Flicker Free ETTR Timelapse: - -Beginners Guide & Basic Post Processing --

Started by RenatoPhoto, May 26, 2013, 01:35:58 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

dmilligan

My script uses a preview of the actual output from ACR to do the deflicker. From that perspective it may actually be more accurate, not less. The curve for ACR's exposure slider is not perfectly linear and somewhat 'image adaptive' i.e. putting in 1 EV in there doesn't mean ACR is always going to be exactly 1 EV in the output (or not all areas of the image are going to be adjust by exactly 1 EV). So even if you measured 1 EV perfectly in the RAW file, ACR may not actually correct by exactly 1 EV. Additionally with my script you have the flexibility to deflicker when using strong amounts of other ACR settings that tend to cause additional flicker b/c they are 'image adaptive' (such as highlight and shadow recovery and clarity).

If you think of ACR as a black box, the in camera deflicker only has the capability to make a guess as to what it's going to do. My script can actually examine the output and make sure the 'black box' did what it expected, and then make additional corrections. This is the whole reason the script does multiple iterations.

If you use the ufraw workflow, this isn't the case (ufraw is not a 'black box') and the in camera deflicker should be approximately the same.

TBH, I don't think there's going to be really any noticeable difference between the two methods. From my experience, deflickering in post is far more convenient, and you can have a lot smaller interval between pictures (in camera deflicker requires about 5-10s to run), which is important when you get to night time and you need on the order of 30s exposures.

Please feel free to do a comparison though, I could be wrong.

NickZee

@arturochu  I think 5 seconds is a good place to start experimenting.  It really depends on how fast the cars are traveling.   Go out the night before and shoot a few test stills to see the trails you get.

@dmilligan.  Thank you for the thoughts.  I've been using the Bridge script and everything seems wonderful.
Architectural Photography
NickZimmerman.com
5D MK3 & 600D | 24-70mm USMII L 2.8 | Nikon 14-24mm 2.8 | Nikon Nikkor 24mm 2.8 AIS & 50mm 1.4 AIS ** Windows 8.1 Pro | 32GB Ram | i7-4770 @ 3.4GHz

Signum

Hi all, im totally newbie in timelapses.
Mi 6d crash all the time in the ettr test pic, this is the log:

ASSERT: FALSE
at ./ASIF/ASIF.c:486, task ASIF
lv:0 mode:3


Magic Lantern version : v31337.TRAGIC.2014Feb28.6D113
Mercurial changeset   : fd027ab570aa (unified) tip
Built on 2014-02-28 19:16:45 UTC by User@PC.
Free Memory  : 354K + 1855K

Can anyone help me plz??
THX

Signum


Stongtea

Hi all,

Spent some time getting this to work over the last few days, I am deeply grateful you guys have made this is possible on a camera as cheap as a 600D.

However, when i have been shooting, the size of the "stop" changes are really obvious in my timelapse videos (i.e. the jump in exposure is too big) is there a way of bringing the jump size down?

Kind regards

Rob

andyshon

@dmilligan
I've been trying your script and having a few problems. My first test worked pretty well but I thought not quite as well as the original in-camera deflicker. But I'd processed the shot with default ACR setting, where as the in-camera deflickered version I was comparing it to had been tweaked. So I tried to synchronise the ACR settings and re-run the script, at which point things started to go wrong and I've not had any joy since. I've been trying to do too much at once so I really don't know if the script or my system is the problem. I've never used Bridge before, which doesn't help. I've been trying to get my head around it but I'm really not in a position to make a coherent fault report at the moment, except that it often seems to freeze and when I try to quit Bridge I get an error message saying 'resize is undefined'. Sorry for the non report report. I'll persevere if I get the chance and post more coherently if I can.

