Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - chadfx

Pages: [1] 2 3 ... 14
Bug Reports / Re: GPU error CUDA_ERROR_OUT_OF_MEMORY (2) at line 123
« on: September 05, 2018, 12:58:34 PM »
Hi, just one last update. With the latest graphics driver update from Nvidia last week (v399.07) the problem seems to have been resolved.

Cheers, -Chad

Bug Reports / Re: GPU error CUDA_ERROR_OUT_OF_MEMORY (2) at line 123
« on: August 23, 2018, 10:31:23 PM »
I think there must be some issue between these two graphics cards. I have another piece of software which also is having semi-cryptic error messages with both cards. I actually have to pull out the 780 in order to get the other software to work.

I am not much of a PC guy, so I feel very limited in my troubleshooting abilities in this area. I'm happy to follow instructions if you have some ideas of where I can look for issues.

Cheers, -Chad

Bug Reports / Re: GPU error CUDA_ERROR_OUT_OF_MEMORY (2) at line 123
« on: August 23, 2018, 10:06:00 PM »
Hello Alexey, thanks for following up on this.

I just tried to run the Build Dense Cloud phase which also failed, but it did seem to try and do a few steps of reconstructing depth before it failed. The console said that the 780 did 100% of the calculation and the 1070ti 0%.

It also gave the same error, but with an additional error; "Warning: cudaStreamDestroy failed: out of memory (2)

I'm running 1.4.3 build 6529

Bug Reports / Re: GPU error CUDA_ERROR_OUT_OF_MEMORY (2) at line 123
« on: August 23, 2018, 08:34:57 PM »
I've since updated the NVidia driver to 398.82 and I am still getting the error if both graphics cards are enabled.

Bug Reports / Re: GPU error CUDA_ERROR_OUT_OF_MEMORY (2) at line 123
« on: August 20, 2018, 04:23:09 PM »
Hitting the exact same error. Nothing fancy in my setup, but I did swap out a GTX 780 for a 1070ti recently.
If I disable the 780, leaving just the 1070ti, it runs fine.

The project is a small number of moderate res TIF images and has run fine in the past.

NVidia driver is 398.36 in Win764

 AlignPhotos: accuracy = Highest, preselection = generic, keypoint limit = 80000, tiepoint limit = 20000, apply masks = 0, filter tie points = 0, adaptive fitting = 1
2018-08-20 15:16:25 Matching photos...
2018-08-20 15:16:25 Detecting points...
2018-08-20 15:16:25 Using device: GeForce GTX 1070 Ti, 19 compute units, 8192 MB global memory, compute capability 6.1
2018-08-20 15:16:25   driver version: 9020, runtime version: 5050
2018-08-20 15:16:25   max work group size 1024
2018-08-20 15:16:25   max work item sizes [1024, 1024, 64]
2018-08-20 15:16:25 Using device: GeForce GTX 780, 12 compute units, 6144 MB global memory, compute capability 3.5
2018-08-20 15:16:25   driver version: 9020, runtime version: 5050
2018-08-20 15:16:25   max work group size 1024
2018-08-20 15:16:25   max work item sizes [1024, 1024, 64]
2018-08-20 15:16:26 [GPU] photo 0: 80000 points
2018-08-20 15:16:26 Finished processing in 1.529 sec (exit code 0)
2018-08-20 15:16:26 Error: CUDA_ERROR_OUT_OF_MEMORY (2) at line 123

General / Re: Raw file batch calibration/sync of image exposure?
« on: July 11, 2015, 01:30:11 AM »
+1 for bigben's approach (or do the same in Lightroom)'s usually fairly quick work and most of the time is 'close enough'. I haven't found Photoscan to be particularly sensitive to minor variations anyway

If you really want to get fancy, you could put all of the images as a sequence into a compositing tool, like Nuke, sample/average the values and automatically apply the necessary corrections to the images...but it also might get thrown off if the subject matter in the images varies too greatly

I didn't see anything on the RAWTherapee web pages that would do any sort of automatic exposure balancing across a series of image.

General / Re: New Drone from DJI- Inspire
« on: July 09, 2015, 01:55:55 AM »
Just a quick note that the Ricoh GR still exists and has been (very minorly) updated to a Mk2 version (and yes, it's still relatively expensive).

General / Re: Compact Canon CDHK RAW & Lens Distortion
« on: July 03, 2015, 11:31:54 PM »
That makes sense, thanks. I know that outside perspective correction is a usually a strict no-no for Photoscan. It sounds like it will figure it all out on its own either way (camera corrected JPGs or uncorrected RAWs). I certainly haven't had poor results from the RAWs, but it would be worth a comparison at some point I suppose (if only time were more plentiful...sigh)

General / Re: Compact Canon CDHK RAW & Lens Distortion
« on: June 30, 2015, 03:15:11 AM »
Hi Borgy, I can't say exactly why you are seeing these differences in the PS computations. When I did my initial RAW vs JPG comparisons, it was only on individual photos and not through the entire PS pipeline. I haven't done much testing with this setup since them, sorry. I'm not if I ever really compared the results as you have here. It's certainly possible that the camera JPGs are a better fit for PS processing.

When you are converting your images from RAW, there are some additional adjustments might be helpful, such as noise reduction, which is likely happening in the camera's JPGs. I would probably suggest starting from a comparison of the photos themselves and verifying that a converted RAW image actually does look better than the camera's JPG. Hopefully that would improve your PS results. There are also a number of different RAW conversion tools, some better than others. I am usually doing this with Adobe's Camera Raw (or Lightroom)

Cheers,  Chad

Face and Body Scanning / Re: masking an object with background color
« on: June 26, 2014, 07:39:45 PM »
Welcome to the wonderful world of extractions...if your foreground is a similar color as the background, you are almost always going to have a difficult time separating the two. You will likely need to hand trace (rotoscope) the edges that the software cannot detect. You can try using more sophisticated extraction tools/plugins in something like Nuke, After Effects or Photoshop...but similar colors will almost always be a challenge.

Sorry, and good luck! 


"On some Nikon cameras shooting in Live View doesn't automatically mean it uses mirror lockup."

oh, right...the Canon's can do that as well. I think it depends on what sort of live view focusing option you choose. on the Canon's they call it Quick Focus I think, so it flips the mirror back down to focus via phase detection, then flips the mirror back up so you can use live view. although in this case I would highly recommend using manual focus, so I believe you shouldn't get any mirror slap. some live view modes have an extra 'quiet shutter' mode, which also might be useful

You can get by without a remote, just set the self timer to 2 can also use mirror lock up, which can help eliminate any other vibrations. although if you use live view, you don't need to worry about that aspect

iso800 is probably not going to help you, though...just use a good tripod, make sure everything is stable after you tweak focus and fire off the self timer.

I've had good results with Helicon Focus...I hope you will as well.

Cheers and good luck!

Morgan, are you sure those two images match their captions? I would bet that the photo on the bottom is from the DSLR, as it has a shallower depth of field (focus is not sharp front to back) due to the much larger sensor of the camera as compared to the small point and shoot. Any part of the image that is out of focus or otherwise unsharp will cause problems for Photoscan to solve correctly. (no points to lock onto when it's all blurry)

Try using a smaller aperture (f16 or f22) and use a tripod (due to the resulting slow shutter speed). You hopefully should see better sharpness across your object in the resulting photos.

Cheers, -Chad

Meshmixer is also a great tool for fixing problems with less than perfect geometry...more interactive than Meshlab, I have found.

Pages: [1] 2 3 ... 14