HDR is not working as expected...
enridp
Maybe I'm doing something wrong but I don't know what, this is what
I'm doing:
1) Open the images with Camera Raw (5.6)
2) Save them as tif (I see an option at the bottom it says: Adobe RGB
(1998); 8 bit; ... do I need to change to 16 bit?)
3) then change Exposure to 1 and save tifs again
4) then change Exposure to -1 and save tifs again
5) Load all the exported tifs (0,+1 and -1) in PTGui
6) the dialog for working with HDR never appears, and if I go to
Exposure/HDR and click on Enable HDR stitching I see the message:
"All source images were taken with the same exposure settings. To
create an HDR panorama , take bracketed images..."
I think the reason is because in the EXIF data all my tifs have the
same exposure time.
Does somebody know how to solve this?
Because if I stitch the 0EV tifs, and then apply a template to the
others, I have a blurred result after converting to HDR with
Photomatrix (I don't know why because the images comes from the same
RAW).
Regards!!
Enrique.
Ryan Gardner
You are doing it wrong.
To do hdr you need more than one exposure. Faking it by spitting out multiple 8 bit tiffs is a waste of time.
Spit out a single 16-bit tiff with all the info of the original and stitch a 16-bit ldr panorama and that will be the best you can get.
Faking hdr gives you nothing.
Ryan
> You received this message because you are subscribed to the Google Groups "PTGui" group.
> To post to this group, send email to pt...@googlegroups.com
> To unsubscribe from this group, send email to ptgui+un...@googlegroups.com
> Please do not add attachments to your posts; instead you may upload files at
> http://groups.google.com/group/ptgui/files
> For more options, visit this group at http://groups.google.com/group/ptgui
michael crane
> You are doing it wrong.
>
> To do hdr you need more than one exposure. Faking it by spitting out
> multiple 8 bit tiffs is a waste of time.
>
> Spit out a single 16-bit tiff with all the info of the original and stitch a
> 16-bit ldr panorama and that will be the best you can get.
>
> Faking hdr gives you nothing.
no it is useful to extract different tiffs from the raw and to blend
them in some other software.
Kevin Wilton
You guys are now all systematically tearing the lungs out of version 9.
I am have been working with version 6 for years and am really still very happy with the performance of the program, whether making simple 3 or 4 image pans to 360x360 worlds. And, if I need to do an HDR, I know I can do a brilliant job in other software, not Photoshop I might add and then produce the pan.
So my question is this, what would I gain from moving up to the next stable version, which I presume is version 8?
Open question to everyone.
Kev
Ryan Gardner
Nope. A single 16 bit tiff will contain more info than those 3 faux exposures.
Just render it as a 16 bit tiff and stitch it.
Then if you want to get your jollies by running it through an hdr app you can.
But regardless of what you call it, its never going to be hdr since you are limited by the dynamic range of your source.
Ryan Gardner
In my experience Hdr in ptgui 9 works just fine... you just have to have an hdr source.
michael crane
> In my experience Hdr in ptgui 9 works just fine... you just have to have an
> hdr source.
Kevin Wilton
I am curious what you could be doing, because if you cannot split out 3, 5
or7 different exposures from a RAW file and make a great HDR then you must
be doing something wrong.
Just make the required number of copies of the original image, in the same
folder, call them 01 - 07, or whatever.
Open them all in ACR, Lightroom, PhaseOne, Nikon whatever you use and make
+1, +2/3, +1/3, 0, -1/3, -2/3, -1 exposure adjustments, or +1.5, +1, +.5,
0, -.5, -1, -1.5, whichever sized step you wish to work with then save the
changes and process in your normal HDR software. Or create TIFFs and then
work the HDR if the software you use doesn't support RAW file input.
If you want to get really interesting results, and I mean funky, make 5
copies of your final HDR output and open all of those in ACR and make +1,
+.5, 0, -.5, -1, exposure adjustments, then combine these for a second time
in a bespoke HDR program.
Kev
--
Ryan Gardner
2 or more bracketed files... either as camera raw or 16 bit tiffs.
Or a 32-bit hdr file (hdr, exr etc)
For the faux hdr if you could merge your exposures into one of those files first.
You can export all you want from a single raw file, but it is just a waste if time since you are providing no additional dynamic range compared to the 16-bit tiff export
Ryan Gardner
can only be stored in 32-bit floating point file (like exr, 32-bit
psd, hdr) etc.
Saying you can extract multiple exposures from a single raw is true -
but it's a waste of time. The camera RAW stores 12 or 14 bits of
information, so a 16-bit TIFF can store ALL of that same information.
Exporting a +/- 1 or +/- 2 EV tiff file from a source raw provides you
ABSOLUTELY NO additional highlight or shadow detail. It might look
different because the raw converter will apply a default S-Curve to
each exposure so when you merge them you get a different looking
result - but you could achieve the same look by applying a curve to
the 16-bit tiff file.
