[Papywizard v2] Tethered shooting and intelligent bracketing  

Everything you need to motorize your head
User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

[Papywizard v2] Tethered shooting and intelligent bracketing

by fma38 » Fri Feb 06, 2009 12:05 am

I finally asked on the gphoto2 ML, and, with the help of the developpers, found a way to change the exposure bias on my Canon EOS 20D. So, as soon as this feature will be implemented in libgphoto2, I will start to work on the intelligent bracketing feature, by analysing the pictures...

And it seems that other guys are also working on this, so it should progress quickly.

Let's discuss this here....
Frédéric

User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Fri Feb 06, 2009 2:21 pm

fma38 wrote:I finally asked on the gphoto2 ML, and, with the help of the developpers, found a way to change the exposure bias on my Canon EOS 20D. So, as soon as this feature will be implemented in libgphoto2, I will start to work on the intelligent bracketing feature, by analysing the pictures...

And it seems that other guys are also working on this, so it should progress quickly.

Let's discuss this here....

Hi Frédéric!

Sorry for being here so rarely - lots of stress these days (bad weather for outside assignments - and bad times for assignments anyway . . )!

I tested the Breeze DSLR RemotePro shortly last week. Very fine functions to expand the bracketing options of the 20D. A second interesting option is to shoot burst of varying focusses! I didn´t do it yet but i learned it´s done by a script you can edit which tells the camera to focus editable steps between selected focus-points.
That´s a great tool.
The only negative aspect in my eyes is to do it on a laptop running Windows . . (i had to borrow my daughter´s)

Especially in combination with a Merlin these features are fantastic!

On the weeekend i´ll try your latest PW-update and from next week on i´ll be here some more often again . .

best to you, Klaus
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon Feb 09, 2009 9:39 am

Tethered shooting control software for Nikon DSLRs:
http://www.oxfordeye.co.uk/TetherPro/TetherPro.aspx
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

User avatar
Paul
Member
 
Posts: 778
Likes: 0 post
Liked in: 0 post
Joined: Sat Aug 30, 2008 4:46 pm
Location: Bonn, Germany
Info

by Paul » Mon Feb 09, 2009 9:55 am

Paul

close, but no cigar ... ... ...

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Mon Feb 09, 2009 10:28 am

klausesser wrote:A second interesting option is to shoot burst of varying focusses! I didn´t do it yet but i learned it´s done by a script you can edit which tells the camera to focus editable steps between selected focus-points.

Is it really working with your 20D? Or with any Canon DSLR?
Frédéric

no avatar
steve_g2
New member
 
Posts: 1
Likes: 0 post
Liked in: 0 post
Joined: Mon Feb 09, 2009 7:23 pm
Info

by steve_g2 » Mon Feb 09, 2009 7:33 pm


Hi
By way of a quick introduction I'm the owner of the go2graphics blog. Just to clarify, diyphotobits.com is the site owned by the guy who came up with the tethered script, I use it for some shoots and find it really good for my needs. Works very well with Adobe Lightroom 2 using LRs ability to monitor a folder in real time.

User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Mon Feb 09, 2009 11:19 pm

fma38 wrote:
klausesser wrote:A second interesting option is to shoot burst of varying focusses! I didn´t do it yet but i learned it´s done by a script you can edit which tells the camera to focus editable steps between selected focus-points.

Is it really working with your 20D? Or with any Canon DSLR?

Hi Frédéric!

As i learned it works fine - notes on the Breeze-Site - but i didn´t test it myself yet. I´ll do as soon as i can.
I just own ONE Canon-lens: a 17-58 kit lens - i very rarely use it, but for testing it should do.

http://www.breezesys.com/DSLRRemotePro/focus_stacking.htm

best to you, Klaus
Last edited by klausesser on Mon Feb 09, 2009 11:20 pm, edited 1 time in total.
Simplicity is the keynote of all true elegance. Coco Chanel

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Mon Feb 09, 2009 11:23 pm

I don't think it will work with the 20D:

"What do you need?

