Messages

I miss that they didn't go with Foveon sensor. 

But they plan to launch a Foveon model next year according to PetaPixel and Fstoppers.

Could be downsampled 4k from 6k.

But 12 bit cdng is external recording only, but not specified if you need an actual recorder or just SSD drive. The concept video shows it rigged up with a SSD, like the BMP4k.

There is so much talk of RED owning patents on wavelet compression e.g. the reason Blackmagic removed cdng from the P4k when they introduced Braw. That is atleast word on the street (the internet). Interesting to see if this camera will run with uncompressed cdng only.

In the camera web page it says "storage media: Portable SSD USB 3+". They don't specify that it should be a tape recorder. --> https://www.sigma-global.com/en/cameras/fp-series/specifications/

Tested in my 5D2 and it works even if the camera reaches shutter or ISO limits.

Advanced Bracket
Type: Exposure
Frames: 5
Increment: 2EV
Sequence: 0 - + -- ++
Delay: Auto
Shifting: Half

Check that type is not set as Autodetect.

+1

In my case, shooting interiors, I also would like to have a way to mark those pictures that are part of a manual bracketing set or reject the test photos.

Maybe to be able to assign color tag, flags and rating to the photos? This way you could flag as rejected the test shots and once you import them you know the rest are good to go.

As far as I know, I don't think it's possible, and if it were I suppose aliasing and moiré would increase.

I would like to help, but since English is not my native language, I don't think I should update the wiki because of any mistakes I might make.

I can, however, translate everything into Spanish. Is I18N planned?

Everything looks fine with the cr2hdr Lr plugin and the new Lr 7.3 version. Nevertheless, Lr crashed trying to convert a batch of 65 Dual ISO files, something that never happened to me with previous Lr versions.

@garry23,

That again gives me equally apart shots, good for exposure fusion but not perfect when working with layers and luminosity masks.

Just another question, do you know if my request could be done using LUA? If so I could take a look to all the documentation and try something myself.

Thanks

If this project ever goes ahead, count on me to translate ML to Spanish and Catalan.

@R

Have you considered using the ML advanced bracketing together with the ML raw spotmeter, to nail the exposure for the shadow area where you require details, ie set the Ev 'zone' you want, ie x Ev down from 100% blown out.

...

@garry23, I already use the spotmeter in EV feedback. Do you mean then placing my shadows in the desired zone and Autodetect frames with a - -- sequence in the Advanced Bracketing?

Running @a1ex bracket.lua on lua_fix.2018Mar21.5D2212 all 4 shots trigger the external 270EX flash at -2 FEC.

[Intro]

Hi developers,

[Jump to Request if you don't want to read the Intro]

Intro:

Although I studied some programming that was almost thirty years ago and all the knowledge is buried deep in my memory, so deeply that now, even if I wanted to, I could not program myself what I am going to ask of you. I can be a good tester though; I use ML every single day in interior photography shooting hundreds of photos every day.

Here's the thing. Dual ISO is the best feature of ML in the field of photography in which I work. Measuring for high lights and setting the power of my external flashes according to the volume of the room and the reflectance index of the walls and ceiling gives me a very good exposure base of the interior and the exterior to get all the DR I need. But there are some situations where flash is not an option. Main one is white balance. If the color temperature and brightness of the outside or the inside light is too extreme and there's no time to color balance the external flashes, using them is going to give weird color cast which will be difficult or time consuming to remove in postpo.

That's the moment when I use bracketing. But as you know, equally EV apart shots in a bracketing is not the perfect way to get the perfect bracketing. For simple scenes you can expose for HL and use + ++ in Advanced Bracketing or expose for midtones or shadows and use - --, depending on what you're trying to achieve. But for complex scenes, let's say bright exteriors with dark furniture on a mahogany wood floor and ochre walls lit with orange incandescent bulbs of different power, you get the perfect bracketing exposing specifically for each luminance zone where you want to preserve information.

Request:

Would it be possible to make the Advanced Bracketing even more advanced  adding a way to store in memory a set of different exposure values previously measured?

In this "Custom Advanced Bracketing (CAB)" you would have let's say 2, 3 or 4 memories: lights and shadows; lights, midtones and shadows or lights midtones 1, midtones 2 and shadows.

This feature could work in two ways. One could be, after you access the CAB you measure highlights and press SET to store that EV in memory, measure midtones and press SET, ... and so on. Another way, even cooler, could be using the AF box to define where those areas of interest are. You could move the AF box to a window and press SET, then move the AF box to the wall and define midtones and move the box to a shadow and store in memory where the shadows are. In the CAB settings you could then tell ML where you want those areas to fall in the zone system, for example memory 1 (highlights) Zone IX, memory 2 (midtones 1) Zone VII, memory 3 (midtones 2) Zone XI and memory 4 (shadows) Zone IV. This way, knowing ML where to measure, you could get the perfect bracketing several times just pressing the shutter button, without having to manually measure again when, for example a cloud changes the brightness of the exterior or you decide to turn on or off some lights.

About https://www.magiclantern.fm/forum/index.php?topic=7139.msg196445#msg196445, it turned out to be incompatibilities between Lr plugin and exiftool .

cr2hdr.3.0-170316-02-MLV-DEV Plugin uses exiftool 10.43 and cr2hdr-win 2017-12-04 00:01 comes with exiftool 10.02. If you want to update the plugin bin folder with new releases of cr2hdr-win, just replace cr2hdr.exe and dcraw.exe but not exiftool.exe.