Ryan
Kevin Wilton
-----Original Message-----
From: pt...@googlegroups.com [mailto:pt...@googlegroups.com] On Behalf Of
Ryan Gardner
Sent: Sunday, 14 November 2010 11:05 AM
To: pt...@googlegroups.com
Subject: Re: [PTGui] HDR is not working as expected...
Ryan
> ++.5,
> 0, -.5, -1, -1.5, whichever sized step you wish to work with then save
> the changes and process in your normal HDR software. Or create TIFFs
> and then work the HDR if the software you use doesn't support RAW file
input.
>
> If you want to get really interesting results, and I mean funky, make
> 5 copies of your final HDR output and open all of those in ACR and
> make +1,
> +.5, 0, -.5, -1, exposure adjustments, then combine these for a second
> +time
enridp
Well, maybe it's not a true HDR, but we can make a fake HDR....
There's a lot of examples of HDR from a single RAW (even a single
JPG!)
And when I asked this here a few time ago, one of the options was just
that, extract the exposures from my RAW, I'm more confused now :(
I've tried what you said, save my RAW with Camera Raw as tif (changing
to 16 bits before), stitched those 16bit tiffs, and I have a nice
stitching, but without tone mapping :(
Also, I can't use it with Photomatrix...
I want to make dramatics tonemappings, I'm not interested in HDR
specifically.
I want to make things like these:
http://imagengratis.org/images/30322029284d8.jpg
http://imagengratis.org/images/img246678tone.jpg
http://imagengratis.org/images/gdanskoldcity.jpg
can you help me to do that from a single RAW? (I need to shoot action
panoramas that's because I need to use a single image)
On 13 nov, 21:06, Ryan Gardner <ryeb...@gmail.com> wrote:
> You are doing it wrong.
>
> To do hdr you need more than one exposure. Faking it by spitting out
> multiple 8 bit tiffs is a waste of time.
>
> Spit out a single 16-bit tiff with all the info of the original and stitch a
> 16-bit ldr panorama and that will be the best you can get.
>
> Faking hdr gives you nothing.
>
> Ryan
Ryan Gardner
Right. That's why until you get a true 32-bit sensor you must merge multiple exposures to get real HDR image.
Ryan Gardner
You can do that with the 16 bit image.
Photomatix should be able to do it fine, but if it can't you could take the stitched panorama and convert it to a 32bit format in Photoshop and then run it through whatever tone mapping app you want.
michael crane
inform others so I am going to ignore what you say for the time being.
toodle pip
Ryan Gardner
pretty helpful link I found after 10 seconds of searching:
http://www.hdrsoft.com/resources/dri.html
in particular #raw :
http://www.hdrsoft.com/resources/dri.html#raw
"Can't I just create the exposures from one RAW file?
Not really. Your RAW file contains data captured by the sensors for
only one exposure. The total dynamic range you can reconstruct from
one photo converted with different exposure settings can never be more
than the dynamic range captured by your camera, and this is rather
limited (see above).
When you are using only one exposure to capture the scene, your RAW
file is already your HDR image.
Converting the RAW file to images with different exposure levels is a
bit like slicing the dynamic range of the RAW into several parts.
Combining the parts back into an HDR image will at best re-produce the
dynamic range of the initial RAW file...."
It's from HDRSoft - the people who make Photomatix (But then again,
maybe you aren't sure if they understandthings sufficiently to be able
to inform others?)
Another HDR software maker - Unified Color has this to say on their FAQ:
"Can I create 3 bracketed images (JPEG or TIFF) from a single RAW file
and merge them in HDR Expose?
You can certainly do so, however you must note that this process will
not create any new information in the image. You can just load the
original RAW file into HDR Expose and process it from there to obtain
same level of details in the final image. The image result may be
different because conversion to JPEG or TIFF changes the image
information (typically RAW converters apply various S-curve
corrections), however to note again this manipulation cannot create
any new image information (shadow or highlight details) — you need to
capture it from the original scene with true camera bracketed shots."
Although maybe they also don't understand things enough to explain it?
The idea of splitting a single raw file out to multiple 8-bit
exposures for the sole purpose of getting more dynamic range out of
the source is pretty ridiculous. It's almost as stupid as cropping all
of your images into 3 images manually with some overlap and then
loading them into PtGui and stitching them and thinking that by having
more images you are going to get a better result.
If you are just interested in "tone mapping", I've already explained
that to someone else - but using 3 8-bit exposures instead of one
16-bit one not only increases your workflow for not benefit, it
doesn't give you anything as a benefit.
