Analyzing scene SBR with an app

[Paul Ozzello]

I'm working on that missing wheel, I try to uploaded new build currently.

The green dot is probably the iPhones "camera on" light on status bar.

OK great - it's working now!

 

[radiant]

OK great - it's working now!

Good! There is something funny going on with the grid size configuration. But this is betatesting at it's best.

 

[markjwyatt]

This may not be important (or may be obvious), but I imagine with LF you could put the phone under the dark cloth and image the ground glass directly?

 

[koraks]

This may not be important (or may be obvious), but I imagine with LF you could put the phone under the dark cloth and image the ground glass directly?

Results will be extremely disappointing due to lack of contrast in the GG image and extreme falloff in the corners. Also the phone camera may struggle a bit with the low light levels, although with today's phones, this may not be a severe issue. The problem with falloff/vignetting and inherently lower contrast will spoil the broth.

 

[radiant]

The problem with falloff/vignetting and inherently lower contrast will spoil the broth.

Exactly. I did an camera app that flips the image from ground glass to help viewing the composition but it is really difficult because of this. You basically see a small spot of the scene with camera..

 

[radiant]

Doest any betatesters had time to try this in real use? Do you think it is worth developing further / releasing to app store?

 

[distributed]

Sorry for the late answer! I added a comment in Testflight about a week, but I am not sure it made it to you? Currently I can't see the analysis output on my iPhone SE.

I can see myself using the app for some time to get a feel for scene SBR. I would expect to use it less after I have achieved a certain proficiency in guessing SBR.

A second point is precision: If the app routinely provides inaccurate results, I will probably not use it much. Of course this depends in the largest part on how good your raw data is. I am sometimes using Luxi, without the attached accessory dome, as a lightmeter. I find myself often correcting its exposure advice in e.g. direct sunlight. I suspect this is because the exposure metainformation of the camera images is not accurate, but admittedly I have no proof. Under these circumstances, I only use the app when I'm the shade.

 

[radiant]

Sorry for the late answer! I added a comment in Testflight about a week, but I am not sure it made it to you? Currently I can't see the analysis output on my iPhone SE.

Sorry, I must have missed that. Apple haven't sent any notifications at all so I would probably need to login time to time to see if there are comments.

I've actually put halt on the testing, so no new updates coming any time soon. Busy with other things..

edit: and you are right about that the app should be reliable.

 

[distributed]

No worries! In case the report doesn't arrive: There's basically nothing shown on my screen blow "Analysis", only a red "corr." towards the lower right of my screen.

 

[radiant]

Resurrecting this project ..

I was doing a band shoot with direct positive paper and it was a bit tedious to measure the dynamic range of the scene I was shooting. So I decided to continue working on this app.

Basics are the same but now it shows EV(100) or difference (to locked yellow block) in stops. So it is easy to see what is the difference in stops to certain area on the photo. The one example photo here has black block as reference (captured previously with different color as indicator)

I've also added light meter, I tested this against Sekonic spot meter and it seems to measure correctly!

Some screenshots so you get the idea:

The blue blocks are indicating "underexpose" (configurable over/underexposure indicator).

When block is selected and difference is showing, the light measurement is done from that block. So basically kinda spot meter functionality. Except on this app you can define the area size where the measurement is done.

 

[Andrew O'Neill]

This is really cool,

 

[radiant]

This is really cool,

Yeah I think too. I haven't seen this kind of approach yet + it seems to work. Maybe this could a useful tool for others too..

 

[radiant]

btw if you want to try this out, you can betatest it with this link: https://testflight.apple.com/join/MfSnZoX7

 

[Andrew O'Neill]

btw if you want to try this out, you can betatest it with this link: https://testflight.apple.com/join/MfSnZoX7

I'd test it out in a heartbeat, but I only have an android phone...

 

[radiant]

I'd test it out in a heartbeat, but I only have an android phone...

Good reason to bite an apple! No really, I don't have any Android devices and never learnt to develop those. Mostly writing apps for myself but end up publishing to the AppStore..

