How to create a good HDR photo?


Status
Not open for further replies.
Well I post this in response to some posting on using a single RAW, manipulate the levels, obtain 2 more sets of over and under exposed images and treated the 3 of them using HDR formula.

I personally find this rather silly because RAW only means more information retained and without compression, would means you get pixel to pixel colour values of 1 single snap and does not means you capture the highlights or shadow which may be over or under exposed

True enough recover highlight for those areas that are not totally washed out and recover shadow for those areas which are not totally saturated at zero. But manipulating the level is not going to make wash you values back to their correct values nor the shadow area.

Think about this. I colour an area 50% gray. then I perform level manipulation to produce 75% gray and 25% gray, does it means I succeed in recovering wash out or shadow areas ? It's just a mathematical formula to create an image, but capture an image. Obvious enough is one is just trying to fake for the sake of doing so.

HDR is about tackling the deficiency of linear sensitivity nature of the sensor or film as oppose to the non-linear vision that our eyes can offer us. Using multiple images overly exposed to recover shadow details and under exposed to recover wash out highlights.

If one bend of creating HDR for the sake of doing so, perhaps a direction is make your original image looks nicer.

Just my 2 cents worth.

Thanks :)
 

If one bend of creating HDR for the sake of doing so, perhaps a direction is make your original image looks nicer.

i agree with this, but there ARE uses for hdr from single raw -

the swan photo on my pbase is an example of that - the original photo couldn't quite bring out the rays, and yes, perhaps i could happily spend a few days dodging and burning carefully to bring them out, but hdr made it simpler, and even threw in some extra benefits like a slight softening (yes, softening, hurhur) of the image.

i won't say NO all the time for the sake of saying NO either. ;) have you tried hdr with a single raw? do you understand what it does - or are you preaching what you have not tried?

that said, understanding light is far more important than arguing about technical stuff like this. raw format allows you retain extra information from the scene, like it or not. a jpg's blown out highlights are permanently lost, but you can recover MORE highlight detail from the un-blown out segment from a raw file. and so long as one knows what he is doing, what his aim is, then i would say there is nothing wrong absolutely in using the method, be it hdr from single raw, or any other photoshop/processing method available in the whole wide world.
 

Last edited:
Well... I won't say NO to NO as well. But regarding what you mentioned, can't that extra light ray be brought out by regional contrast increase and regional brightness adjustment ?

Of course, I'm not questioning on how u do your work and perform your art. Just casually asking only.

You have your rights and options to choose whichever way you feel comfortable to PP your photos :)

My earlier statement was just trying to instill the concept of working on the original image to bring out the goodness rather than trying tricks that is not meant to be that way in the first place. (^.^)
 

can't that extra light ray be brought out by regional contrast increase and regional brightness adjustment ?

My earlier statement was just trying to instill the concept of working on the original image to bring out the goodness rather than trying tricks that is not meant to be that way in the first place. (^.^)

? i don't understand. i specifically mentioned that yes, it could be done. maybe if i posted the original file to provide a good comparison between before and after, you'll see what i mean.

how is hdr a "trick"? isn't burning and dodging also a trick, if one wants to be pedantic? or are you simply being argumentative for the sake of being so? one can only be purist or non-purist, it reeks of double standards when you call "hdr from single raw" a trick, when you would not call burning and dodging (or regional contrast/regional brightness adjustment, if you will) a trick.. ;)

if that is your view, then what do you think of hdrs where the hierachy of light is lost? are they "trying tricks that is not meant to be that way"? should a tree be as bright as or brighter than the sky? should it even glow unnaturally? (i.e. haloing)

last of all, have you tried hdr from single raw before to make a valid comparison for yourself? or are you rattling off something which you view, but have not experienced? :)

if it gives you additional DR, i say, go ahead. after all, there are so many situations where a single shot rather than a conglomerate of many shots would do better, even with photomatix's later versions with anti-ghosting. take for example the swan shot again. do you think the swan will keep still for 6 or 7 shots, even if i am the quickest button pusher in the world?
 

Last edited:
just taken with a k100d

exif is still there, it's 100% crop from the same file with same settings, shot in raw.

jpg direct from raw, along with attempts to burn in details in a "localised contrast/brightness adjustment" bid:

2947074016_a2b5330686_o.jpg


adjustment in RAW converter, THEN output again.

2947073722_d868126654_o.jpg


roughly the same haphazard burning/dodging applied. the same? no, i don't think so. i don't really have the time, else i'd go out and capture a few more clearer examples just for you.. ;)
 

Last edited:
Not sure if this has been discussed before. Anyway this is what I did with my single raw file.
a. ensure that the photo is correctly exposed and convert the raw file to jpg
b. decrease the exposure and convert into jpg
c. increase the exposure and convert into jpg
d. use you favorite hdr tool and combine the 3 jpg files into HDR

some of samples that I did using this method http://www.flickr.com/photos/jarrodstone/sets/72157607923053766/
 

Hi guys,

This is my first attempt at using HDR. I took 3 JPGs with different exposures ( +- 2 ) and combined using Photomatix. Do tell me what you think! I didn't really like the composition of the shot but it was the best I could get from the front of my house :P

2950804399_13ace4a671_b.jpg
 

I think we have some miss understanding here. I suggest that we go back to the basics and clarify some terms here. Goggle on "HDRi/HDR" and "tonemap/tonemapping" to clarify the terms.