So to answer your original question - just use a single 16-bit TIFF
file spit out then load it into whatever you want to tone map /
dynamic range map / whatever you want with it it will work just fine
and be a heck of a lot faster than trying to use 3 8-bit tiff files to
do who knows what.
Now, if you are insisting on using them - it sounds like PtGui is
reading the EXIF and seeing that they were all taken at the same
exposure - in advanced image parameters you can go in there and tell
what the exposure value is on the other ones to force it to think of
them as HDR - but like I said... it's a waste of time because you
aren't adding any more dynamic range.
If you only shot a single exposure, just blend a single exposure (in
16-bit more). it's a much more streamlined workflow and you end up
with a single 16-bit image as the result. From that single 16-bit
image you can do the whatever you want with that has as much dynamic
range as possible from your source image.
Ryan
On Nov 13, 2010 6:17 PM, "michael crane" <mick....@gmail.com> wrote:
enridp
image from a single RAW, but I want to apply tone mapping to the final
panorama.
So, there are two options:
1) I'm doing it wrong
2) Extracting differents exposures doesn't add any information, but
helps to Photomatrix to make a good tonemapping.
If I'm doing something worng I don't know what, please tell me if you
know :(
Here are my tests:
http://www.imagengratis.org/images/tonemappingfr.jpg
I like the tonemapping obtained with 3 exposures, and I don't know why
I couldn't obtain the same result with my TIF16.
I'm tonemapping after stitching, is that an error? (anyway, it's a lot
of work to tonemap before stitching and Photomatrix it's sooo slow
opening RAW files...)
I've tried what you said too, changing the exposure in Advanced Image
Parameters, but I don't know what numbers I must enter, default is
1/400, so I put 1/200 (for +1EV) and 1/800 (for +1EV). It worked, and
I can change the tonemapping inside PTGui directly without
Photomatrix, that's great, although I think I must practice because
the shadows are too dark.
This is my output from PTGui (3 exposures from RAW, 0,+1 and -1,
exported as TIFF8, loaded in PTGui, changed exposures, linked images,
tonemapped inside ptgui)
http://www.imagengratis.org/images/tonemappedfro.jpg
Can we apply tonemapping inside PTGui with a single TIFF16?
Kevin Wilton
You're being a bit pedantic about this, like a few others I could mention on
other issues, but I won't.
Okay, while the output from a range of adjusted images from a single RAW
image is not by a strict and narrow definition a true HDR, the end result
has the appearance of one and for all intense and purposes is the very next
best thing.
The final result has the 'over cooked' feel of an HDR image, there is detail
in areas that otherwise did not have detail and the effect is more than
satisfactory in most circumstances.
The additional advantage of making an HDR, oops, a fake HDR image from a
single RAW file this way is the non-requirement to produce half a dozen
images per shot, which takes up valuable set up time and equally precious
card space. Not to mention the need for a totally stationary subject and a
tripod for every set of shots. It also does not require you to set out with
the intention of producing an HDR, or the need to carry a tripod everywhere
on the off chance you see a shot that needs the treatment. It also frees
you, well me and the rest of us in this thread to make an HDR, oops did it
again, fake HDR later when back in the studio/office from an image that was
never shot for that purpose.
Does that make you happy?
Kev
-----Original Message-----
From: pt...@googlegroups.com [mailto:pt...@googlegroups.com] On Behalf Of
Ryan Gardner
Sent: Sunday, 14 November 2010 12:12 PM
To: pt...@googlegroups.com
http://www.hdrsoft.com/resources/dri.html
in particular #raw :
http://www.hdrsoft.com/resources/dri.html#raw
details) - you need to capture it from the original scene with true camera
bracketed shots."
Ryan
Ryan Gardner
look, and you can get a decent looking image that way.
I see what you are talking about - the middle one looks better than
the other two because it is recovering the shadow detail a lot better.
I've not played with photomatix in a long time - so I'm not sure how
good it is at keeping the color information from the single image, but
I have spent a lot of time recently using 32 Float in photoshop (and
before that using HDR Expose standalone) - and I'm pretty sure that it
could get a similar result as what you are looking for as the end
result.
Can you email me that original raw file? Or if it's bigger than 20
megs maybe put it on filedropper or something? I'm pretty sure I can
get a result that will look very similar to what you are saying you
like with the 3 exposure version just using the raw file (or a 16-bit
tiff, actually) using HDR Expose or 32 Float - which wouldn't help
you much, but if you can streamline your workflow to be exporting
16-bit TIFFs from a RAW file and then stitching in PTGui and then
adjusting the contrast in a tool (from the workflow of opening your
raw and spitting it out 3 times and then adjusting the contrast in a
tool) you'd see it go a lot faster.
Ryan
Eric O'Brien
bracketed exposures to feed into a tone mapper to get a look you
want. I don't know if you can tone map a single image in PTGui... I
don't personally care for the "distressed" look of cranking the dials
way over in a tone mapper, so I haven't tried doing this.
