureMerger will make it even easier.
Is there a way to get rid of the edges/borders of each tiles while capturing?
What is the exact (real world) size of the tiles - any tile size. I cannot seem to find info anywhere, or did I just missed something.
Thank you.
Hi Tamas, what do you mean by edges of each tiles? If you want to merge them together, you can Ctrl+J them, then Alt+M > By Distance to fuse common points. The real world scale unfortunately depends on the capture, but 1. LilyCaptureMerger can handle this 2. afaik once you measured it once you can multiply it by 2 or 4 or ... or 1/2, 1/4 if it is different for another one.
@@ElieMichelFr Just bought LilyCaptureMerger and it works very well with the newest Blender and Chrome browser.
Still I have the "borders". All the rectangular areas - no matter what size/height it was captured from - if you zoom in in Blender, have a thin border, like if you imagine they were of paper and the borders were bent down. Unfortunately I cannot post screenshots here...
@@tamaseichert8752 Oh yeah I see what you're talking about now. There is no way to get rid of it (not without non trivial mesh processing at least) :/
Thank you Élie, no worries, thought so that it is how it's coming out from the GPU.
I am not getting anything when i import, any idea why? I followed exactly as said here
My renderdoc always goes back to the pre-inject page when I click "ok" on the chrome tab, then chrome just generates another code. I can't get past this point, I believe I've used the recommended versions, and have tried different versions as well, always with the same result. Any idea what I've got wrong?
same here. @Élie Michel any idea?
@@kidoctulus01 close all other chrome pages
Cool! Is it possible to export the model with a coordinate system (so that it can be georeferenced)?
Hi, I have followed the steps but when I open the google shortcut, the PID doesn't show & it just loads normally. Do you know how I can fix this?
Close chrome) Then open it with GPU hacking, you can use Firefox to contine watching tutorials
In blender 3.6 i have this Error: Failed to load the RenderDoc Module.
Please report to MapsModelsImporter developers providing the full console log with debug information.
Any solution for this problem?
Hey Elie, do you plan on updating the addon for blender 3.1?
Bonjour ELie et merci pour ce formidable outil. Superbe pont de Cheviré :)
Je suis donc arrivé grâce à toi à ma première étape en important le modèle 3D de Google vers Blender.
Comment augmenter le niveau de détails (LOD) lors de la capture avec RenderDoc ?
Quelle procédure recommandes-tu pour en faire ensuite l'export vers MSFS2020 ?
A te lire.
Pour augmenter le niveau de détail tu peux déjà dézoomer dans le navigateur (ctrl+molette) pour faire croire à google maps que tu as un plus grand écran. Ensuite si ce n'est pas assez tu peux utiliser LilyCaptureMerger eliemichel.gumroad.com/l/KSvXuu pourb fusionner plusieurs captures
@@ElieMichelFr Houa hou ! je viens de visionner la vidéo de LilyCaptureMerger ... impressionnant !
Can't find the inject into process on RenderDoc :/
When in renderdoc go to Tools>Settings>General>Enable process injection. For some reason it is not on by default
hello Elie, thanks a lot for the great workflow. It appears that maybe Google now is blocking the injection. Tried it on all different browser and my guess is there is a safeguard in the maps itself. Do you have any ideas on how to work around it? Thank you for looking into this!
same issue. cant see the chrome gpu in renderdoc
@@KashishSharma93 shift + esc on chorme brings up taskmanager for chrome specifically. There you can find the PID for gpu. But there is some code(I gather) in new version of chrome especially to block the process.
@@Patheticbutharmless any workaround that block? i'm stuck in that error still
@@leephelipe Sorry. I totally gave up on the tech. As of now.