1. A Canon DSLR which supports live view e.g. EOS 40D, EOS 50D, EOS 5D Mark II, EOS 450D/Rebel XSi, EOS 1000D/Rebel XS, EOS-1Ds Mark III or EOS-1D Mark III"

:(
Frédéric

User avatar
Castillonis
Member
 
Posts: 264
Likes: 0 post
Liked in: 0 post
Joined: Thu Mar 06, 2008 8:49 pm
Location: Oakland, Ca
Info

by Castillonis » Tue Feb 10, 2009 11:03 pm

I also use Granite Bay's time lapse software which supports the Canon 20D

http://www.granitebaysoftware.com/CamerasSupported.aspx

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Wed Feb 11, 2009 8:55 am

Do you know if it is possible to control it from an external script?
Frédéric

no avatar
GURL
Member
 
Posts: 2943
Likes: 0 post
Liked in: 0 post
Joined: Tue Dec 06, 2005 1:57 pm
Location: Grenoble
Info

by GURL » Sun Feb 15, 2009 12:23 am

[quote=" www.kaiser-fototechnik.de ]"Zigview S2 Live can currently be used with these cameras: Canon 450D, 40D, 1D MkIII, 1Ds MkIII, Nikon D3, D300, D700, D90, Sony Alpha 350, Alpha 300."

http://www.kaiser-fototechnik.de/en/news/produkte/1_1_zigview_s2live.asp

Zigview S2 Live was not designed to control a motorized pano-head but to control a DSLR (from a short distance up to 100 m) and is certainly useful for photographing wide life animals, macro photography and other circumstances where a tripod is used.

For many of these applications being able to control the camera orientation would be useful...
Georges

no avatar
jeremyp
Member
 
Posts: 33
Likes: 0 post
Liked in: 0 post
Joined: Sun Feb 15, 2009 4:28 am
Location: Sydney, Australia
Info

by jeremyp » Mon Feb 16, 2009 2:52 am

Hi Frédéric,

Is your intention to automatically shoot high dynamic range images? I.e. a bracketed sequence that overall contains no clipped shadow or highlight information?

What about using wireless USB hub rather than tethered shooting? You could then have USB -> RS232 for the mount control.

Regards,

Jeremy (new member)

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon Feb 16, 2009 8:48 am

fma38 wrote:I finally asked on the gphoto2 ML, and, with the help of the developpers, found a way to change the exposure bias on my Canon EOS 20D. So, as soon as this feature will be implemented in libgphoto2, I will start to work on the intelligent bracketing feature, by analysing the pictures...

And it seems that other guys are also working on this, so it should progress quickly.

Let's discuss this here....

Frederic,

How to you envisage this working?

What form of analysis do you propose to undertake to analyse the pictures?

Could something similar be incorporated into APP/APG do you think to determine which images from a set of bracketed images should be used to make a pano?

I appreciate that you are seeking to avoid having 'useless' bracketed images in the first place, to speed up shooting when using Merlin/PW, but what about a similar process of automatic analysis/assessment/selection after shooting a full set of bracketed images being incorporated into APP/APG for those who don't shoot with Merlin/PW? Or for those shooting with Merlin/PW on a platform - such as a small handheld (like the Nokia Tablets) which cannot handle the processing, or who wish to shoot untethered, or whose cameras are not suffiiecently controllable in a tethered mode.

In a tethered mode do you think the automatic analysis/assessment/selection procss while shooting will be fast enough to save time compared to shooting a complete set of bracketed images and doing the automatic analysis/assessment/selection after shooting the pano?
Last edited by mediavets on Mon Feb 16, 2009 9:06 am, edited 1 time in total.
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Mon Feb 16, 2009 8:59 am

jeremyp wrote:Is your intention to automatically shoot high dynamic range images? I.e. a bracketed sequence that overall contains no clipped shadow or highlight information?

That's the idea ;)

What about using wireless USB hub rather than tethered shooting? You could then have USB -> RS232 for the mount control.