But you certainly *can* use just a single image with Photomatix Pro
4.0 -- open a raw file, or open a tiff file and adjust the controls
all you want. (Photomatix even has a preset named "Grunge!")
I don't want to take the time right now to see if Photomatix treats a
single image any differently than a synthesized (faked) set of
bracketed photos. I suppose it might. Right now, I just don't see
the point in taking the time to create synthetic bracketed exposures.
As others have said, if you are looking for *true* HDR, there is no
point, technically, in doing so.
As far as see at the moment, neither is there any point, practically,
in doing so if all you want is a tone mapped "look."
eo
PTGui Support
need to make sure that they have the right fake exposure times.
After loading your images in PTGui switch to Advanced mode and go to the
Image Settings tab. Enter fake values in the Exposure column, for
example 1/100, 1/400 and 1/25. The exact values don't matter, as long as
the differences match the EV values that you've extracted.
But I agree with Ryan that this is a bit of a waste of time. If you just
want the tone mapped look, PTGui can tone map non HDR images as well.
Convert your images to 16 bit TIFFs, load those in PTGui. Then go to the
HDR tab, enable HDR and configure the tone mapping or exposure fusion
settings.
Joost
>> any new image information (shadow or highlight details) � you need to
PTGui Support
Quite a lot has changed, here's the version history:
http://www.ptgui.com/versionhistory.html
And these are the changes in the upcoming version 9:
http://www.ptgui.com/beta.html
If there's something in it for you consider upgrading; otherwise feel
free to stick with version 6.
Joost
On 11/14/2010 1:41 AM, Kevin Wilton wrote:
> You guys are now all systematically tearing the lungs out of version 9.
>
> I am have been working with version 6 for years and am really still very
> happy with the performance of the program, whether making simple 3 or 4
> image pans to 360x360 worlds. And, if I need to do an HDR, I know I can
> do a brilliant job in other software, not Photoshop I might add and then
> produce the pan.
>
> So my question is this, what would I gain from moving up to the next
> stable version, which I presume is version 8?
>
> Open question to everyone.
>
> Kev
>
> *From:*pt...@googlegroups.com [mailto:pt...@googlegroups.com] *On Behalf
> Of *Ryan Gardner
> *Sent:* Sunday, 14 November 2010 10:07 AM
> *To:* pt...@googlegroups.com
> *Subject:* Re: [PTGui] HDR is not working as expected...
>
> You are doing it wrong.
>
> To do hdr you need more than one exposure. Faking it by spitting out
> multiple 8 bit tiffs is a waste of time.
>
> Spit out a single 16-bit tiff with all the info of the original and
> stitch a 16-bit ldr panorama and that will be the best you can get.
>
> Faking hdr gives you nothing.
>
> Ryan
>
> On Nov 13, 2010 4:49 PM, "enridp" <enrique...@gmail.com
> <mailto:pt...@googlegroups.com>
> > To unsubscribe from this group, send email to
> ptgui+un...@googlegroups.com
> <mailto:ptgui%2Bunsu...@googlegroups.com>
> > Please do not add attachments to your posts; instead you may upload
> files at
> > http://groups.google.com/group/ptgui/files
> > For more options, visit this group at
> http://groups.google.com/group/ptgui
>
> --
> You received this message because you are subscribed to the Google
> Groups "PTGui" group.
> To post to this group, send email to pt...@googlegroups.com
> <mailto:pt...@googlegroups.com>
> To unsubscribe from this group, send email to
> ptgui+un...@googlegroups.com
> <mailto:ptgui+un...@googlegroups.com>
Matthew Rogers
Matt
Matthew Rogers
I just wish people would stop confusing HDR images with tonemapped images. Yes you need to produce one to get to another but that's where it should end.
Matt
Bjørn K Nilssen
If you're lazy and looking for a fast way to make tonemapped (fake) HDR
images you'll find that it will be much faster if you spend some time
learning how to use PS's layers, layer masks, layer blending (don't forget
the split black/white point options) and blending modes, as well as all
the adjustment layers, shadow/highlight etc. The high-pass filter is
particularly useful for getting that over-cooked and haloed 'HDR-look'
(besides being very useful for other purposes).
Ever tried to save a hopelessly over-exposed shot by multiplying a copy on
top of the base, and seen the magical appearance of details that were 'not
visible' in the original? It's just like the fake HDR technique - if there
are any details there it will be possible to make them more visible by
using the right tools. One of these tools is the fake-HDR method, but it
is just so much faster to use PS on the 16bit (or even 8bit) - and you get
a lot more control.
> Does that make you happy?
