Thanks for the report back. @jcasman and I will test this again with the THETA Z1.
This happens if the āNumber of pictures with noiseā settings is more than 2.
And how to make the application take pictures of raw and not jpg?
In the current version 1.0.1, the denoise was only tested at 1. There is no RAW support right now. At the time the plug-in was submitted, Kasper hadnāt received his Z1 yet and was unable to test it properly.
The community did basic testing. There are definitely issues with the Z1 due to the larger image size, but I think itās a useful introduction for people with the Z1.
The Z1 support should improve once Kasper has a Z1 in hand. Heās ordered one.
Only two out of ten panoramas are gathering ā¦
In the camera settings you can not change the resolution of the photo?
It was tested with 9 pictures (9 sets of 3 pictures, so 27 pictures) at denoise 1. However, the testing was very limited.
The camera can change the resolution, but the app was only tested with JPG at the highest resolution, not RAW.
Thereās likely a problem with some resource such as memory usage in the camera that is causing problems with the larger Z1 images. This can likely be fixed with more hands-on use.
Thanks for all the time you spent testing the app and the reports. Iām sure the developer, Kasper, will be able to use this feedback.
Please make it possible to shoot in raw without stitching.
What do I need to do so that the Authydra plugin does not hang? Change the resolution of the photo in the application āRicohā? But the āimage sizeā settings are closed there. They cannot be changed.
Today I talked with a man who also has z1. His āAuthydraā works without freezes.
This problem can only be with my camera?
I have a similar, or the same, problem on my Z1. After it takes the photos it loses wifi and I get the browser message ācanāt connect to the serverā. I have to reboot the Z1 to reestablish wifi.
One thing that has seemed to helped a bit was turning off sleep, or changing it from the default 3 to 10 minutes. It still loses wifi while processing after taking the photos, but once processing is done, it comes back instead of hanging.
Also, the wifi logo doesnāt change color, etc. for me. Does it not work on the Z1?
Iām using the current version of Authydra on the Theta plugin page.
Any ideas?
Iāve got a bit more information about Authydra running on my Z1. It takes 2 minutes 30 seconds to take the photos with 3 images per exposure selected. It then starts the merge which is when I lose connection. At the 9 minutes and 30 seconds the screen on the Z1 goes out and at 10 minutes the camera goes to sleep. FYI, I have sleep turned off.
Maybe itās by design, but I canāt connect to the camera again until I completely power off and on.
Iām using firmware Version 1.11.1
I reinstalled the plugin, and the same thing happened. The camera seems to shut off before the exr is made. Also, all of the images used to create the exr are still laying around without the .jpg extension, so it doesnāt seem to have cleaned up, which makes sense. Iām not sure how I previously got one exr made, out of many tries.
Do you think this a firmware Version 1.11.1 issue or something else?
Hi Bret,
strange issues. I still havenāt received my own Theta z1 so it is still a bit hard to help.
What happens when you set the denoise number to 1? and the number of exposure pictures to something low like 5.
Does it complete then?
Maybe it is something with the amount of memory the z1 has and that the images have a highr resoluation which takes more memory.
I hope to recieve my z1 soon so I can make a stable version and do some good testing.
Thanks for the response, Kasper. I suspect youāre correct about it being a memory issue. I reduce the exposures to 5 with 2 images per for noise reduction and it completed in 3m:01s. Increased to 3 images per for noise reduction and it crashed. Note that I still have to reboot the camera to be able to connect with it.
Thanks again for your hard work. Iām sure youāll be able to add lots of cool things unique to the Z1 once you receive yours.
One more thing. All files created by the plugin are time stamped āDec 31, 1969 at 7:00 PMā.
Hmm that is also interesting Iāll look into it!
Thanx!
Aan wo, sep. 25, 2019 om 23:23, Bret via THETA 360 Developer <> schreef:
I wanted to add that Iām having the same issue as described above but on the Theta V. Of the 7 HDRIās Iāve attempted to capture (all with default settings), only 1 of them has produced an EXR while the others just leave directories full of jpgs.
The 1 that actually worked was initialized through the web interface, the others were initiated via button press on the unit when out in the field :(. Autosleep is also disabled on my device.
You mention on the todo that youād like to export the default python script for offline processing. Where is this? Is it embedded as a string somewhere in the JNI? Or did you mean to port some openCV C++ into equivalent python openCV? I might take a stab at it and send over a PR if you can point me to the entry point.
Hey all - any updates on DNG support for the Z1?
Also - I cannot seem to access the web interface on the Z1. The address 192.168.1.1:8888 doesnāt do anything, and āStart Pluginā in the Theta app (which takes you to 192.168.1.1/plugin) just times out.
Hi Zach,
hmmm strange behaviour. Are you on the latest firmware? There seems to be something in the older firmware that didnāt work so well with the opencv library.
When you press the button it takes pictures with the default setting, which is 11 pictures with 3 denoise. When you use the webinterface and set it to these default setting and then take pictures is it able to complete?
yes the python script was one of my plans but it is still on the to-do list so I havenāt gotten around to doing it. So it is not there yet. It is a bit on the backlog because if the app works as it should then the script shouldnāt be necassary. (btw it would pretty much be the steps outlined on this website https://www.learnopencv.com/high-dynamic-range-hdr-imaging-using-opencv-cpp-python/ )
let me know!
Yes.
That was my experience but I donāt think it is causation. I went through the steps outlined and attempted to manually process these and discovered that alignment was failing. It seems root cause in all 6 cases for me was that I had at least 2 exposures that were pure white, literally 100% of pixels clipped. It seems like calculating a better middle exposure is the key here.
in that cause you could try to set the stop jump / settings as low as possible. that will hopefully prevent to many full white pictures.
The alignment step can be skipped as the pictures should all align (assuming you didnāt move the cam that is).
Also creating the response curve can be replaced with the curve you can download from the text file camercurve.txt but you will need to write something to read that file into a variable. (I have attached a file which I use to test stuff in python so with a bit of work you can find the code to read the curve file. It contains lots of othe junk aswell).
good luck!
(Attachment hdr2.py is missing)