 

[radiant]

The app is now in store: https://apps.apple.com/us/app/zoneview/id1555845301

You can use these redeem codes if you want the app for free:

6TNJNWJHTXY4
E3PXPHKTHEF6
KKX7TJJ6E64A
XMX3MF7FNNJR
XTK4PY7LMN4R
HKNPXXH9L76L
4XNYTP3YXEFJ
FJEJ74LTLL6Y
MLXNKTE697LF
LKJR9W7Y669X
JLJY634THA6Y
LJ4HYLKL9YW9
PAPT7KEMMH3N

These are single use, so if you the code doesn't work, try another one. If you don't know how redeem codes work, here is instructions: https://support.apple.com/en-gb/HT201209

 

[radiant]

The app got an update. Now it shows Zone numbers (Roman numericals) if you choose it to. Actually that is pretty cool feature, I have to say myself.

Anyone here using / tested the app?

 

[distributed]

I just redeemed a code. Thanks, @vedostuu! I will give it a spin once I get outside today.

 

[Nikola Dulgiarov]

I couldn't redeem a code (an issue with my App Store I can't seem to resolve), but I did get the app and played around the other day.

Testing on an iPhone SE II, the reading agree with a trusted light meter both in absolute accuracy and in dynamic range. I did notice situations where the readings would drift after changing the scene, as if there is some running average being calculated. I'll spend more time using it to see. Otherwise, I really like it, it's intuitive and I'm far more willing to trust my phone as a light meter using it.

 

[radiant]

Testing on an iPhone SE II, the reading agree with a trusted light meter both in absolute accuracy and in dynamic range. I did notice situations where the readings would drift after changing the scene, as if there is some running average being calculated. I'll spend more time using it to see. Otherwise, I really like it, it's intuitive and I'm far more willing to trust my phone as a light meter using it.

Wow! So nice to hear, really!

There is no running average at all, I assume it is maybe the phone camera algorithm that adjusts the exposure, perhaps? Also what fools me time after time is that the "box" value is really calculated from the whole box area, not just center of it. So if you have even tiny bit of drastic change in luminosity, it will affect to the reading. The boxes can be set so small that even slightest change of the measured area can change the output value. I don't know if this is the case, but I have to check this myself too.

I'm really glad that it is working for you!

 

[Paul Ozzello]

Wow! So nice to hear, really!

There is no running average at all, I assume it is maybe the phone camera algorithm that adjusts the exposure, perhaps? Also what fools me time after time is that the "box" value is really calculated from the whole box area, not just center of it. So if you have even tiny bit of drastic change in luminosity, it will affect to the reading. The boxes can be set so small that even slightest change of the measured area can change the output value. I don't know if this is the case, but I have to check this myself too.

I'm really glad that it is working for you!

I'm using it too - I'll leave some feedback once I know what I'm doing ;-)

 

[radiant]

Update to the app; now it goes to ISO 0.005 for daguerreotype use.

 

[wiltw]

OR just take a spot meter and point it at the brightest location with detail and then the darkest location with detail and compare the two. No app needed.

^
This is what I have always done in lighting and shooting product photograph for reproduction via offset press, which supports narrower dynamic range than can be captured on film.

• I can see merit of quickly assessing a single photograph of complex scene which is captured with smartphone camera, but the issue is that FOV will be different from the FOV of the actual camera+lens.
• Another issue I can raise, using lower right photo posted in Post 61...the upper left corner of that bottom right rectangular area (which was assessed at -2EV), is that the trees there are considerably darker than the rest of that rectangle. So while that rectangle is assessed at -2EV, the trees themselves are darker than that value!
  So a spotmeter would better assess that area than the average value of the rectangular area.

 

[toyoboyo]

Thanks for making this app vedostuu. I just bought it to play around with. It looks like a nice way to visualize exposure.
Only request -
Is there a way to control the camera zoom or increase cell granularity?

 

[radiant]

Another issue I can raise, using lower right photo posted in Post 61...the upper left corner of that bottom right rectangular area (which was assessed at -2EV), is that the trees there are considerably darker than the rest of that rectangle. So while that rectangle is assessed at -2EV, the trees themselves are darker than that value!

I think it is affected by the darker corner area + tree branch.

As the app is averaging the area you probably have to more the device a bit to get confirmation on the reading on this kind of edge cases.

Is there a way to control the camera zoom or increase cell granularity?

Zoom would be nice feature! I have to consider that. I wouldn't increase cell granularity more than what it is.