I hope it will make you happy to save all the time spent on making all
those fake RAW conversions? ;)
--
Bjï¿œrn K Nilssen - http://bknilssen.no - 3D and panoramas.
floz
not mentioned at least, is the contrast setting / curve and clipping
settings applied to the RAW file.
And Ryan, no, what the thread starter is trying to do is not nonsense;
what you are saying however is more so.
Here's an illustration to go along with this post: [link]
(1) You shoot an image in RAW. This file contains the dynamic range
the camera / sensor is capable of (that's what the DXO marks and so on
tell you), say 12 stops altogether. The RAW file stores this range of
brightnesses and uses 12 or 14 bit per channel to sample this range of
brightness (like a 8 bit or 16 bit WAV file samples the audio levels,
where more bits result in finer graduations).
(2) You then open the file in your RAW converter, which de-mosaics the
data etc. to cook up an image you can actually look at. It also
applies an S-curve (adding contrast) so that the images looks more or
less "normal" and "punchy". Say you use the default contrast curve,
contrast setting 25 and black clipping 5 in ACR. The resulting image
you can store as 8 or 16 bit TIFF. 16 bit is more finely graduated,
but here's the thing: 0 brightness in all three channels still
describes the *same* (darker) physical brightness value in both files;
so does the 255 or 65535 in each channel describe the *same*
(brighter) physical brightness value in both files. *So just storing
your file as 16 bit, if I'm not mistaken that is, only gives you more
values in-between the darkes and brightes point, but not beyond!!* So
no, Ryan, while saving as 16 bit does give you more details in the
darks and brights, it does not extend the *range* of brightnesses you
have store in the file. ==> See (3).
(4) Now you change the parameters in Lightroom, set black clipping to
0, reduce contrast everywhere and even use the recovery slider. This
will extend the overall range of brightnesses you squeeze into your
TIFF file, again, be it describe with 8 or 16 bit per channel.
(5) This is what the thread starter is intending to do; and I think it
does make sense... Of course, truly shooting different exposures will
give you better results (static scene, tripod assumed), as, for
instance, +2EV image will be much noisier than an image actually
captured with a 4 times longer exposure time . . .
Anyway, did I explain all this correctly? I haven't done any
background reading on this, but this is how I understand the way
brightness and exposure works in cameras and files. [Of course, all
the brightness distributions are logarithmic, but I'm not trying to
complicate things.]
floz
enridp
This is the link to download the RAW files (70MB, 6 RAW):
https://www.yousendit.com/download/dklyV0p4Z1BtNEt4dnc9PQ
I'm really interested not only in learning more to make the things
right, also for working faster, even more, my goal is to make an
automated process, because I'm planning to take a lot of panoramas
(like Google Street View, well... not so much) and I need a way to
automate the whole process.
Joost:
I have tried what you said, opening my TIFF16, aligning, selecting HDR
(omit the red warning) and tonemapped the result, I think this could
work too, but I have a problem with tonemapping in PTGui: it's a
really great feature, but Photomatrix has more options, and my big
problem is the size of the preview, I can't see details so when I
stitch the final result I alwways see that my settings was not
correct.
are you planning to add more options for Tonemapping in PTGui9? or
maybe is possible to add Photomatrix like a plugIn for PTGui...
is possible to zoom an area to see details in the tonemapping editor?
Matthew and Bj:
I know that it's not a true HDR, and I have remarked that my goal is a
tonemapped image, maybe it's possible to simply edit it in Photoshop
to fake that effect, but I need the easiest and fastest way, and also
automated... and I'm not an expert in Photoshop :(
Tonemapping is easy to me, I just need to slide some controls and see
what happens. And I think we can automate the process, I'm pretty sure
PTGui can do that (but I don't know how), also there's a "Batch
Processing" button in Photomatrix.
I'm a beginner in this, and I'm sure I'm doing many things wrong but
I'm reading and testing all your tips.
Also I think there's a lot of confusing information in internet, maybe
could be a good idea to put the best workflow for (fake)HDR with a
single RAW inside PTGui FAQs.
Bjørn K Nilssen
<enrique...@gmail.com>:
> Ryan:
> This is the link to download the RAW files (70MB, 6 RAW):
> https://www.yousendit.com/download/dklyV0p4Z1BtNEt4dnc9PQ
>
> I'm really interested not only in learning more to make the things
> right, also for working faster, even more, my goal is to make an
> automated process, because I'm planning to take a lot of panoramas
> (like Google Street View, well... not so much) and I need a way to
> automate the whole process.
>
> Joost:
> I have tried what you said, opening my TIFF16, aligning, selecting HDR
> (omit the red warning) and tonemapped the result, I think this could
> work too, but I have a problem with tonemapping in PTGui: it's a
> really great feature, but Photomatrix has more options, and my big
> problem is the size of the preview, I can't see details so when I
> stitch the final result I alwways see that my settings was not
> correct.