Do you have links to these wireless usb hubs? But the main problem will be the protocole to talk to the camera... That's why we discuss about 3rd party shooting softwares which can be scripted.
Frédéric

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Mon Feb 16, 2009 9:11 am

Andrew, I don't think that the speed will be the major improvement during shooting. The intelligent bracketing main goal will be to ensure that you will get all the dynamic of the scene.

Most of cameras can't shoot more than 3 exposures in auto bracketing mode, limited to +-EV. It is sometime noe enough. Not because the dynamic is greater than the auto bracketing can handle, but because the 0EV exposure is not always centered; exposure algos of modern cameras are so complex than it is impossible to predict what they will do. Then, +2 or -2EV is not enough. And using manual mode is not possible, because a wide pano will have too much difference in ligh from one side to the other.

I don't know yet how to analyse an image to predict the next exposure to shoot, but I think your idea is good: as long as you get enough pictures to cover all the dynamic (using a lot of bracketed pictures, and so having much more than needed in most cases), it will be possible to remove those wich don't add usefull information...

I'm not a math expert, so I will need some help for this part. If you find usefull papers, feel free to share ;) GURL's idea was to use the algos implemented in tufuse/enfuse:

http://research.edm.uhasselt.be/~tmertens/papers/exposure_fusion_reduced.pdf

Look at chapter "3.1. Quality Measures": the "Well-exposedness" param could be a good starting point...
Frédéric

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon Feb 16, 2009 9:17 am

fma38 wrote:Do you have links to these wireless usb hubs? But the main problem will be the protocole to talk to the camera... That's why we discuss about 3rd party shooting softwares which can be scripted.

http://www.usb.org/developers/wusb/
http://www.itreviewed.co.uk/article.php?id=2572
http://www.wisair.com/
http://www.everythingusb.com/cables-unlimited-wireless-usb-adapter-kit-15823.html
http://www.ritzcamera.com/product/EP53008144.htm
Last edited by mediavets on Mon Feb 16, 2009 9:27 am, edited 1 time in total.
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

User avatar
Paul
Member
 
Posts: 778
Likes: 0 post
Liked in: 0 post
Joined: Sat Aug 30, 2008 4:46 pm
Location: Bonn, Germany
Info

by Paul » Mon Feb 16, 2009 9:49 am

the max. distance limitations are very restrictive:

CWUSB max 10m

versus:
bluetooth class 2 30m
bluetooth class 1 100m
Paul

close, but no cigar ... ... ...

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Mon Feb 16, 2009 10:10 am

The main problem will be the drivers... Anyway, this is not the main problem for now. Let's first make something working with a wire ;)
Frédéric

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon Feb 16, 2009 11:21 am

fma38 wrote:I don't know yet how to analyse an image to predict the next exposure to shoot, but I think your idea is good: as long as you get enough pictures to cover all the dynamic (using a lot of bracketed pictures, and so having much more than needed in most cases), it will be possible to remove those wich don't add usefull information...

I think GURL has already experimented (sucessfully) doing this manually.

Taking far more bracketed shots than may be necessary is perhaps a rather heavy handed approach - and time consuming with longer exposures - but it should work as a first pass at the problem, with selection done after shooting, perhaps via an import filter for APP/APG? Rejecting bracketed images that will add nothing should at least be pretty simple. Could that be one way to determine/limit the end points for a range of exposures while tethered shooting - ie. to detect when images become either totally over exposed or totally underexposed.

I'm not a math expert, so I will need some help for this part. If you find usefull papers, feel free to share ;) GURL's idea was to use the algos implemented in tufuse/enfuse:

http://research.edm.uhasselt.be/~tmertens/papers/exposure_fusion_reduced.pdf

Look at chapter "3.1. Quality Measures": the "Well-exposedness" param could be a good starting point...

The paper is interesting. I don't begin to understand the maths but I can see that they get decent results and that the process is reasonably fast.