I've been trying the latest builds for ETTR timelapse on a 5D2. I've not been getting the crashes I was before and ETTR seems to be functioning really well, but I'm still having problems with deflicker. In the shot below about 20 frames, distributed seemingly randomly through the sequence, did not have xmp files associated with them. And there was three sections where the brightness jumped 0.38 stops for about 150 frames, before jumping down again. Looking at the back of the camera during the shot, global draw did not seem to be working properly. It had been at the beginning but by half way through the image and intervalometer info was displaying correctly but no histogram or zebras, instead some text that looked like code. I've forgotten what it said I'm afraid, should of written it down. Despite this ETTR clearly functioned fine.


dmilligan

Quote from: andyshon on April 10, 2014, 06:14:47 PM
@dmilligan
I'm really not in a position to make a coherent fault report at the moment, except that it often seems to freeze and when I try to quit Bridge I get an error message saying 'resize is undefined'.
I've seen some intermittent issues with 'resize is undefined'. It typically has something to do with Bridge not giving me valid image preview data. I request the preview from Bridge as 'BitmapData' and then I call a function on it called 'resize' which is supposed to be a function of that class. If Bridge gives me bad data for some or another reason then that function isn't there and you get the 'resize is undefined' message. I think it has something to do with the cache and synchronization stuff. Supposedly I'm in 'synchronized' mode, so that when I request stuff from Bridge, it blocks until the data is valid, but it seems like sometimes that doesn't happen. I don't think there is anything I can really do about as its probably a bug in the Bridge scripting engine.

TLDR: Best bet is to select all the photos and flush the cache and then wait for the previews to regenerate, then run the script.

QuoteIn the shot below about 20 frames, distributed seemingly randomly through the sequence, did not have xmp files associated with them.
The deflicker might not have had time to finish running. What was your interval time between shots? Getting below about 10s and you can start running into the occasional 'dropped' xmp file, if for some reason the CPU was a little busier. It also has to contend with ETTR's analysis using the CPU as well.

diogo.fagundes

I have noticed that the ETTR can be linked with an external shutter. This can make the shutter time get over 30'' ? If yes, is there any means of linking the ETTR with the Bulb Timer from ML? Because I would like to have a bigger minimum shutter in the beginning to make its end more smooth (shooting a sunrise).
Another question, the ETTR module can work with Tv mode, what makes the camera choose the aperture based on the expo. Is there any ways of combining the ETTR with the Tv mode to get more Ev's during a night to day timelapse, without getting big flickering and neither confusing the ETTR?
Premiere cs6, LightRoom 4, LR Time-Lapse 3, Mac OSX 10.8.5

dmilligan

When you are in a semiautomatic mode like Tv or Av, ETTR sets the exposure compensation, and acts more like a 'correction' to Canon's metering.

So for example, you're in Tv and Canon meters the exposure and then takes the image. ETTR analyzes this image and determines how much it could be shifted right (or left if it's overexposed) without overexposing and sets that as the EC. The next photo you take will still be metered by Canon, but it will be offset by this EC.

Using Tv or Av with ETTR for night-day timelapses is not particularly useful because you are still relying on the Canon metering system somewhat. It's best to bypass it completely and let ETTR set the exposure itself.

a1ex

Quote from: diogo.fagundes on April 18, 2014, 01:57:33 AM
linking the ETTR with the Bulb Timer from ML?

I miss this one every time I get into a cave (which sadly doesn't happen often :P )

But on my todo list there is an algorithm to meter a long exposure from a really really dark test picture in one iteration. Whether I'll be able to make it work... we'll see.

andyshon

Quote from: dmilligan on April 11, 2014, 11:25:17 PM
If Bridge gives me bad data for some or another reason then that function isn't there and you get the 'resize is undefined' message.

Yeah, this seems to be what's happening. I think it may be caused by something outside Bridge. Or at least it seems to happen less if Bridge is the only application running. I'm also seeing problems like below (sometimes), where certain frames are processed very differently from those around them in one iteration, the following iteration will then try to correct them. This was after 3 iterations of the script. They weren't there after two iterations and the fourth started correcting them. This may have started happening after a 'resize undefined' problem, not sure.