> are you planning to add more options for Tonemapping in PTGui9? or
> maybe is possible to add Photomatrix like a plugIn for PTGui...
> is possible to zoom an area to see details in the tonemapping editor?
You could render/stitch your 16bit file first, and then use the result as
a source.
Then you can "zoom in" on areas to test different settings by setting
output to rectilinear, and then adjust cropping so that you only get a
smaller area.
A lot faster workflow?
There are many HDR tonemappers around, like PictureNaut.
There's also Artizen, which comes with PS plugins too.
And then there are some PS actions called "poor man's HDR" or something
similar.
Try to Google it, and you'll probably find a lot of info.
> Matthew and Bj:
> I know that it's not a true HDR, and I have remarked that my goal is a
> tonemapped image, maybe it's possible to simply edit it in Photoshop
> to fake that effect, but I need the easiest and fastest way, and also
> automated... and I'm not an expert in Photoshop :(
> Tonemapping is easy to me, I just need to slide some controls and see
> what happens. And I think we can automate the process, I'm pretty sure
> PTGui can do that (but I don't know how), also there's a "Batch
> Processing" button in Photomatrix.
> I'm a beginner in this, and I'm sure I'm doing many things wrong but
> I'm reading and testing all your tips.
> Also I think there's a lot of confusing information in internet, maybe
> could be a good idea to put the best workflow for (fake)HDR with a
> single RAW inside PTGui FAQs.
--
Bjørn K Nilssen
Which may be fine if you intend to use the HDR for something, like IBL for
lighting a 3D scene.
The problem is that to make a viewable image you'll have to convert back
to LDR, which is limited by 0.0 -> 1.0, no matter how many steps you get
in between. And when you do that it doesn't really matter anyway, because
you're still limited by the original 14bits from the RAW file.
A 16bit file format is fully capable of storing any detail 14bits. You
still don't get the fine details in the deep shadows and bright areas with
one single shot.
Try to shoot a fairly dark apartment with bright sun outside the windows,
and see if you can fake a HDR with that data, with no overexposed windows,
and no black shadows!
Erik Krause
> I have tried what you said, opening my TIFF16, aligning, selecting HDR
> (omit the red warning) and tonemapped the result, I think this could
> work too, but I have a problem with tonemapping in PTGui
No one stops you from stitching to 16bit TIFF and processing the result
in photomatix (which is the correct name, not ..matrix).
You can even stitch single exposures and output an HDR file format if
this simplifies anything...
--
Erik Krause
http://www.erik-krause.de
michael crane
so is it true then that a 16bit tiff can hold all the information that
is in the CR2 file ?
mick
Erik Krause
> Correct me if I'm wrong, but one thing you are all missing, or have
> not mentioned at least, is the contrast setting / curve and clipping
> settings applied to the RAW file.
You're not wrong. That is exactly what Matthew writes: You have to know
your RAW converter in order to bypass these limitations. Not all RAW
converters are capable to do this.
Canon DPP f.e. is not. But Adobe Camera Raw of course is. The procedure
is simple: switch on highlight and shadow clipping and play with the
sliders just to the point where the clipped areas are minimized (or at
least anything is included you want to).
Don't bother about the contrast being low, it can be increased later be
it with a tone mapping operator or by simple large radius USM.
Erik Krause
>> > You can even stitch single exposures and output an HDR file format if this
>> > simplifies anything...
> so is it true then that a 16bit tiff can hold all the information that
> is in the CR2 file ?
Absolutely. In fact it can hold even more information, since a RAW file
is in linear space and 16bit TIFF usually is in gamma 2.2 space, which
means that levels are more evenly distributed. See
http://www.normankoren.com/digital_tonality.html#Human_vision
for a comparison.
michael crane
so what is the current perceived wisdom for work flow to extract best
information if having to take single photos and stitch panorama ?
nods to Ryan
mick
enridp
OK, now I know nothing again :)
I thought that saving as TIFF16 I was saving the full RAW information
(which is 12bits...)
I was searching more and I found this from Photomatix (not matrix(!))
FAQs:
http://www.hdrsoft.com/support/faq_photomatix.html#tips
It satys:
There are three techniques for using Photomatix Pro with one single
exposure taken in RAW mode:
· Technique 1: Open your RAW file in Photomatix Pro to tone map it
directly.
· Technique 2: Convert your RAW file into a 16 bits/channel image in
your favorite RAW converter, open it in Photomatix Pro, and tone map
it.
· Technique 3: Create two or three exposures in your RAW converter and
combine them in Photomatix Pro (or Photomatix Light) as it they were
"real" bracketed shots
Technique # 3 often gives the best results. Also, it has the
advantage of working with Exposure Fusion.