It would be very interesting to have such a process incorporated in to APP/APG especially if it was able to use GPU power.
Last edited by mediavets on Mon Feb 16, 2009 11:22 am, edited 1 time in total.
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Mon Feb 16, 2009 11:44 am

mediavets wrote:Taking far more bracketed shots than may be necessary is perhaps a rather heavy handed approach - and time consuming with longer exposures - but it should work as a first pass at the problem, with selection done after shooting, perhaps via an import filter for APP/APG? Rejecting bracketed images that will add nothing should at least be pretty simple. Could that be one way to determine/limit the end points for a range of exposures while tethered shooting - ie. to detect when images become either totally over exposed or totally underexposed.

It would be better to avoid taking under/over-exposed pictures, rather than remove them. And as I said, you are not even sure to capture all the dynamic using only -2,0,+2 exposed pictures. As the maths will be the same...
Frédéric

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon Feb 16, 2009 12:07 pm

fma38 wrote:
mediavets wrote:Taking far more bracketed shots than may be necessary is perhaps a rather heavy handed approach - and time consuming with longer exposures - but it should work as a first pass at the problem, with selection done after shooting, perhaps via an import filter for APP/APG? Rejecting bracketed images that will add nothing should at least be pretty simple. Could that be one way to determine/limit the end points for a range of exposures while tethered shooting - ie. to detect when images become either totally over exposed or totally underexposed.

It would be better to avoid taking under/over-exposed pictures, rather than remove them. And as I said, you are not even sure to capture all the dynamic using only -2,0,+2 exposed pictures. As the maths will be the same...

Perhaps the maths would not be quite the same if one was only looking for totally over/underexposed images; might that not be simpler maths/process? I was just suggesting what I thought (with no understanding of the maths) might be a more simple way to determine the end points for a range of exposures given that the range and number of exposures required to capture full dynamic range cannot, as you say, be easily determined in advance for any or all shooting positions in a pano.
Last edited by mediavets on Mon Feb 16, 2009 12:31 pm, edited 1 time in total.
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

User avatar
fma38
Moderator
 
Topic author
Posts: 5850
Likes: 2 posts
Liked in: 2 posts
Joined: Wed Dec 07, 2005 6:21 pm
Location: Grenoble, France
Info

by fma38 » Mon Feb 16, 2009 12:33 pm

I think maths *are* very close: even if it is easy to see that a picture is under/over-exposed, how do you know that you can remove it or not? You have to know that the next one covers the parts of that wrong picture. So, it is the same as predicting the next correct exposure bias to cover all the dynamic...

Am I wrong?
Frédéric

no avatar
mediavets
Moderator
 
Posts: 16415
Likes: 2 posts
Liked in: 130 posts
Joined: Wed Nov 14, 2007 2:12 pm
Location: Isleham, Cambridgeshire, UK.
Info

by mediavets » Mon Feb 16, 2009 1:27 pm

fma38 wrote:I think maths *are* very close: even if it is easy to see that a picture is under/over-exposed, how do you know that you can remove it or not? You have to know that the next one covers the parts of that wrong picture. So, it is the same as predicting the next correct exposure bias to cover all the dynamic...

Am I wrong?

I don't know.

I guess it all depends on the starting point exposure in a bracketed series.

If shooting manually with manual pano head and my oh-so-simple D40 I'll set my desired aperture in aperture priorty mode swing the camera around the scene watching the metering and then 'guesstimate' a manual exposure for all shots, but of course that's with FE lenses usually. I'll then set the manual exposure, take a test shot and look at the histogram and adjust accordingly my 'universal' manual exposure setting - with FE lenses the D40 metering usually seems to underexpose shots left to itself. It's quite easy with the D40's using the LCD for shooting data display and the simple system for adjusting the main parameters. I don't know whether this is a good/ideal approach but it seems to work for me. Perhaps this too could be automated?

Of course my camera lacks AEB, but as you say most camera's AEB capability is too limited to be of much use anyway. So that's the best I can do other than manually bracketing - I tried it once and it was a PITA.

Anyway, if one took that sort of 'average/universal' manual exposure as a starting point - set by the user - then I reckon that with a typical pano scene you could then shoot brackets automatically incrementally adjusting exposure up and down (in user defined steps), using third party software perhaps to control the camera, until you hit the limits at which the image was either totally underexposed or totally overexposed.