Any way "HDR Generation" and "RAW tonemapping" are 2 separate commands even in Photomatix. They are different and I think this thread is about HDR right?:)

Just hope this will make the issue less complicated here.;)
 

Not sure if this has been discussed before. Anyway this is what I did with my single raw file.
a. ensure that the photo is correctly exposed and convert the raw file to jpg
b. decrease the exposure and convert into jpg
c. increase the exposure and convert into jpg
d. use you favorite hdr tool and combine the 3 jpg files into HDR

some of samples that I did using this method http://www.flickr.com/photos/jarrodstone/sets/72157607923053766/

I don't disagree with your approach. But there's a much simpler workflow than this; there's no need to convert to JPG and then make +2EV or -2EV files to create the 3 images for HDR merging. Take the single RAW file or a single 16-bits image like TIFF and then open in Photomatix the rest is up to you. One main reason of against using JPG in merging is you may lose some colour details/highlights inherrent of the JPG format.

Enjoy!
 

Last edited:
Hi, LifeInMacro is correct. Just open the raw in Photomatix, then apply the "ToneMap" command. It should produce better and more straight forward solution. Btw, this is what I mean by Tonemap, its not HDR (thats why you don't use Generate HDR command.:)

Not sure if this has been discussed before. Anyway this is what I did with my single raw file.
a. ensure that the photo is correctly exposed and convert the raw file to jpg
b. decrease the exposure and convert into jpg
c. increase the exposure and convert into jpg
d. use you favorite hdr tool and combine the 3 jpg files into HDR

some of samples that I did using this method http://www.flickr.com/photos/jarrodstone/sets/72157607923053766/

I don't disagree with your approach. But there's a much simpler workflow than this; there's no need to convert to JPG and then make +2EV or -2EV files to create the 3 images for HDR merging. Take the single RAW file or a single 16-bits image like TIFF and then open in Photomatix the rest is up to you. One main reason of against using JPG in merging is you may lose some colour details/highlights inherrent of the JPG format.

Enjoy!
 

just my thought, guess those hard anal ppl shouldn't be doing photography in the first place.

ppl who restrict themselves to certain genre or 'this is wrong, that is right' ppl i feel shouldn't take up photography in the first place.

photography can be regarded as art and art holds no boundary.
 

err...washout means what huh? :p

okie, more C&C anyone to confirm? Please give me your advice how I could improve it...

thanks hor :angel:

the foreground is too dark or underexposed with very little detail.
 

just my thought, guess those hard anal ppl shouldn't be doing photography in the first place.

ppl who restrict themselves to certain genre or 'this is wrong, that is right' ppl i feel shouldn't take up photography in the first place.

photography can be regarded as art and art holds no boundary.
everyone has a right to a point of view

it is ironic that you allow art to have its own point of view

but people can have no point of views on art. does it not seem hypocritical to yourself? :o

in fact, i think i prefer "hard anal people" to "all encompassing care bears".. photography is all about points of view.. and if one has a "hard anal pov" usually he is going to have a more unique style than the care bear.
 

? i don't understand. i specifically mentioned that yes, it could be done. maybe if i posted the original file to provide a good comparison between before and after, you'll see what i mean.

how is hdr a "trick"? isn't burning and dodging also a trick, if one wants to be pedantic? or are you simply being argumentative for the sake of being so? one can only be purist or non-purist, it reeks of double standards when you call "hdr from single raw" a trick, when you would not call burning and dodging (or regional contrast/regional brightness adjustment, if you will) a trick.. ;)

if that is your view, then what do you think of hdrs where the hierachy of light is lost? are they "trying tricks that is not meant to be that way"? should a tree be as bright as or brighter than the sky? should it even glow unnaturally? (i.e. haloing)

last of all, have you tried hdr from single raw before to make a valid comparison for yourself? or are you rattling off something which you view, but have not experienced? :)

if it gives you additional DR, i say, go ahead. after all, there are so many situations where a single shot rather than a conglomerate of many shots would do better, even with photomatix's later versions with anti-ghosting. take for example the swan shot again. do you think the swan will keep still for 6 or 7 shots, even if i am the quickest button pusher in the world?

To add on,

Well HDR is really, just like any other action script in photoshop - albeit the requirements are either working off differently exposed JPEGS or a single raw; and there are instances whereby its impossible to capture manually bracketed shots due to various reasons; e.g. no tripod, moving subjects etc..