Also they recommend to use low ISO and overexpose the shoot, then
subexpose the RAW with Camera Raw for exmaple, and then tonemap
Look this examples:
http://www.hdrsoft.com/examples.html#stmark
I have tried Topaz Adjust too and I could obtain a nice result from my
TIFF16 stitch, but I think is not possible to batch the processing
with Topaz Adjust.
I'm very very confused, and I'm sure there are more people like me,
maybe this is a good chance to leave this topic clear for everyone.
This is my conclusion for the moment:
1) We can't make a true HDR, 32 bits information from a single 12bits
RAW
2) Anyway, we can make a fake HDR with a single RAW
3) The best result for a single RAW is obtained following these steps:
a) Open RAW files with Adobe Camera Raw (or any other capable to make
the nexts things)
b -> optional) I don't know if this is neccesary :(
Set this:
-------------
black clipping = 0
recovery = 50
contrast = -50 (?)
c) In ACR set:
EV = 0 and save as tiff16 (or jpg8)
EV = +2 and save as tiff16 (or jpg8)
EV = -2 and save as tiff16 (oj jpg8)
d) stitch in PTGui the 3 exposures (we need to change the exposure
inside Image Properties so PTGui recognizes and link them like "true"
HDR)
e) export the stitch like HDR and open with Photomatrix
f) tonemap your HDR and save as JPG
Please correct me if I'm wrong (again :P)
Erik Krause
> I thought that saving as TIFF16 I was saving the full RAW information
> (which is 12bits...)
Not with the default settings of your RAW converter.
> I was searching more and I found this from Photomatix (not matrix(!))
> FAQs:
>
> http://www.hdrsoft.com/support/faq_photomatix.html#tips
> It satys:
> There are three techniques for using Photomatix Pro with one single
> exposure taken in RAW mode:
> � Technique 1: Open your RAW file in Photomatix Pro to tone map it
> directly.
In this case you must take it for granted that photomatix RAW conversion
extracts all the dynamic range from the image. I would assume a decent
RAW converter could do better.
> � Technique 2: Convert your RAW file into a 16 bits/channel image in
> your favorite RAW converter, open it in Photomatix Pro, and tone map
> it.
In this case you need to know how and whether your RAW converter will
extract all the dynamic range.
> � Technique 3: Create two or three exposures in your RAW converter and
> combine them in Photomatix Pro (or Photomatix Light) as it they were
> "real" bracketed shots
> Technique # 3 often gives the best results. Also, it has the
> advantage of working with Exposure Fusion.
All RAW converters should have the ability to extract at +/- 2 EV. Since
you don't need to know much more, this might indeed give the best results.
> Also they recommend to use low ISO and overexpose the shoot, then
> subexpose the RAW with Camera Raw for exmaple, and then tonemap
They surely not recommend to overexpose, since clipped highlights from a
digital sensor can't be recovered. However, most cameras don't show a
correct histogram or correct clipping display if you shoot RAW. Most
often there is still room, even if you expose "to the right" (means
such, that the histogram touches the right side but there is no clipping
yet).
> I'm very very confused, and I'm sure there are more people like me,
> maybe this is a good chance to leave this topic clear for everyone.
As usual there are more than one ways to the desired result. It is good
to know and to have tried all of them in order to choose the right one.
No advice can replace experience.
michael crane
On Sun, November 14, 2010 8:14 pm, Erik Krause wrote:
> All RAW converters should have the ability to extract at +/- 2 EV. Since
> you don't need to know much more, this might indeed give the best results.
so why are you saying a tiff can have all the information that is in a CR2
and still need to go through this process ?
mick
Erik Krause
>> > All RAW converters should have the ability to extract at � 2 EV. Since
>> > you don't need to know much more, this might indeed give the best results.
> so why are you saying a tiff can have all the information that is in a CR2
> and still need to go through this process ?
The fact that a TIFF *can* have all the information does not necessarily
mean that it *will* have it. Whether it will have it depends on the
abilities of the RAW converter *and* the user. So if someone is able or
not willing to find out how to extract the full DR from a CR2 the fake
exposure technique might be recommended.
Matthew Rogers
Matt
michael crane
On Sun, November 14, 2010 8:38 pm, Erik Krause wrote:
> Am 14.11.2010 21:22, schrieb michael crane:
>>> Since
>>> > you don't need to know much more, this might indeed give the best
>>> results.
>> so why are you saying a tiff can have all the information that is in a
>> CR2
>> and still need to go through this process ?