I don't know whether that makes any sense, but its what I was thinking about just off the top of my head as a conceptual approach.

I'm sure GURL would be able to come up with a far more scientific, precise and rigorous plan.;)
Last edited by mediavets on Mon Feb 16, 2009 4:13 pm, edited 1 time in total.
Andrew Stephens
Many different Nodal Ninja and Agnos pano heads. Merlin/Panogear mount with Papywizard on Nokia Internet tablets.
Nikon D5100 and D40, Sigma 8mm f3.5 FE, Nikon 10.5mm FE, 35mm, 50mm, 18-55mm, 70-210mm. Promote control.

User avatar
klausesser
Member
 
Posts: 8836
Likes: 5 posts
Liked in: 64 posts
Joined: Mon May 22, 2006 12:18 am
Location: Duesseldorf, Germany
Info

by klausesser » Mon Feb 16, 2009 2:05 pm

fma38 wrote:I don't think it will work with the 20D:

"What do you need?

1. A Canon DSLR which supports live view e.g. EOS 40D, EOS 50D, EOS 5D Mark II, EOS 450D/Rebel XSi, EOS 1000D/Rebel XS, EOS-1Ds Mark III or EOS-1D Mark III"

:(

Hey Frédéric!

Somewhere i read about it works also with 20D - will look for it.

best, Klaus

P.S.: the expanded bracketing mode works on my 20D - i tested it last year. In the Breeze´s specs the 20D doesn´t appear - i guess it´s simply too old to list it. But it should work. I´ll test again this afternoon!
Last edited by klausesser on Mon Feb 16, 2009 2:12 pm, edited 1 time in total.
Simplicity is the keynote of all true elegance. Coco Chanel

no avatar
GURL
Member
 
Posts: 2943
Likes: 0 post
Liked in: 0 post
Joined: Tue Dec 06, 2005 1:57 pm
Location: Grenoble
Info

by GURL » Mon Feb 16, 2009 7:21 pm

mediavets wrote:Of course my camera lacks AEB, but as you say most camera's AEB capability is too limited to be of much use anyway. So that's the best I can do other than manually bracketing - I tried it once and it was a PITA.

The pity is just that you made a single attempt where, for ordinary peoples, severals attempts are needed... ;)

If you want to know more about that you should get The HDRI Handbook - Christian Block - rockynook - http://www.hdrlabs.com.

This book includes some boring chapters like chapter 2 (file formats) and describes some matters most photographers don't care (chapter 7) but:
1) a large number of bracketed shots (ordinary or panoramic) are included in the CD-ROM and, as they were used for the examples shown in the book, you can find the right tricks in the book and can compare your own results. It's both faster and easier than shooting your own examples before any attempt at using the described applications (they are included on the CD, too)
2) the difference between HDR recording and HDR rendering is clearly stated. As a rule, peoples don't set aside tone mapping and recording and this is very confusing
3) last but not least, the kind of moderate tone mapping where the results don't share the so called "HDR look" is well covered...

Besides that, the theory and the maths behind HDR recording are straightforward and easy to understand: just take as many shots as you need to insure that for each of the final image pixels and each of the red, green and blue colors at least one of the source images recorded a value next to 128. The maths? a simple rule of three will be enough! (the huge amount of computing is to be left to computers, don't mind ! [small color changes could be introduced when adjustment steps are too large, best results probably implies that using RAW values would be better: avoiding clipped colors is more important.])

How the photographer will latter use the bracketed shots should not matter and should be unknown from PapyWizard (that is PapyWizard should work for any tone mapping applications, *fuse included.)

On the other hand the practice hurts a lot.
- the more shots one need the more moving parts in the subject will cause ghosts
- the more shots one need the more stitching will be difficult (a dark forest of links with the current APP for example)
- the higher the exposure differences between shots the more stitching will be difficult (no CP to found between the darkest and the brightest image - future versions of APP/APG will avoid this problem.)

fma38 wrote:I think maths *are* very close: even if it is easy to see that a picture is under/over-exposed, how do you know that you can remove it or not? You have to know that the next one covers the parts of that wrong picture. So, it is the same as predicting the next correct exposure bias to cover all the dynamic...