@R: Sharing an unprocessed Dual-ISO file might be helpful ... content of plug-in bin dir, too. Exiftool, dcraw, cr2hdr

Like the one here? --> https://www.magiclantern.fm/forum/index.php?topic=7139.msg196445#msg196445

( https://www.amazon.es/clouddrive/share/2hOTnDMQlaR59dlhFOp5faiVnOftB4sWivsdVIiHL5K )

The ..\Adobe Lightroom Classic CC\cr2hdr.lrplugin\bin looks like this:

I used the latest version for mac and it transcodes just fine over here, camera name and lens name is represented. Are you using cr2hdr directly or some other app?

I'm using Dual ISO Converter plugin for Lightroom. Could it have something to do the fact that I've downloaded last build of cr2hdr and overwrite the cr2hdr found in the Lr plugin?

- unprocessed dualiso: https://www.amazon.es/clouddrive/share/2hOTnDMQlaR59dlhFOp5faiVnOftB4sWivsdVIiHL5K

- processed dualiso: https://www.amazon.es/clouddrive/share/N5smrclgCuyIQOPP5bO80qnh0TMaryMC825RSzsGakM

Things I have noticed since build 2017-12-04 00:01.

1) Metadata loss. Unprocessed dng has the right metadata.
- Exposure: 1/125 at f8.0
- Focal Length: 30mm
- Model: Canon EOS 5D Mark
- Lens: EF 17-40 f4 L USM

Processed dual iso has lost metadata.
- Exposure: 1 sec at f1.0
- Focal Length: ?
- Model: Canikon
- Lens: ?

So Lr warns that it's unable to locate a matching lens profile automatically.

2) Even after manually assign lens profile something seems wrong. Before the last build 90% I needed -2 points of distortion correction for the 17-40mm. Now I need about +10. ?¿?¿?¿ Weird.

3) Orientation info seems lost? Before last build Lr automatically rotate the images to right orientation, now I have to do it manually.

Lr 7.1 on Windows 8.1.

Has anyone experienced problems with last version (2017-12-04 00:01) of cr2hdr processing 5DII files?

I notice two things in Lightroom that I think they happen since I updated cr2hdr.

First, it seems the processed Dual ISO DNG it has lost orientation info as now LR doesn't turn my images automatically.

Second, LR is unable to find a lens profile of the DNG processed with cr2hdr but it can find it automatically in the original Dual ISO file.

Anyone has noticed this also?

The last screen looks a bit odd - how does it look on a real camera?

Here's mine.

You get that even with a 4" ISO 100/800 exposure.

http://www.magiclantern.fm/forum/index.php?topic=7402.msg188470#msg188470

http://www.magiclantern.fm/forum/index.php?topic=7402.msg188605#msg188605

Spanish test passed. Google Translator is good but also "funny", DeepL seems almost human to me.

Awesome.

In your user case, wouldn't be better to work with HDR, instead of dual iso?
I've already tried both (using HDRMerge for processing) and HDR gives me better results (of course).
Dual Iso is very useful for high speed photography and MLV video, but if you have a tripod and time, HDR gives much better results.

I do not only use Dual ISO because of its DR expansion but also because it helps me optimizing my workflow in most of my sessions. HDR means usually 3 bracketed shots + exporting those images to an HDR software + importing the tone mapped image to LR again... and that means extra time added to the session and therefore high prices for my clients. Of course that in high level shootings I bracket the scene and work with layers in PS but for mid-level ones DISO is the best option in my opinion to 1) get good results, 2) good prices and 3) be able to shoot many properties during high season. Anyway, as I wrote, I also took some bracketed shots just in case and these photos were a kind of experiment to see what the results would be using DISO in a very dark bedroom.

When trying to compensate interior with exterior exposure I mostly use flash + Dual ISO. You can check the results here http://www.ramonclemente.com/portfolio/

I usually try to bounce flashes on walls, ceilings or any other neutral surface I can find in the room. This particular room had two walls of red bricks, red bricks also in the ceiling, light brown tiles on the floor and very dark mahogany furniture (behind me) so bouncing the flashes would have cast red tints to the scene. The room had three 20W incandescent light bulbs.

This time I had no umbrellas or diffusers with me and the only way of getting a kind of natural look was using the few light coming from a dimly lit courtyard so I took a couple of Dual ISO long exposures. I also did bracketing as a back-up because I have to admit that I had not done long exposure using Dual ISO before. When photographing interiors my shutter times are usually from 1/4 to 1/160, this time I took one at 4" and another at 8".

I like the result. The only issue I find is that dead, hot and stuck pixels are more noticeable; they even seem bigger.

Another example of stripes appearing after cr2hdr. In this case stripes appear in a column about 50px wide in the left third of the image.





Here's the link to the Dual ISO RAW, unfortunately I don't keep the original RAW, just the processed after cr2hdr:
https://www.amazon.es/clouddrive/share/3DjuCaSJLRcLTcP9oKvgjrdH4Hx7PLr9Vwub8q90rTv

--no-stripe-fix softens the problem but creates new stripes through all the image clearly visible in dark areas.





Here's the original Dual ISO RAW download link:
https://www.amazon.es/clouddrive/share/bg65hvtYaeodNzBKjIODaOb9KuBam1sSEYdZg53b7sC

Editing this image I noticed a kind of ghost pattern at the bottom that seems to come from the wood table. Is this known and normal?

5D Mark II 2.1.2, Lr 6.7, Dual ISO plugin 3.0-170316-02






image hosting