Auto-ramping not working [solved with Android App 3.0.3]

We tested as good as we can. But there will always be things we didn't find. If you found things that apparently don't work as they should, you can report them here so we can take care of them as soon as possible.
Oliver
Posts: 1127
Joined: October 9th, 2018, 4:17 pm

September 12th, 2023, 6:29 pm

Hi again.

we published an update on the PlayStore, that should fix the issue!

Sorry we missed that.
Founder & CEO of Foolography, Hardware & Firmware developer.
henrikmatzen
Posts: 8
Joined: August 30th, 2023, 7:31 pm

September 14th, 2023, 8:43 pm

Oliver wrote:
September 9th, 2023, 9:28 pm
Hi Henrik,

sorry for the trouble.

We just found the issue: the Android app sent the wrong command to the Unleashed, which started a normal timelapse rather than an Holy Grail timelapse. We'll issue a new App release at the beginning of the week.

Hi Oliver,

I have tested the new update tonight - and it works now with holy grail! Sadly I am back from my big trip, and had to ramp manually and missed some amazing planned spots where I would not be able to physically be there all the time to manually ramp.

However, it does no longer support my Syrp Genie Mini - it won't find them as accessories. It used to work, but not anymore, so I am afraid something else is still wrong for this part - but glad the holy grail part seems fixed finally, thanks for that.
henrikmatzen
Posts: 8
Joined: August 30th, 2023, 7:31 pm

October 10th, 2023, 10:05 pm

Hi Oliver,

I had a sunrise shoot a few days ago - it didn't ramp. The update works with sunsets, but somehow it failed with sunrise. I am not sure if it was just me having bad luck, so I am going to try it again, but I had to manually ramp it - also it seems like it didn't pick the settings from my camera as the starting point, it was using the sunset settings I had been using last time (aparture, shutter, ISO)
Oliver
Posts: 1127
Joined: October 9th, 2018, 4:17 pm

October 23rd, 2023, 3:09 pm

Can you tell me which Algorithm you used for ramping?
The Unleashed will always remember your last choice, so if you did an LRT Sunset before, and didn't change to LRT Sunrise, then that would explain your issue. The EV24h Algorithm works for both sunrise and sunset.

Also, be sure to check your Ramping order, and the limits you set. It can happen that the Ramping order was changed, and if you then start a holy grail with Ramping order limited to one setting, and might even be close to the setting's day limits, then the Unleashed won't continue ramping, by design.
FYI: the Ramping order setting is shared by Holy Grail and Exposure bracketing, just something to keep in mind!
Founder & CEO of Foolography, Hardware & Firmware developer.
bernhardjost
Posts: 3
Joined: August 15th, 2023, 3:26 pm

October 24th, 2023, 10:45 am

Update 23.10.2023
Tested most recent App version 3.03 & Firmware with Nikon D850

Auto Ramping seems to work again! Tested a few frames with EV24h, LRT modified Sunrise Beta, LRT modified Sunset Beta basically by adding and removing the lens cap.

However I still get the warnings "Ramp Mode is read-only...." and "Ramp Order is read only..." as soon I start the timelapse. Not sure if these warnings actually matter since the ramp order is done correctly
Oliver
Posts: 1127
Joined: October 9th, 2018, 4:17 pm

October 24th, 2023, 11:41 am

Yeah, don't worry about those errors. They're supposed to be there, but shouldn't be displayed in the app. They just say that you can't change those settings while the timelapse is running. We'll make sure they won't be displayed in the next version of the app, which we'll publish this week.
Founder & CEO of Foolography, Hardware & Firmware developer.
henrikmatzen
Posts: 8
Joined: August 30th, 2023, 7:31 pm

December 20th, 2023, 8:49 pm

Oliver wrote:
October 23rd, 2023, 3:09 pm
Can you tell me which Algorithm you used for ramping?
The Unleashed will always remember your last choice, so if you did an LRT Sunset before, and didn't change to LRT Sunrise, then that would explain your issue. The EV24h Algorithm works for both sunrise and sunset.

Also, be sure to check your Ramping order, and the limits you set. It can happen that the Ramping order was changed, and if you then start a holy grail with Ramping order limited to one setting, and might even be close to the setting's day limits, then the Unleashed won't continue ramping, by design.
FYI: the Ramping order setting is shared by Holy Grail and Exposure bracketing, just something to keep in mind!
Sorry for the late reply - was travelling again. Had to shift to use QDSLR, because I had too many failed holy grails with my Foolography. To answer the question, I used the LRT Sunrise algorithm, so that was correctly set. I have not tried the EV24h, but I will give that a try to see if that works better for me. Ramping order and limits was correctly set, however, I fail to be able to manually auto-adjust if I can see things are starting look off. With QDSLR I can override the auto ramping and go full manual (and go back to auto if I want) if I start to see that the auto ramping is going bad, and can manually take over and remote-adjust via the tablet. Would be really nice if I could do that in Unleashed too, but it seems to not really work (managed to make it adjust once in a while, but it is not stable and working every time) :-)

Sometimes when shooting lapses with a bunch of ND filters on top, in mixed light conditions, it can be a bit hard for the auto ramping to work well, that goes for QDSLR too , so the ability to take over manually would be nice.
Oliver
Posts: 1127
Joined: October 9th, 2018, 4:17 pm

December 23rd, 2023, 1:55 am

The issues you mentioned were fixed in the Android app version 3.1.0 a while back. This also fixed the but that the "Adjustment" value in the Holy grail screen was read-only during a timelapse. This Adjustment is basically a method for you to intervene with the Algorithm - if you find the images are starting to be too bright, set it to -0.3 EV (or -10% for LRT). If you try the LRT modified Algorithms, we added an undocumented feature - mainly for testing purposes - where instead of adjusting the reference value by X %, you can make an instant adjustment by EV values in either direction, and the Algorithm would then use the next image as a new reference photo.

The problem with the adjustment of the regular LRT reference value is that because the algorithm is unidirectional, while you can see a fairly immediate change in one direction, in the other direction you have to wait for the scene to get so much darker that you'll actually notice that the algorithm skipped making brightening adjustments so that your image is now also darker... This makes it feel as if the adjustment is not working, and makes it impossible to judge whether you adjusted by the right amount.

(the only other difference of the LRT modified algorithms is that we don't normalise the histogram before calculating the weighted average according to Gunther's formula. While unlikely in a real-world scenario, the normalised histogram can actually result in an image being determined to be darker than a previous one, if there is one huge peak in the histogram (for example car headlights briefly shining into the camera creating a spike in the completely white column).

The problem with manual adjustments of the actual settings during a timelapse is twofold:
1) there is usually too little time between shots for you, or the Unleashed to be able to make an adjustment (except the automatic ones, since the Unleashed explicitly waits for those to complete before taking the next shot)
2) as soon as you've made that adjustment, the Algorithm will then slowly make adjustments in the opposite direction again. Since it always keeps using the reference brightness value of the first image for LRT algorithms or tries to stay between +0.3 and -0.3 on the EV meter in the EV24h Algorithm. -> that's the reason why we have a separate "Adjustment" value for the Algorithms.

I hope this info and the app and firmware updates help!
Happy Holidays!
Founder & CEO of Foolography, Hardware & Firmware developer.
Post Reply