Am I wrong?

:cool: possibly !

The easiest criterion to avoid useless shoots could be :
1) stop using longer exposure values when every image pixel reaches a value higher than 128 (or another threshold value to be found by appropriate tests, this could take into account the fact that the lowest recorded level in a given shot is larger than some ISO dependant noise-free level)
2) stop using shorter exposure values when every image pixel reaches a value lower than 128 (or another threshold value to be found by appropriate tests, this could take the used camera highlights clipping level into account)
3) it's quite obvious that in most occasions accurately recording the light sources level (like the sun or other light sources) is useless. On the contrary being able to record some details on the moon in a night landscape or reflected lights could be desirable. This suggest a maximum exposure value being used and the corresponding setting being available to users.

Selecting the bracketing starting point should be left to the photographer because the best method will depend of the camera (presently I'm using matrix metering and average metering on my camera: no conclusion yet, I'm not sure one is better than the other, looks like it depends of the circumstances.)

The step size should be left to user (1 EV or 2 EV is common, I still have to test 3 or 4 EV for my Fuji S5)

This is not easy stuff but I'm quite confident that avoiding useless bracketed shots is possible for a computer (or removing them latter, whichever is easiest would be better than nothing). My guess: analyzing the thumbnails rather than the real shots should be enough.

:lol: Another solution would be to use spot metering. PapyWizard could scan the subject step by step using the the motorized panohead! This could look very smart, I doubt it would be useful when lighting changes are the rule (not very often useful, that is.)

The plain truth: PapyWizard shooting a non bracketed series first (using a given "manual" exposure or any kind of automatic exposure) and then suggesting some bracketed shots because it guesses they could be useful is something I would like to have. Presently, the best I can do to avoid large bracketed series is to shot one image and inspect the corresponding histogram before going to the next position. This sucks. The histogram displayed by many (if not all) cameras is difficult to read when the camera is on a panohead (under the sun, etc.) Blinking highlights are often easier to read but (... just try it! :o )

Manually removing useless bracketed shots (where useless points to shots that don't carry any more missing data values) works when using Enfuse and anyways Tufuse was explicitly designed in such a way that this can work (an example is shown in Max Lyons documentation.) Autopano seems to prefer that, too. Photomatix: ?

fma38 wrote:exposure algos of modern cameras are so complex than it is impossible to predict what they will do.

J'ai trouvé ça, je suis pas tellement plus avancé : http://www.kenrockwell.com/nikon/matrix01.htm sauf sur un point: sur les exemples (pages 6 à  14) je préférerais le plus souvent (mais pas toujours) le résultat de la mesure non matricielle si j'avais qu'une seule photo à  ma disposition. Combiner les deux avec Enfuse donnerait souvent un résultat meilleur que d'avoir seulement l'une ou seulement l'autre.

En gros (il me semble l'avoir déjà  dit ailleurs) l'expo matricielle telle qu'elle est réglée par les constructeurs tend surtout à  éviter le "elle est trop sombre, on y voit que dalle sur ta photo!" Autrement dit tant pis pour les morceaux de ciel étrangement colorés autour du soleil et les nuages noyés dans un blanc jaunâtre du moment qu'on peut reconnaître la tante Philomène, celle qui est assise à  l'ombre là -bas au fond. Personnelement je préférerais l'histogramme juste calé à  droite (mais pas trop à  droite) ce qui ne semble être jamais disponible sans que la remontée des ombres style D-Lighting soit enclenchée automatiquement par l'appareil (après quoi il me semble que les paramètres EXIF perdent pas mal de leur signification...)
Last edited by GURL on Mon Feb 16, 2009 8:11 pm, edited 1 time in total.
Georges

Next

Who is online

Users browsing this forum: No registered users and 1 guest