>
> The fact that a TIFF *can* have all the information does not necessarily
> mean that it *will* have it. Whether it will have it depends on the
> abilities of the RAW converter *and* the user. So if someone is able or
> not willing to find out how to extract the full DR from a CR2 the fake
> exposure technique might be recommended.
what is the best way to extract the full DR from the CR2 ?
mick
michael crane
On Sun, November 14, 2010 8:39 pm, Matthew Rogers wrote:
> Why ? Just because the data is contained in the file doesn't mean it's
> actually visible. Otherwise there'd be no need for any type of adjustments
> in a converter. The RAW converters job is basically to optimise the visual
> representation of the data contained within the file.
as I have understood it in the past I have used FDRtools and photomatix
but as I do not have a device to view HDR I didn't see the point as I
couldn't make any decisions how things looked like that. Are there devices
available to work in HDR ?
mick
Erik Krause
> so what is the current perceived wisdom for work flow to extract best
> information if having to take single photos and stitch panorama ?
In addition to what I wrote some hours ago [1] you need to use a linear
curve. Most raw converters use a S-shaped curve to boost contrast.
[1] http://tinyurl.com/23v4cxp
Ken Warner
to output 16 bit tiff and make sure that the image is converted
as shot. In other words don't tweak the raw converter's EV setting
to a value other than 0.
But then that raises the question -- what about color temperature
and saturation and gamma correction any of which can raise or lower
the image information out of bounds.
michael crane wrote:
> On Sun, November 14, 2010 8:38 pm, Erik Krause wrote:
>> Am 14.11.2010 21:22, schrieb michael crane:
>>>>> All RAW converters should have the ability to extract at Ä… 2 EV.
michael crane
On Sun, November 14, 2010 9:11 pm, Erik Krause wrote:
> Am 14.11.2010 18:27, schrieb michael crane:
>> so what is the current perceived wisdom for work flow to extract best
>> information if having to take single photos and stitch panorama ?
>
> In addition to what I wrote some hours ago [1] you need to use a linear
> curve. Most raw converters use a S-shaped curve to boost contrast.
>
> [1] http://tinyurl.com/23v4cxp
yes I read that.
I was hoping for a work flow.
mick
Erik Krause
> as I have understood it in the past I have used FDRtools and photomatix
> but as I do not have a device to view HDR I didn't see the point as I
> couldn't make any decisions how things looked like that. Are there devices
> available to work in HDR ?
Yes, of course: http://en.wikipedia.org/wiki/DR37-P
BrightSide, the manufacturer of above monitor was acquired by Dolby.
They now seem to have a monitor, too: http://tinyurl.com/2akdy4r
Eric O'Brien
It is my intention and expectation to stitch each exposure to
individual files, then merge them. Maybe by creating a HDR file, then
tone mapping or maybe by Exposure Fusion.
That being the case, I consider the exposure steps to actually be
*intermediate* versions. So trying to get each of those to "look just
right" would actually be the wrong way to go.
In particular, pretty much the only change I make in the raw
conversion is to neutralize the color balance. I usually take an
additional shot of a neutral card, so I use that as a basis for what
is "neutral" for the entire set. I'll also correct chromatic
aberration, and probably add a small amount of sharpening.
Beyond that, all other adjustments are OFF. In particular, this means
using a *linear* curve. Also, neutral "contrast" "brightness" and so
forth. The point being to get as neutral a starting point as possible
for what goes into the Tone Mapping or Fusion software. Whatever
comes out of *that* should probably be considered an intermediate step
too. So I get as close as I can in there, then open the result in
(for me) Photoshop and fuss with it some more.
My temptation is to say that even if you are shooting only *one*
exposure, that it would be best to leave the raw conversion settings
"flat" or "neutral," and wait until you have the final stitched image
before you go to work adjusting curves or whatever. (I could be wrong
here... I haven't tried that route.)
eo
> --
floz
linked to earlier. Very enlightening (pun intended).
I always get annoyed when people only know half facts but present them
as if they were ultimate truth. Specifically, I come accross this
confusion fairely often, where people think 16 bit automatically means
*broader* dynamic range covered. However, all it means is *finer
graduation* of what would be covered if it were just an ordinary 8 bit
image.
Cameras, in turn, be it with 12bit or 14bit A/D converters (and
there's much evidence that 14bit actually makes little sense beyond
marketing claims) *do* cover a rather broader dynamic range with their
bit depths. It still does make sense to use 16 bit TIFFs though if
there is more (external) processing to be done after raw conversion,
as RAW conversion itself is interpolation, and with 16 bit a finer
graduated interpolation can be produced, if my understanding of all
this is correct.
floz
interesting thing I just stumbled upon...
=> http://www.dxo.com/intl/photo/dxo_optics_pro/whats_new#singlehdr
While I'm not saying marketing is a justification of anything, I still
think the guys at DxO know what they are doing, so single shot HDR
can't be that much of a non-sense...
(Again, this is not for me, I think the resulting effect is rather
outdated and never was appealing in the first place, but for those who
fancy it, why not.)