Quote from: dmilligan on April 11, 2014, 11:25:17 PM
Getting below about 10s and you can start running into the occasional 'dropped' xmp file, if for some reason the CPU was a little busier.

I was shooting every 12 secs with 6 sec max exposure, but some of the frames with dropped xmps are short exposures, so did have over 10 secs processing time. I still have an old build (21st July 2013) that has served me well for ETTR. With this I can reliably shoot with the intervalometer set to 8 secs and the max exposure set to 4 secs.

austinmarti

This is a very n00bish question as I have never shot a day-to-night timelapse.

The ML settings override the manual Canon settings, right? So they take care of the changes in exposure going into night?

diogo.fagundes

Is there any ways of using ettr without giving all the preference to the shutter? I mean, because sometimes I want a timelapse at night with an interval time lower than 30 sec, but every time I set the ettr it makes my shutter speed go to 30 sec and my ISO to the minimum as possible. It would be great if we could lock the iso or give preference to the iso till it reach a determined value.
Premiere cs6, LightRoom 4, LR Time-Lapse 3, Mac OSX 10.8.5

dmilligan

Quote from: diogo.fagundes on April 20, 2014, 02:52:02 AM
Is there any ways of using ettr without giving all the preference to the shutter? I mean, because sometimes I want a timelapse at night with an interval time lower than 30 sec, but every time I set the ettr it makes my shutter speed go to 30 sec and my ISO to the minimum as possible. It would be great if we could lock the iso or give preference to the iso till it reach a determined value.
You can set the maximum shutter speed for ettr. There's no reason in my mind to lock the ISO or give "preference" to it.

Quote from: austinmarti on April 20, 2014, 01:50:19 AM
The ML settings override the manual Canon settings, right? So they take care of the changes in exposure going into night?
Yes, ML takes control of the Canon settings.

diogo.fagundes

Quote from: dmilligan on April 20, 2014, 04:33:32 AM
You can set the maximum shutter speed for ettr. There's no reason in my mind to lock the ISO or give "preference" to it.

ohh... sure... I completely forgot that. hahaha
Thanks!
Premiere cs6, LightRoom 4, LR Time-Lapse 3, Mac OSX 10.8.5

matthewgowan

I went out this afternoon to test my capabilities shooting the day-to-night holy grail using ETTR.

https://vimeo.com/92484919

Next time I'll be sure to keep our very happy bichon named Max away from the tripod legs.  :'(

So for shooting I let Magic Lantern ETTR take care of everything. I did not use the Magic Lantern de-flicker. I used LRTimelapse and Lightroom for the post processing. Exposures ranged from 1/60th to 5 seconds at ISO 100 using the "Sunny" white balance (I think it's 5200).

I mainly wanted to see how well my understanding of Magic Lantern would "ramp" with ETTR. There's certainly no complaints from me on this avenue.

Saramaki

Hi,

I recently found this fantastic Magic Lantern script for my timelapse experiments as I shoot mainly long day2night2day timelapses. Without any kind of bulb ramping it's practically impossible. I did some tests with the bulb ramping feature on the stable ML version (v2.3) a few weeks ago, but I wasn't happy with the harsh ramping, which made post processing difficult. Then I discovered this ETTR and thought that this may help. However, I have not had much luck with my tests. Two time already the ETTR has somehow "freezed" during shooting fixing the shutter speed to some low value like 1/25 sec although the night is dark (requiring at least 15 sec exposure). In the morning all I find is some 300 totally black photos and another 300 all-white overexposed photos due to non-working ETTR. I wonder why is this? I have followed the ETTR for some time during the intervalometering in each trial to make sure it works properly, and it does. Then I close the display (turn it facing inwards) to save battery, is this where it goes wrong? If this is the problem, is it possible to use ETTR with closed display?