The following picture was treated from a single RAW image:

2673664217_df9150a417.jpg


Before i discovered HDR i used to work off a single photo and try to bring out its DR by doing color burn; multiply and various blending techniques.

E.g. A typical landscape scene, i would add multiply and color doge then apply layer masking. Subsequently apply shadow/highlight recovery treatment and more layer masking.

Essentially HDR does pretty much the same thing except it does it alot faster.

Many might dispute the realism of HDR'ed photos, but i think its just a personal preference.
 

in fact, i think i prefer "hard anal people" to "all encompassing care bears".. photography is all about points of view.. and if one has a "hard anal pov" usually he is going to have a more unique style than the care bear.

How true. In good stride, we can probably get better feedback from their anal pov.
( Even if we have opinions of their rigid dogmatism )

2673664217_df9150a417.jpg


Many might dispute the realism of HDR'ed photos, but i think its just a personal preference.

But HDR wise this looks pretty true to my eye bro.:thumbsup:

Ryan
 

actually, just want to clarify one point... HDR or HDRI is not a script or a program function that somehow gives your image more dynamic range... it is an image format that allows the expression of a wider dynamic range, usually in 32bit floating point format... what the HDRI programs do is that they combine multiple bracketed shots, and the wide dynamic range represented in there shots, into a single HDR image (note: one cannot save a HDR image into a regular TIF or JPG file... it has to be tone mapped into a discrete 8bit or 16bit image first)...

what one sees on screen is not HDR because there are no commercially available screens at the moment that can display that amount of dynamic range... so, what programs that handle HDRI do is that they have to use something called tonal mapping to try to squeeze the image so that it may be shown on a monitor... but note that this is just for display purpose and not baked into the image until one does a conversion into a discrete 8bit or 16bit format...

following up from what was mentioned here about single RAW conversions to "HDR" vs. optimizing a single RAW exposure, what can be said is that we can probably get good results from both methods... they are just two ways of getting something done... in less tricky images, optimizing the single processing of a RAW file would probably be easier, but in more complex lighting conditions, the "HDR" method might be better in that it is working in floating point and would allow the tonal values to be stretched more easily (if working the 32bit floating point image before it is tone mapped into discrete 8bit or 16bit images)... YMMV

as for "reality" versus supersaturatedandcontrasty, well, it all depends on what one is trying to show... me, I'm more interested in whether it is effective in what it is trying to show or whether it is just used as a gimmick that does not contribute to the image, and how well the image is processed :)
 

Last edited:
yep, hdr "compresses" the dynamic range.. :)

which is why we need the tone mapping step in photomatix even after the hdr has been generated.. :)
 

This is true. That's why if you have a single RAW, you won't have 32bit of data to be compressed, but as much as only 12-14bit of data for most Digi-Cam. This is why more shots are needed to cover more range. However, if the scene DR is not that wide to start with, then forget about HDR, single RAW tonemap will be enough to pull out all the details you have in the 12-14bit data.:)

actually, just want to clarify one point... HDR or HDRI is not a script or a program function that somehow gives your image more dynamic range... it is an image format that allows the expression of a wider dynamic range, usually in 32bit floating point format... what the HDRI programs do is that they combine multiple bracketed shots, and the wide dynamic range represented in there shots, into a single HDR image (note: one cannot save a HDR image into a regular TIF or JPG file... it has to be tone mapped into a discrete 8bit or 16bit image first)...

what one sees on screen is not HDR because there are no commercially available screens at the moment that can display that amount of dynamic range... so, what programs that handle HDRI do is that they have to use something called tonal mapping to try to squeeze the image so that it may be shown on a monitor... but note that this is just for display purpose and not baked into the image until one does a conversion into a discrete 8bit or 16bit format...

following up from what was mentioned here about single RAW conversions to "HDR" vs. optimizing a single RAW exposure, what can be said is that we can probably get good results from both methods... they are just two ways of getting something done... in less tricky images, optimizing the single processing of a RAW file would probably be easier, but in more complex lighting conditions, the "HDR" method might be better in that it is working in floating point and would allow the tonal values to be stretched more easily (if working the 32bit floating point image before it is tone mapped into discrete 8bit or 16bit images)... YMMV

as for "reality" versus supersaturatedandcontrasty, well, it all depends on what one is trying to show... me, I'm more interested in whether it is effective in what it is trying to show or whether it is just used as a gimmick that does not contribute to the image, and how well the image is processed :)

yep, hdr "compresses" the dynamic range.. :)

which is why we need the tone mapping step in photomatix even after the hdr has been generated.. :)
 

However, if the scene DR is not that wide to start with, then forget about HDR, single RAW tonemap will be enough to pull out all the details you have in the 12-14bit data.:)

Yup. And for scenes that actually do not need the extended range to start with will look wierd when u purposefully HDR multiple exposures to stretch the DR.

Ryan
 

Status
Not open for further replies.
Back
Top