I have also found that my camera drains the batteries quite fast when using ETTR. With battery grip, I can normally take at least 1500 photos. With ETTR (in room temp) I can reach only 600-1000.

My setup:
EOS 600D + battery grip
Tokina 11-16mm f/2.8
ML Nightly.2014Apr29.600D102
Settings like in this tutorial: http://wiki.magiclantern.fm/ettr
Camera in Manual mode, by default 1/250 sec, f/4, ISO 100
Max shutter speed was set to about 30 sec.
Aurora Borealis photography & day2night2day timelapse photography

a1ex

Quote from: Saramaki on May 05, 2014, 04:45:49 PM
Then I close the display (turn it facing inwards) to save battery, is this where it goes wrong?

Quite possible, since ETTR relies on the QuickReview image.

Saramaki

Okay, I did another test with display on and ETTR worked very well for the whole 11-hour test period. So it seems that I have to keep the display on when using ETTR, which is a slight problem at least in cold outdoor conditions. Any workaround?

Also, I noticed that the low battery performance was due to one faulty 3rd party battery in the grip. It was noticeably swollen, so it's straight to recycling bin. Last night I tested using one original and one 3rd party battery and I got 11,5 hours of operational time in room temps.
Aurora Borealis photography & day2night2day timelapse photography

dubzeebass


blainesuque

Hi Alex
I have been using this method for some time now. I have not had any major problems. but lately when i shoot timelapses using the Intervelometer and Auto Ettr or sometimes just the intervelometer by itself. when im done and in post processing i see that sometimes the image size changes or crops randomly at some point during the timelapse but usually towards the end. My average photos per timelapse is around 300 photos. Below is an example of what im talking about, do you have any idea of what may be causing this?

ps. I also get a messege "RAW ERROR" on the display sometimes and im not sure why or what that is coming from. 






https://m.flickr.com/#/photos/92940314@N03/14056319919/

https://m.flickr.com/#/photos/92940314@N03/14056319919/

garry23

I hope someone can point me in the right direction.

I am using a 5DIII, LV mode, silent picture, timelapse, A-ETTR always on, sidecar files selected, review 2 sec.

I get the dngs OK, but no xmps are created.

Am I doing something obviously wrong?

Cheers

dmilligan

The ML deflicker was not made to work with silent pics. It uses the QR data and there is no QR for silent pics.  It would be theoretically possible to add deflicker for silent pics, but I don't think anyone is going to work on that as the preference is now for doing deflicker in post (this would be time better spent working on a post tool).

If you use Adobe you can try this: http://www.magiclantern.fm/forum/index.php?topic=8850.0

garry23

dmilligan

I guessed that was the issue: pity, as shutterlesss timelapse is a godsend.

The issue with non-sidecar dngs out of silent is that they don't have an EXIF data, ie ISO or exposure time.

Thus, for holy grail A-ETTR silent work, the dngs are nearly useless: I think?

For normal (fixed exposure) silent picture timelaspse there is a simple work flow:
1. Note aperature, ISO, exposure etc;
2. Capture Silent DNGs (no ETTR);
3. Ingest into Lightroom;
4. Use LensTagger to add back the missing (silent picture) EXIF;
5. Use LRTimelapse to create your masterpiece!

Bottom line: looks like I should steer clear of sunrise and sunset for silent picture timelapses!

zeinikuzeiniku

So I have read through this thread and the one for the script about deflickering.

I have also done some research on ETTR but can someone help me understand this a bit more?

ETTR is good for maintaining details that might otherwise get lost if underexposed?

I did a time lapse and followed all instructions using ETTR and the deflickering afterward. However, I didn't see why there was any advantage to using ETTR. I was shooting a time lapse during daytime so maybe ETTR isn't needed in this situation? Is it best for changing light scenes (ie. sunrise/sunset)?

Anyway, everything works well for this workflow. I'm just having a hard time wrapping my mind around the benefits of it all and what exactly ramping exposure and such with the script besides deflickering does.